This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Estopedist1 (talk) 07:19, 5 March 2024 (UTC)
Guidance on membership data maintenance
My Canadian colleagues and I have been uploading performing arts associations' members datasets to Wikidata for the last two years. We uploaded 7 such datasets (and soon an 8th dataset) and have been maintaining this data since. Over the course of this work, questions arose on how best to populate member of (P463) statements that are accurate and meaningful at the time of the initial upload and remain so over time. We discussed these questions with the WikiProject Performing arts (Q59956276) community and we arrived at a consensus on a set of good practices for documenting membership statuses in Wikidata. While this guidance was developed for performing arts use cases, we realize it could be useful in other domains as well. Publishing it within the WikiProject Performing arts (Q59956276) consequently doesn't make sense. We should publish it somewhere where the broader Wikidata community will be more likely to search for and find such information.
Does this proposed set of good practices make sense? Do you have further input to provide? If so, please add comments or suggested edits to the Google doc.
Adding this guidance to the "member of" discussion page would definitely be helpful for Wikimedians seeking extra guidance about using the property. As for the WikiProjects, it doesn't look like Organizations has a Data Structure section per se. However, the guidance could be added to the Discussion page, if members of that WikiProject would find it useful. Dlh28 (talk) 19:30, 23 February 2024 (UTC)
Can I use P518 to qualify a P180 statement to clarify what part of the image is meant? For instance if an action is depicted and the subject of the action isn't the main subject of the image
Suppose I have a Qx with a statement P180 Qy. I understand this to mean that Qx is an image that depicts Qy. I have seen a qualification of P518 Qz on such a statement used to clarify which part of the image depicts Qy. Suppose however that Qy is an action or process that may have a subject, Qw. Can I qualify the statement with P518 Qw to indicate this? If not, how else could I indicate that Qw is the subject of the action Qy? Theanswertolifetheuniverseandeverything (talk) 12:25, 1 March 2024 (UTC)
Q123415663 can be merged with Q7699181 (or, Ten Days in a Mad-House at WikiData)
But I get conflicts when I try copying statements over -- like it claims that adding an Internet Archive ID conflicts with a work called a "literary work". This makes no sense to me, and none of the help or discuss links seem to clarify why, nor what to do about it. Help! Gnuish (talk) 19:26, 1 March 2024 (UTC)
They cannot be merged because they are not the same thing. One is the work (in any form) by the author, the other is a specific publication of that work. Such items should never be merged. Nellie Bly's jounralistic exposé was originally published in the newspapers, which is one edition of the work, with its own publication data. Ten Days in a Mad-House (Q123415663) is the 1887 collected paperback edition, which also includes two other works by Nellie Bly. Ten Days in a Mad-House (Q7699181) is the collective umbrella for all editions of Nellie Bly's series of articles describing her experiences in a New York sanitorium for the insane. --EncycloPetey (talk) 19:29, 1 March 2024 (UTC)
Thank you, that's a great explanation! And thanks for reverting my changes. Now, where can a simple explanation like yours be put, so that the next confused person will find it when WikiData starts complaining about deeply technical "conflicts". This all started when I tried to add the Wikisource link into the WikiData entry for the book (that was linked from the enwiki article about the book). The help/discuss links for the conflict notice just led me to a page that talks generally about constraints, and which never mentioned anything about "literary work"s or why they conflict with an "edition" in Wikidata. Gnuish (talk) 19:53, 1 March 2024 (UTC)
Wikidata:WikiProject Books is the place where discussion happens. There is already a set of properties set up there are differentiated between literary works and their editions. But the underlying reason for the distinction is in the nature of library database structure, which is explained at w:en:Functional Requirements for Bibliographic Records on the English Wikipedia. Anyone wishing to add entries into a bibliographic database of library quality would need to look up and learn the basic international standards, and there's no one place to point to for that. --EncycloPetey (talk) 21:44, 1 March 2024 (UTC)
NFSA title ID (P11948) should allow creative works, not just films
I was happy to find there is a property already for NFSA title ID (P11948) - thanks! However, it has the subject constraint that it can only apply to instance of (P31)film (Q11424), but the database also includes hundreds of digital works that are not films but, for example, digital artworks or hypertext fictions. The database itself uses the term Multimedia but since these works are quite varied, perhaps we could add creative work (Q17537576) to subject type constraint (Q21503250)? Please let me know if there is a better way to suggest this kind of change - I'm still pretty new around here.
Thank you for your interest @Lijil, I can confirm that the NFSA uses its Title IDs to represent "creative works" generally (films, TV shows, albums, games etc) so I think you suggestion is a good one. Pxxlhxslxn (talk) 20:43, 1 March 2024 (UTC)
Should we make sure that the imdb entry is not used on either one of the "Rune Bergman"s or is the entry added to them both as an exception or what do we/I do? MythsOfAesop (talk) 01:00, 1 March 2024 (UTC)
We encounter this a lot at VIAF and LCCN. Sometimes the errors are introduced by us when they import Wikidata data. --RAN (talk) 20:18, 2 March 2024 (UTC)
Sorting values in a property
I am not sure if this question has been raised before. Is there a way to sort values in a property? Other than removing all the values and readding them in the desired order. For example, visitors per year (P1174) in Louvre Museum (Q19675). Currently, it is shown in the order they were added to the item. The "correct" sorting could be by point in time (P585), and showing the preferred rank (green value) the first one, giving more visibility to it. Perhaps, Wikidata developers could add a feature so we can define the (by default) desired ORDER BY in a property? Even more, they could add some feature (similar to class wikitable sortable) in the visitors per year (P1174) layout (and any other property) so we can click and sort the values by year, by quantity, by date retrieved, by... Thanks. Emijrp (talk) 08:59, 2 March 2024 (UTC)
The ordering of properties themselves (P31 etc.) was built-in to the Wikibase extension at some point. This used to be handled by a JS gadget but that had poor performance. If one were to make a gdget for sorting claims by ordinals, dates etc. it is likely to face similar performance issues.
The visual ordering doesn't matter for things like queries and data consumers. It might be hard to manually check for missing ordinals in a long list when one is editing an item, but that's something that can be easilly checked for, if anyone is looking for a gadget idea. It might even be a better idea for a new property constraint e.g. "contiguous set constraint"! I also think it's fair to say wikis attract the kind of people who would be annoyed by improper ordering. :) Infrastruktur (talk) 15:36, 3 March 2024 (UTC)
timber framing (Q6438054) and timber framing (Q629000) both have German wikipedia articles. One talks about a general construction method used since ancient times while the other talks about one modern implementation of the method (implying we may have other items for the different modern method). The items use different words in German but appear to overlap in English. From Hill To Shore (talk) 07:49, 4 March 2024 (UTC)
Redirects in SPARQL queries
I've got some SPARQL queries that are returning redirect pages instead of the items they redirect to, which makes the results harder to use. Is there any way to get the query service to follow redirects? dseomn (talk) 18:42, 1 March 2024 (UTC)
This is handled by a bot, but I forget which one, and how often it runs. The only way a redirect show up in a query is if was recently added.
If you need to query newly merged items a lot you can ask for help on WD:RAQ to have the query consider redirects. This should work if they are only likely to show up in one or two places, otherwise it might not be practical. Removing redirects from the result is also an option. If you want to try yourself, the predicate for redirects is owl:sameAs e.g. wd:Q123 owl:sameAs wd:Q456 . Infrastruktur (talk) 20:12, 1 March 2024 (UTC)
According to Help:Redirects#Links_to_redirects it's User:KrBot and "within a few days". But it's pretty annoying to have broken queries for those few days, every time anything being queried was recently merged.
I looked at owl:sameAs, but it looks difficult to use in all the places that redirects are possible. E.g., adding owl:sameAs to the property path would remove normal non-redirect results, and adding owl:sameAs? to the property path would give both the redirect and the target. Getting both non-redirects and redirect-targets without also getting redirects seems like it would make the query a lot more complicated. And removing redirects would break the query in a different way, by giving incomplete results.
Is there any chance of something server-side to just treat the data as if the statement value were already the target instead of the redirect? dseomn (talk) 20:57, 1 March 2024 (UTC)
Hard question. I would very much like to know this myself. There is something called inference and reasoners. This is turned off on Wikidata because it is slow, but if I understand correctly having an advanced reasoner configured will basically magically follow the redirect. If anyone happens to sit on an example for Apache Jena that would be sweet. Infrastruktur (talk) 00:48, 2 March 2024 (UTC)
We already have ISCO-08 occupation class (P8283). Is that not sufficient to model the information from ISCO-08? (Caveat: I haven't made any effort to grasp the ISCO-08 data model, so I really don't know the answer to this.)
Here's your quick overview of what has been happening around Wikidata over the last week.
Discussions
Open request for adminship: Mike Peel (scheduled to end after 9 March 2024 13:08 (UTC))
Closed request for adminship: Madamebiblio requested for admintool to fight against vandalism, and the request was closed as successful.
New requests for permissions/Bot:
Mfchris84-Bot - Task: Will run scripts on toolforge to automate and schedule edits on bibliographic items. Starting with Swiss online magazine Syntopia Alpina (Q112206291), then continuing by adding swisscovery edition ID (P9907) to version/edition items.
BorkedBot_12 - Task: Update number of viewers/listeners (P5436) with the qualifier YouTube channel ID (P2397) for YouTuber content creators and influencers.
LocodeBot - Task: Match UN/LOCODE (Q499348) to Wikidata entries by ISO 3166-2 region code, name and distance (if defined), important for UN organisations.
Join the Queensbridge Tech Lab edit-a-thon - create and enrich local diverse heroes in Queens, NYC as part of the NYC Open Data Week. March 16, 10:00 - 16:00 EST (UTC -5).
Empowering BHL Staff with Wiki Education’s Wikidata Certificate Course - The Biodiversity Heritage Library blog celebrates committee members completing a Wikidata Certification course led by Will Kent, as part of the Wiki Education. Participants levelled up SPARQL queries, visualisations, federation, modelling and bulk loading and editing.
WiQuizz - from Nuxt and SPARQL WiQuizz allows you to create or play quiz games where content is generated from Wikidata. It could be a fun way to get people interacting with Wikidata!
Other Noteworthy Stuff
Claim Analysis Toolkit, collection of Observable notebooks to visualize simple Wikidata claims
date popularized (point in time the subject became well known to the public, if different from its inception)
Bild der Person bei der Tätigkeit (picture of a person in action, especially for a sportsperson, visual artist, musican, actor. P18 is normally used for portraits)
Newest WikiProjects: Te Papa Research Expeditions - create and enrich metadata and trial research expeditions schema for Te Papa, Dominion and Colonial Museums.
Please discuss this edit, there are many points in it: 1) can we (should we) keep history of a position in the item? 2) please estimate use of P31. As for me, I would use P279 for such values. 3) Should we have female forms of a title? Overall, I would suggest to undo this edit. Infovarius (talk) 13:19, 5 March 2024 (UTC)
The Universal Code of Conduct Coordinating Committee (U4C) is a global group dedicated to providing an equitable and consistent implementation of the UCoC. Community members are invited to submit their applications for the U4C. For more information and the responsibilities of the U4C, please review the U4C Charter.
Per the charter, there are 16 seats on the U4C: eight community-at-large seats and eight regional seats to ensure the U4C represents the diversity of the movement.
Read more and submit your application on Meta-wiki.
essentially because of the redundancy. We know "Milrose Games" are a recurring sport event on its item, this implies any of its instance are edition of recurring sport event, so it’s essentially useless to state it on all the instances. We just need
@Vicarage that was my first thought but it seems that with QS one has to remove broken reference first and that then add the fixed one, right ? Kpjas (talk) 09:02, 6 March 2024 (UTC)
Regrettably some Polish sources are rather irresponsible and change or take down the source without the proper provision and notice thus leaving hundreds/thousands links all over the net rotten.
I like the external_ID idea, that way if they change again, you just need to change the url format once at external_ID, and all the entries get fixed at once. --RAN (talk) 16:53, 6 March 2024 (UTC)
Wikimedia Canada survey
Hi! Wikimedia Canada invites contributors living in Canada to take part in our 2024 Community Survey. The survey takes approximately five minutes to complete and closes on March 31, 2024. It is available in both French and English. To learn more, please visit the survey project page on Meta. Chelsea Chiovelli (WMCA) (talk) 23:48, 6 March 2024 (UTC)
Allow false claims due to storage space limitations and a claimed need to be pragmatic
Example: Talk:Q1402057 , a user seems to suggest to conflate a publication (identified by ISSN) and an organisation (identified by ISNI).
The user is aware of the disctinctions, claims it to be subtle: "The newspaper vs the organization can be a subtle distinction, yes, but it's important to consider that Wikidata's storage space is limited and that we need to be pragmatic re: the usefulness and maintainability of different items." CV213 (talk) 15:30, 3 March 2024 (UTC)
TLDR your debate there, but if an organisation does one thing, then having separate items for both can be clumsy, as only one can get the wikipedia items, properties get duplicated , and querying the indirections is harder to write and slower to run. And a system that works for machines can be horrid for human contributors, so pragmatism is everywhere on WD. Vicarage (talk) 17:02, 3 March 2024 (UTC)
Absolutely split. For example: at The New York Times Company (Q2529982) and The New York Times (Q9684). I don't think we should worry about using up server space, we can always add more, other websites have billions of entries in their databases like Ancestry (30 billion records) and Familysearch (1.56 billion people). I would split them since they have their own external identifiers. The head of the organization is not the head of the newspaper, so split. The conflation of the two is giving errors: "An entity should not have statements for both ISSN and ISNI." --RAN (talk) 19:45, 3 March 2024 (UTC)
No one is suggesting to split every newspaper, and if done properly there should be no more maintained issues. --RAN (talk) 03:28, 7 March 2024 (UTC)
Where does the "2x" come from? Even if it were true, why would that be an issue, since the claims are clearly not "2x", since several are disjoint. CV213 (talk) 21:37, 3 March 2024 (UTC)
@CV213: splitting every newspaper into the newspaper and the organization that produces it roughly doubles the number of items? and i didn't say there would be 2x the data. I said it would be 2x the items. BrokenSegue (talk) 16:55, 4 March 2024 (UTC)
Conflating a newspaper and its publisher into one item is a very bad idea. The same goes for a TV station and its broadcaster, or an Internet service and its operator. These are different concepts that need their own item. Let's not take a bad example from Wikipedia, Wikidata are more granular. Note that a newspaper could be published by several publishers over time, just as a publisher could publish several newspapers. Jklamo (talk) 03:29, 4 March 2024 (UTC)
Shall be splitted. If one should make a refererence to an information for a specific date, person and so on, one shall make the reference specifically to the newspaper item not to any combined item mixing two completely different things. I can not see that maintenance is an issue here. Pmt (talk) 08:49, 4 March 2024 (UTC)
Tired of wasting time. Block evasion: sock puppet of User:Tobias Conradi. Blocked CV213 like the previous sock (Ac2wd). Multichill (talk) 18:56, 4 March 2024 (UTC)
I think you should not be adding these quantities in this way. Firstly, there could be various recipes of how to make this cocktail, and there is no reason to give preference to this particular recipe. Secondly, you have not specified how much volume of the cocktail you are making with 45ml of bourbon, so it is rather meaningless. "A few dashes" is so vague and imprecise that you should not attempt to define it as a volume. — Martin (MSGJ · talk) 10:26, 5 March 2024 (UTC)
My company, Object Management Group, has several entries as we are a family of consortia under the Object Management Group. I am not sure who originally wrote the content but it is woefully out of date.
Every time we try to update it, we are told we are too close to the content. The content is completely out of date, with most of the people listed are not at the company.
Hi there, my colleague Evan Birkhead tried to put comments into the Industrial Interent Consortium and was told he was too close to the contact as the company was paying him. I tried to add a page for Digital Twin Consortium and was rejected.
Industrial Internet Consortium - Wikipedia
I can't send the link it won't copy here.
My CEO just updated his name William Hoffman for our former CEO Richard Soley who passed away last year. William's name was linked to a list of people, none of whom is himself.
Long story short, can I make these updates? Yes I am the PR Director.
Assuming that you're talking about the English Wikipedia, there is guidance available at en:Wikipedia:Edit requests and en:Wikipedia:Suggestions for COI compliance. Basically: no you cannot make those changes, but you can request that they be made by somebody who does not have a conflict of interest. Discussions about Wikipedia are best had on Wikipedia, rather than here at Wikidata, though. M2Ys4U (talk) 03:25, 7 March 2024 (UTC)
it sounds like you are asking about English Wikipedia but this is the chat page for Wikidata. Unfortunately we cannot help you here. BrokenSegue (talk) 03:49, 7 March 2024 (UTC)
Save the date - Wikidata Leveling Up Days, April 5-7 and 12-14
The Wikidata Leveling Up Days are a series of online events and a collection of useful contents to help people get on board with Wikidata. The events will take place from April 5th to 7th and from April 12th to 14th, and will include a wide variety of presentations, workshops and discussions covering the basics of Wikidata: get started, browse and query Wikidata, enhance your editing experience, ensure data quality, and connect Wikidata with sister projects.
The event will give the occasion to both new and seasoned editors to get together and exchange ideas and tips. Some sessions are already planned in the schedule, but the program is open: community trainers and organizers are welcome to schedule their own self-organized sessions, at the time that suits them best, and in the language that would be accessible to their peers. If you’re interested in scheduling your own session, you’ll find guidance on this page.
The Leveling Up Days are also a collection of introduction videos, recorded in advance with experienced Wikidata editors, that will soon be available online. We will make sure that the videos are available ahead of the event, so people can watch them at their own pace. The videos will include translatable captions, so you will be able to help us make this useful content accessible to a wide range of Wikidata newcomers.
The online event is open to everyone without registration, but if you’re interested in attending, feel free to sign up here. You can also join our help desk channel where we will have Q&A sessions.
And last but not least: please help us reach out to people who may be interested in learning more about Wikidata: feel free to share or transfer this message to the channels, mailing-lists and groups where it could be relevant.
If you have any questions, feel free to reach out to me.
Strange that it was named "lake outflow" at the time of creation, when the proposal only used the term "outflow". The property likely should be renamed to what was actually proposed, just as P200 was renamed to match proposal. — Huntster (t@c)05:38, 5 March 2024 (UTC)
Entities using the outflows property should be instances of one of the following classes (or of one of their subclasses), but Sassenage hydroelectric plant currently isn't:
Rapid and blitz Elo ratings within chess players articles
Hello, on chess players articles the classical rating is the only rating provided. What about blitz and rapid ? These paces have a growing importance in the chess world and they are necessary to understand one's chess career. I talked to the creator of TemirovBot, which is in charge of updating those ratings in the biography infoboxes, and he told me he would try to implement the other ones if the authorization was granted. Hence, I am requesting this authorization.
If this isn't the right place, where should I open a discussion to have these questions answered ?
Thank you for your answers, Verbalemya (talk) 10:32, 5 March 2024 (UTC)
Hi, I created this with "instance of" as only "given name": Q124753851 Other examples I checked had more specific ones, like "male given name", "female given name" or "unisex given name". Anyway what I want to ask, is it OK to create with only "given name" if you don't know if it's only female or male, or unisex given name? Tehonk (talk) 21:44, 6 March 2024 (UTC)
According to Vojtěch Dostál (on the property proposal page), some IDs which can be reconciled via links to NL CR AUT ID (P691) present in some of its entries. Infrastruktur (talk) 14:21, 8 March 2024 (UTC)
I am about to do a mass-import of P5587 claims. What is the proper way to mark them as being inferred from the P906 claims? Thanks for quick answer. Infrastruktur (talk) 19:38, 7 March 2024 (UTC)
I don't have more right now. It's been evaluated with this script that only outputs the result that I have mentioned above. Could easily be expanded to include more results, though, but my free time is limited. —MisterSynergy (talk) 19:48, 6 March 2024 (UTC)
run an ipython notebook server locally? though not sure why you wouldn't want to log in. you're already logged in here and it's the same login (iirc) BrokenSegue (talk) 18:20, 8 March 2024 (UTC)
One can install Python locally, copy all the code into a .py file, install the requests package with pip, and run it locally then without any login. The advantage of PAWS is that this is a managed Python environment, ready-to-go with most important packages for interaction with Wikimedia content already installed. —MisterSynergy (talk) 22:17, 8 March 2024 (UTC)
If one inputs the wrong property, say inception instead of service entry there is no easy way to edit one into the other: there is an easy way to change the value of a property but not to change the property to which one refers. There should be an easy editing facility for this. David.Monniaux (talk) 13:57, 9 March 2024 (UTC)
Disregard, sorry I just figured out how to do this after posting this query. I clicked on the trash can icon which ungreyed my publish button. Ottawahitech (talk) 18:35, 9 March 2024 (UTC)
@Ottawahitech: If you think you may need to move sitelinks again, you may want to turn on the "Move" gadget in your preferences. "Move: Adds a move tool to the sitelink edit toolbar to move a sitelink to another item." From Hill To Shore (talk) 21:41, 9 March 2024 (UTC)
Preferred is a call for python through the pywikibot wikidata api.
Alternatively I can call a sparql script retrieving that single object.
The brutal way is to create the sitelink and let the creation fail, if the category is used somewhere else. (Will this overwrite a Commons-Gallery-Link with a Commonscat-Link?)
commonssite=pywikibot.Site('commons','commons')cat=pywikibot.Page(commonssite,"Category:Vienna Central Cemetery")if(cat):try:x=pywikibot.ItemPage.fromPage(cat)print(x.id)except:pass
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Herzi Pinki (talk) 21:17, 15 March 2024 (UTC)
@S8321414: If you want to move that discussion forward it would be good to actually make a clear proposal that can be voted up or down. If you want to add a new right than we need criteria about how to give out that right. ChristianKl ❪✉❫ 15:53, 10 March 2024 (UTC)
Cannot find a link I just created on a wikidata existing item
I was watching Fareed Zakaria on CNN this morning interview Noah Feldman about his new book "To Be a Jew Today". I thought I would check to see if this information is available on WD and discovered it was not. So I added a new item for the book (To Be a Jew Today (Q124808842)) and then added the new item to Noah Feldman (Q1959850) using a notable work (P800) link.
@Ottawahitech Statements follow a certain order, so if you add a statement and later come back to the same item, it will no longer be listed at the bottom. In this case, I see "notable work" listed below "educated at". You can also search within the text of the page (Ctrl + F) to make it easier to find the statement. –FlyingAce✈hello20:39, 10 March 2024 (UTC)
Interesting. How would one find out in what order statements are displayed on items? They do not seem to be ordered by the property number.
I use my Android Samsung Galaxy A20 for almost all of my Internet access, and discovered that I have to use my laptop to edit Wikibase items because of a lack of mobile support for adding statements to Wikibase items. This issue has been reported before.
I recently purchased a Oneplus 11 smartphone with 16 GB of RAM. My laptop has 4 GB of RAM. My wish is to run Wikibase locally on my Oneplus 11 to be part of the Wikibase Ecosystem as described at wikibase.world. The idea is also to run a digital agent on device to assist with adding and editing items to my local Wikibase.
Both use m.wikidata.org, and run a local instance of Wikibase on my Oneplus 11, along with a digital agent that can act as an interface to both the website and the local server.
There is an incredible amount of work being done training models that can recognize entities and relationships within text and images, and can create SPARQL queries.
I am a retired librarian, from St. Louis Public Library, and would need help installing and testing the software on my Oneplus 11, and could offer a limited amount of compensation to someone for their help. I have been thinking about putting a Help Wanted sign in the Computer Science building at Washington University, wustl.edu, to see if someone there might be interested. That campus is just a few miles from my home. Jkintree2 (talk) 20:49, 8 March 2024 (UTC)
Running wikibase on a phone sounds like a lot of work for not much gain. Have you instead considered https://www.wikibase.cloud/ which would host it for you? A proper mobile website has been on "wanted features" list for a while and is unlikely to be build with just a help wanted sign. Not sure what you mean by "digital agent". These kinds of AI powered features are still in the "publishing papers" and "prototypes" phase of development (I know the WMF is working on "text" to SPARQL models but they still aren't turn key). BrokenSegue (talk) 00:09, 9 March 2024 (UTC)
Thanks very much for reality checking me.
.
I might actually want to run two instances of Wikibase in my Oneplus 11. One of the instances would be for private items. Examples of private items would be my email messages, news and scientific papers that I read, music and videos. I want my digital agent to learn about me, and my perspective of the world, and a graph structure is excellent for learning. I might ask my digital agent to make copies in local cache of music that I listen to from YouTube repeatedly. When I ask my agent, call it Jaybay, to play music that I like, if there is a local copy from which advertising has been removed, it should play that local copy instead of wasting network resources to get it again.
.
A second local instance for Wikibase could be for public items. I might argue that carbon dioxide is a greenhouse gas that can alter the climate of the planet if the atmospheric concentration changes, and this may result in an increase in my homeowners insurance. That argument would be an item. It would be linked to other items and statements with supporting and refuting evidence and reason. An item in my public Wikibase would be merged with items in other Wikibases that refer to the same entity. The strength of the evidence and reason supporting a statement would be a property of that statement.
.
It might take an immense amount of processing within a Wikibase Ecosystem to calculate the weights of the confidence in the evidence and the reason for all of the arguments, debates, input to a deliberative process. A Oneplus 11 can do trillions of operations per second.
.
The digital agent should respond to spoken as well as touchscreen input. Among other tasks, it would add items to the private Wikibase by default, and only add items to the public Wikibase when explicitly instructed to do so, for the purpose of contributing to a decentralized global platform for collective human and digital intelligence.
Wikibase is not digital agent. While it would be valuable to have software that has a local knowledge base and an LLM that interacts with that, that's not what Wikibase can do currently. ChristianKl ❪✉❫ 16:08, 10 March 2024 (UTC)
Understood, thanks. I spent several hours the last couple of days using termux on my Oneplus 11 installing php, apache2, mariadb, and mediawiki so I could then install Wikibase. I got stuck trying to enable php on apache2.
.
As you say, even if I succeeded in running Wikibase on my phone, there is no digital agent at this time with a language model that can translate English instructions into SPARQL to serve as an interface to Wikibase for nontechnical users. My reach has exceeded my grasp, again. Jkintree2 (talk) 16:54, 11 March 2024 (UTC)
I will conclude by observing that the OnePlus 11 smartphone with 16 GB of RAM had about as good of hardware as any smartphone on the market in 2023.
.
How many such Oneplus 11, or comparable smartphones, were sold worldwide in 2023, maybe millions of them. There will be smartphones with superior hardware available later this year, and smartphones with superior hardware to the 2024 devices that will coming on the market in 2025.
"note that some people will likely object to this since transcripts are copyrighted and so this might be facilitating piracy."
@BrokenSegue: Just curious why you believe some people would view making links to online-transcripts as an invitation to infringe copyright laws any more than any other WD-links to online information? Ottawahitech (talk) 20:49, 10 March 2024 (UTC)
because I've seen people here that have objected to sources associated with piracy. In the extreme case you could imagine people objecting to linking to torrents associated with films. BrokenSegue (talk) 20:54, 10 March 2024 (UTC)
I guess in that case we are stuck using AI that can translate verbal input in language X and producing text output in language Y? Ottawahitech (talk) 21:13, 10 March 2024 (UTC)
I appreciate the warning. Actually it appears there is little appetite here for new property proposals, and rightfully so, with the ever growing number of properties and the complexity of WD in general.
You would likely need to call the property something like "official transcript URL". That way it would be clear it's not going to be used with illegal pirated transcripts. ChristianKl ❪✉❫ 00:32, 11 March 2024 (UTC)
do official transcript websites of tv shows even exist? i thought they were all vaguely pirated (though generally nobody cares) BrokenSegue (talk) 15:39, 11 March 2024 (UTC)
Wikidata weekly summary #619
Here's your quick overview of what has been happening around Wikidata over the last week. This is the Wikidata summary of the week before 2024-03-11. Translations are available.
Discussions
Open request for adminship: Zafer - (RfP scheduled to end after 12 March 2024 20:03 (UTC))
Closed request for adminship: Mike Peel 2 - Adminship re-granted, welcome back Mike!
New requests for permissions/Bot:
Mfchris84-Bot 1 - Task: Add swisscovery edition ID (P9907) to instance of (P31) version, edition or translation (Q3331189) based on checking ISBN-13 (P212) or ISBN-10 (P957).
NinoBot - Task: Semi-automatically create version, edition or translation (Q3331189) entries for Spanish Wikisource Index pages from Commons files.
Emijrpbot 10 - Task: Create a written work (Q47461344) and version, edition or translation (Q3331189) item for works and editions in the National Library of Spain datos.bne.es project.
Queer Data Days - Wikidata Gender Diversity Project Event, a pair of virtual events welcoming a range of speakers investigating queer approaches to data and technology. Sign up for March 15 or March 16, 15:00 (CET).
Wikidata Leveling Up Days, an online event and a collection of resources to welcome new people to Wikidata. The online event takes place on April 5-7 and 12-14 with presentations, workshops and discussions. The program is open to contributions. Learn more about the project at the website WiGeDi.
Wikibase Intro (Dutch) - Bart Magnus writes an article about the benefits of Wikibase for the Flemish Association for Library, Archives & Documentation (VVBAD).
Multilingual entity linking and global insights - Ontotext are developing AI products that can read text and establish of the same concept in different languages (MEL) and between knowledge graphs and plaintext.
Bosnia and Herzegovina - This project aims to be a central hub for the curation of all items (biographical, cultural, geographical, organizational, etc.) related to Bosnia and Herzegovina.
Te Papa Research Expeditions - aims to create or enrich Wikidata with data on research expeditions undertaken by Te Papa/ Dominion Museum/Colonial Museum and institutional staff
WikiProject Highlights: Looted heritage - this WikiProject is calling for Item suggestions to work on. The scope includes all forms of looted heritage; artworks, sacred items, human remains, and other forms of tangible and intangible artifacts that may be considered a part of a people's heritage.
Newest database reports: unmarked supercentenarians - contains instance of (P31) human (Q5) with period of life greater than 110 years and not marked with subject has role (P2868) supercentenarian (Q1200828) or subject has role (P2868) alleged supercentenarian (Q106991708).
I merged them. The date in Q123221110 was from a Wikipedia article that was created with one date in the text and another in the infobox (possibly from copying the layout from the article at Dauke Vreys (Q123177554)). Peter James (talk) 15:30, 12 March 2024 (UTC)
Wikidata Elastic Search configuration files
Hello everyone,
I wanted to ask if there is a way to find out index configuration/mappings files of the main ElasticSearch in Wikidata. Either, the one when browsing to the main wikidata website (the panel on the upper right corner) or the mediawiki actions api wbsearchentities.
I wonder mainly how the labels/aliases/descriptions are stored in the index and what is the returned entities priority/ordering process when searching, e.g. how come when searching a word "human", it returns human (Q5) as the top result.
I would appreaciate any response.
I don't want to hasten anyone or to be impolite in any way, but it would be really great if someone with property creator rights could create this property. This is something that is important to us as the Polish-language Wikimedia community. This issue has been a subject of discussions involving a number of users both in Polish Wikipedia's project chat and on the plWiki Facebook and Discord forums. I fully understand that there are always too few volunteers in our movement to cope with the workload, but if someone could spare a few minutes, that we be really excellent. Thank you in advance! Powerek38 (talk) 12:59, 12 March 2024 (UTC)
I've once clicked on a special link showing all edits with a certain comment. Would it be possible to show all edits belonging to #temporary_batch_1709666797939?
This year, the term of 4 (four) Community- and Affiliate-selected Trustees on the Wikimedia Foundation Board of Trustees will come to an end [1]. The Board invites the whole movement to participate in this year’s selection process and vote to fill those seats.
The Elections Committee will oversee this process with support from Foundation staff [2]. The Board Governance Committee created a Board Selection Working Group from Trustees who cannot be candidates in the 2024 community- and affiliate-selected trustee selection process composed of Dariusz Jemielniak, Nataliia Tymkiv, Esra'a Al Shafei, Kathy Collins, and Shani Evenstein Sigalov [3]. The group is tasked with providing Board oversight for the 2024 trustee selection process, and for keeping the Board informed. More details on the roles of the Elections Committee, Board, and staff are here [4].
Here are the key planned dates:
May 2024: Call for candidates and call for questions
June 2024: Affiliates vote to shortlist 12 candidates (no shortlisting if 15 or less candidates apply) [5]
June-August 2024: Campaign period
End of August / beginning of September 2024: Two-week community voting period
October–November 2024: Background check of selected candidates
Board's Meeting in December 2024: New trustees seated
Learn more about the 2024 selection process - including the detailed timeline, the candidacy process, the campaign rules, and the voter eligibility criteria - on this Meta-wiki page, and make your plan.
Election Volunteers
Another way to be involved with the 2024 selection process is to be an Election Volunteer. Election Volunteers are a bridge between the Elections Committee and their respective community. They help ensure their community is represented and mobilize them to vote. Learn more about the program and how to join on this Meta-wiki page.
Best regards,
Dariusz Jemielniak (Governance Committee Chair, Board Selection Working Group)
[5] Even though the ideal number is 12 candidates for 4 open seats, the shortlisting process will be triggered if there are more than 15 candidates because the 1-3 candidates that are removed might feel ostracized and it would be a lot of work for affiliates to carry out the shortlisting process to only eliminate 1-3 candidates from the candidate list.
The band The Donnas (Q1056412) used to have the twitter handle the_donnas, as evidenced by their website (archive). The band faded from existence in the early 2010s, and the handle must have been deleted. Since 2017 the handle points to another account (ID 823574352729964544) that is unrelated to the band. But since the handle is still listed in places like Wikipedia and Musicbrainz, it is still used in The Donnas (Q1056412) - with all the data belonging to the new account.
So in essence, the handle "the_donnas" was correct until some unknown time before 23 January 2017, but apart from that one reference to their archived website, everything else in X username (P2002) at The Donnas (Q1056412) is wrong.
Is there a way to find the ID the handle was associated with before?
Thanks, that helps! Strangely, if you compare the archives for both handles, it seems like they were active in parallel for a while - or content from the new handle was somehow mirrored to the old one by Twitter ...? It's confusing. El Grafo (talk) 11:05, 13 March 2024 (UTC)
Thanks, meanwhile I've set the old one to deprecated instead – would it be better to do it the other way around? I've also added latest start date (P8555) to the new handle based on what I could pull from the archives, but for some reason that's not an allowed qualifier for X username (P2002) ...? El Grafo (talk) 11:09, 13 March 2024 (UTC)
Weird item about a name of a renamed environmental monitoring program
I’m puzzled by Copernicus Programme (Q124788787) recently created by @WiK14ot0g: which is an item linked to a disambig page, about a former name of a European program. I think we should not do this. (by the way, impressive number of natively spoken languages WiK14ot0g !). Later the item was changed from disambig page instance of (P31) to a legit program value. But it seems the program was just renamed and is essentially the same, so the page on frwiki is really just a redirect / disambig page from the former to the new name.
I found this item only because the linked redirect on frwiki was set as a good article, which is nonsense. What is going on with this item ? author TomT0m / talk page10:55, 12 March 2024 (UTC)
Hey. Thanks for the chat. I just merged it, I did not seen that was the previous name. I was driven by the mistake of somebody else in selecting the disambig page through "follows".
Please check whether Manufacturing in the Russian Wikipedia should really be Обрабатывающая промышленность, not Производство. In the definition that is used in English for the article Manufacturing and the one that is used in Russian Wikipedia for the article Производство, they reflect the same thing.
Manufacturing is the creation or production of goods with the help of equipment, labor, machines, tools, and chemical or biological processing or formulation.
Производство — процесс создания какого-либо продукта с использованием первичных (труд и капитал) и промежуточных факторов производства (сырье, материалы и тому подобное).
Обрабатывающая промышленность — отрасль промышленности, в качестве сырья в которой используются продукты сельского хозяйства (хлопок, шерсть и т. д.) или добытые добывающей промышленностью (нефть, природный газ, руда и т. д.), обрабатываемые вручную или с помощью машин.
According to these definitions, Manufacturing = Производство, not Обрабатывающая промышленность.
It should be noted that the article Manufacturing is included in the 1000 most important for all Wikipedias, and it is unlikely that in Russian it includes precisely the Обрабатывающая промышленность or Промышленность, but rather the broad concept of Производство. In this regard, I consider it necessary to correlate Manufacturing in the English Wikipedia with Производство in the Russian one. Производство in Russian language is the most basic concept, промышленность and обрабатывающая промышленность are less basic. Reprarina (talk) 09:28, 14 March 2024 (UTC)
Dear Wikidata users, join us today and tomorrow for the Queer Data Days event, organised by the Wikidata Gender Diversity project! The event will run from 2pm GMT to 7:30pm GMT and feature presentations from speakers working broadly on the topic of queer data (including our own project). The event is hosted by University College London on the Zoom platform. Please find all the instructions on the event page. Thank you! Mushroom (talk) 09:41, 15 March 2024 (UTC)
How to combine Surinam country data Q730, Q19901436 and Q7646305 to find all Surinam data?
I've entered and cleaned some Surinam plantations and all were recorded with country code Q7646305 (Surinam as dutch colony).
However people looking for this data don't realise that the code Q730 doesn't include the colony data. To complicate this, the colony Suriname lasted to 1954, then the country became a country within the Netherlands (Q19901436) till 1975.
Basically we have one geographical country registered in 3 codes, colony, dutch country and republic.
My question is how to solve this. One solution could be to add countrycode Q730 to all plantations, but I wonder or making the colony a sub of Q730 could solve this structurally. Nico5038 (talk) 18:09, 15 March 2024 (UTC)
I don't think there is a good way. country (P17) is muddled and my proposals to clarify for historical items were met with very different viewpoints. I suggest you use have both country (P17) and country of origin (P495) for when it was formed, ignore any warnings and don't be surprised if someone fiddles with them. Nesting historical countries within existing ones is unwise as the borders often change. Vicarage (talk) 08:13, 16 March 2024 (UTC)
Thanks for the response. I do see the merit of using the (P495) country of origin as it does ty the plantation to the colony aera. However it would also require all plantations to get a P17 country, so two different properties, while the country is the same name, only a different type of governement. Thus my question how we could ty the threetypes of governement to the same countryname "Surinam" and include the three types in a query using just the Q730. Nico5038 (talk) 22:23, 16 March 2024 (UTC)
Overlong names
Sometimes scientific articles have extremely long names - e.g., Q57278717, whose proper title is "Excited states and solvatochromic shifts within a nonequilibrium solvation approach: A new formulation of the integral equation formalism method at the self-consistent field, configuration interaction, and multiconfiguration self-consistent field level", but which we describe as "Excited states and solvatochromic shifts within a nonequilibrium solvation approach: A new formulation of the integral equation formalism method at the self-consistent field, configuration interaction, and multiconfiguration self-consistent field lev". I've seen worse; this is just the one I was looking at when it occurred to me that I could make a post about it.
I don't like the idea that the best way of dealing with this is just to let long fields be truncated. Suggestions? DS (talk) 20:46, 16 March 2024 (UTC)
If you look at the recent edits, it seems like Wikidata is able to have the full name. Maybe, this is a QuickStatements issue? ChristianKl ❪✉❫ 21:56, 16 March 2024 (UTC)
I'm not sure about scientific articles, but I'm dealing with bibliographical info for books, and the [ : ] character usually denotes a title : subtitle pair. In this case, the title of the article would be "Excited states and solvatochromic shifts within a nonequilibrium solvation approach" and the subtitle "A new formulation of the integral equation formalism method at the self-consistent field, configuration interaction, and multiconfiguration self-consistent field level". IMHO much more readable. --Ignacio Rodríguez (talk) 23:02, 16 March 2024 (UTC)
Data type max length
I can see on Wikibase cloud, that there is probably max length for string and single langauge text. What is the max lenght for these datatypes? Juandev (talk) 23:02, 16 March 2024 (UTC)
This request is forwarded from a Discord groupchat:
I want to put a constraint on Threads username (P11892) such that its value is identical to Instagram username (P2003). (Justification and tutorial on how a threads account is bound to the instagram account are omitted)
I dived into the list of available constraints and concluded that such constraint does not exist yet. I have managed to come up with the following SPARQL query:
This query finds items that have both properties but don't have the same value. However, it disregards qualifiers.
How can I write a refined query so that I can use in a custom constraint, or is it possible to implement such constraint in the software? MilkyDefer07:11, 16 March 2024 (UTC)
but why couldn't someone have a threads account with one username (and an associated instagram account they never use) and a differently named instagram account BrokenSegue (talk) 12:53, 16 March 2024 (UTC)
I cannot contact the OP now, this is my own opinion: Having an associated instagram account they never use does not change the fact that they have the instagram account. Adding it to the list of instagram accounts then immediately tagging it to deprecated rank makes the item more complete. MilkyDefer14:40, 16 March 2024 (UTC)
You probably want to learn what the deprecated rank means in wikidata, Milky. Deprecation is for statements that are not and were never true. If the person or organisation does have an IG account so named, but does not use it, then deprecation is incorrect. The used account should be preferred rank. The unused but nevertheless true account should be normal rank. See also Help:Ranking#Deprecated_rank "used for statements that are known to include errors (i.e. data produced by flawed measurement processes, inaccurate statements) or that represent outdated knowledge (i.e. information that was never correct, but was at some point thought to be).". --Tagishsimon (talk) 00:33, 17 March 2024 (UTC)
Then set the one in active use to preferred rank and other placeholder accounts normal rank, that sounds sensible. MilkyDefer13:51, 17 March 2024 (UTC)
Hi, At Big Brother (Q549337) it's missing 2 presenters - AJ Odudu and Will Best - both started hosting it on 8 October 2023,
Could someone be so kind as to add these for me please as I'll be honest I'm far too incompetent for WikiData!, Many thanks, Warm Regards, Davey2010 (talk) 17:20, 17 March 2024 (UTC)
Property:P8645 has a URL formatter that points to a European database that seldom seems to contain French electrical plants (perhaps only the largest?). So I tried to install a formatter that deals with numbers starting with 17 (France, RTE), to no avail. Can somebody look into this? David.Monniaux (talk) 06:39, 18 March 2024 (UTC)
Wikidata weekly Summary #620
Here's your quick overview of what has been happening around Wikidata over the last week. This is the Wikidata summary of the week before 2024-03-18.
Discussions
Open request for adminship: KonstantinaG07 3 (RfP scheduled to end after 20 March 2024 17:37 UTC)
Closed request for adminship: Zafer Adminship granted, welcome Zafer!
New requests for permissions/Bot:
PagesBot - Task: Adds no. of pages statements based on page(s) on items of type: scholarly article. Then adds a reference with based on heuristic equal to inferred from page(s).
Wicci'o'Bot - Task: Import identifiers of new properties created for TheTVDB (Q15616250), nientepopcorn (Q123434360) (and more properties I plan to request to be added shortly) inherent in the film and TV industry, using IMDb ID (P345) as source.
Closed request for comments: Duplicate Refs, Data Model and UI - (RfC closed 15 March 2024 18:57 UTC). The Community showed support for the proposal to change data model and UI for duplicated references. A Phabricator ticket has been opened: T360224
Next Linked Data for Libraries LD4 Wikidata Affinity Group call March 19, 2024: Please join us on March 19 for a community discussion about how GLAM professionals can incorporate Wikidata/linked data into their work week. Bring your success stories about institutional buy-in and how you’ve communicated the value of linked data to managers and colleagues. We hope to share some strategies that you can take back and use in your own workplace. This will be a followup of sorts to our call last May, and you can review some of what we talked about then, plus slides from our survey here: Notes from May 30, 2023 Meeting on Advocacy for Wikidata & Linked Data in Libraries.Agenda
Connecting Local Archive Data to Wikidata: Focusing on the Archives of National Debt Redemption Movement. By Yun, J., & Oh, S.G. Documents the findings of connecting Wikidata with the Korean NDRM.
Wikidata Lab: Dear Diary (Portugese) - Aims to share resources and capabilities for integration of Wikidata with other free knowledge projects, in particular the Querido Diário project, from Open Knowledge Brasil. Hosted by Porto, É. & Giulio Carvalho, G.
ORCID Scraper - helps users get DOIs from orcid.org for a specific author and curate them using Scholia. It gives users an overview of which DOIs are missing and helps them easily import missing articles one by one.
Newest WikiProjects: Polish Socialist Party - aims to document and enrich data and items (Leadership, sports clubs, commemorated buildings, art, publications etc.) that are connected with the Polish Socialist party.
WikiProject Highlights:
RfC: Slowking4 (2) - Comments are required on actions against the named editor.
Showcase Items: Mahatma Gandhi - The "Great Soul", lawyer, political ethicist and leader of the non-violent resistance to end Imperial British occupation of India.
Check of consensus for use of amateur drawings on items about living people
There is currently a discussion on Commons Village Pump about 5 amateur drawings that are currently in use at Wikidata (and then transcluded through infoboxes to multiple Wikipedias). The concern is whether they should be retained or deleted at Commons. However, a key part of any Commons deletion discussion is whether the files are in use on another project. They are used here on the following items;
Is there consensus either to retain or remove any of these images from Wikidata? This will have a big impact on which way any deletion discussion goes on Commons.
@Hibrideacus, MHM55, Quick1984, Jmabel: as uploader and other editors involved in the discussion at Commons. From Hill To Shore (talk) 12:13, 16 March 2024 (UTC)
This has come up before. See for example Topic:Xpf6vbtk7ccolrng. I do not know how we are to determine which drawings are "good enough". We have no explicit policy but I don't see how we can justify removing images that resemble the subject. BrokenSegue (talk) 12:55, 16 March 2024 (UTC)
If you don't judge the quality of drawings, how are you going to judge their resemblance? Even many photographs provoke complaints from the people depicted, and we make the excuse that they were taken in a public place. Here, feeble semblances of portraits are created without any participation of the subjects. Quick1984 (talk) 13:11, 16 March 2024 (UTC)
is this image of sufficient quality? Genuine question. How do you account for stylistic interpretation? I agree we need a policy. I don't know how to write one. BrokenSegue (talk) 15:32, 16 March 2024 (UTC)
To the extend that their's stylistic interpretation, we would want it to be neutral (or maybe positive) when showing images for Living People on Wikidata. Having images that make negative stylistic choices, seems bad.
I haven't dealt much with images, so I don't feel like I'm the ideal person to write a policy, but it would be great if someone who actually has more experience with dealing with images on Wikidata to write a policy draft to how to select images. ChristianKl ❪✉❫ 23:13, 16 March 2024 (UTC)
it's funny that some people are like "this has so little resemblance that it's insulting" and some are saying "this is so similar it's a derivative work". BrokenSegue (talk) 13:31, 17 March 2024 (UTC)
BrokenSegue, I don't see a contradiction here. The reference to a derivative work here is applicable in a technical and legal sense - this image, and not the original, was before the eyes of the author when they tried to create the portrait. But the result, due to the lack of proper skills, was unsatisfactory, so there is practically no resemblance. By the way, violation of c:COM:DW policy (if detected) allows to delete images even if they are in use. --Quick1984 (talk) 13:49, 17 March 2024 (UTC)
Actually, despite its flaws, the Leikeki image is close enough to the image on OneEarth.org that I might consider it a derivative work and thus not acceptable on Commons. I'll file a DR. DS (talk) 15:41, 17 March 2024 (UTC)
The reason it's not currently allowed as a qualifier is that the statement requires an "applies to jurisdiction" qualifier, and you can't have qualifiers on qualifiers. If the restrictions are loosened if will no longer be possible to alert people that they should add jurisdiction information. IMO the usefulness of this check speaks for keeping it as it currently is. Infrastruktur (talk) 16:45, 19 March 2024 (UTC)
On Plato (Q859)RISM ID (P5504) there is a depreciated and normal. On the AC bar in the articles, it shows as an error, but it shouldn't be. I purged both the item and the article and still got the same result. I tried removing, replacing, swapping, etc., and the error persisted. Can someone please tell me what is going on with it and how to fix it? Thanks! Thanks - Da LambTalk to me!Please don't eat da 🐑!01:34, 19 March 2024 (UTC)
Just guessing that AC = authority control and "article" means the article on Simple Wikipedia. If I have guessed correctly what you mean, then please note it is the valid value people/30108312 which is being marked as invalid. So perhaps something is wrong with the template on simple wiki. — Martin (MSGJ · talk) 11:05, 20 March 2024 (UTC)
City article Bamberg (Q3936) contains world heritage site references (heritage designation, world heritage criteria, world heritage site id) when in fact the world heritage designation only applies to Bamberg's old town, not the entire city. The old town has a dedicated entity Town of Bamberg (Q114559132), named like official Unesco web page. Property P757 has a uniqueness constraint, so duplication is not wanted. I think the more specific article should get the reference and I've added missing data there. However, I'm not allowed to edit Bamberg to remove world heritage information because it is protected. Can someone help?--178.201.237.22721:10, 23 March 2024 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 12:47, 26 March 2024 (UTC)
Wikidata weekly Summary #621
Here's your quick overview of what has been happening around Wikidata over the last week. This is the Wikidata summary of the week before 2024-03-25.
2024 Women's Wiki Month Datathon (Indonesian) Between 24 - 26 March 2024, Wikimedia Indonesia is hosting a competitive Data-thon with Prizes available for editing and enriching Wikidata items on the topic of Women's Health. Full details on the project page.
Levelling Up Days 2024 is an online event designed to welcome new people to Wikidata, upgrade their editing skills and find out about useful tools and apps. Join us for a series of videos and discussions regarding Wikidata and its uses in and outside of the Wikimedia projects.
Program - the live portions of the event take place across the weekends 5-7 and 12-14 April
Ontology Explorer, also an Observable notebook, explore the subclasses in front or behind a Wikidata Item.
Tool of the week
Members of European Parliament by pac02. Explore visualisations of MEP's in the European Parliament. Country of origin, gender, political alignment and other properties from Wikidata can be explored.
Wikimedia Foundation Research Award of the Year 2023 - We encourage you to nominate and vote for research papers published in 2023 that focused on or benefited the Wikimedia projects (we're not biased but hope you vote for Wikidata-based Papers). Please submit your nominations by April 18, 2024.
In preparation for the Research Track of the 11th Wiki Workshop, there is now a Call for Contributions for the event scheduled virtually on June 20, 2024 (tentatively 12:00-19:00 UTC). The Wiki Workshop aims to bring together researchers who study all aspects of the Wikimedia Projects. We hope Wikidata and Wikibase are well represented! Submission deadline: April 22, 2024 (23:59 AoE)
Showcase Lexemes: अग्नि -the Sanskrit word for 'fire' and also the Hindu God of fire 'Agni'
Development
Wikibase REST API:
We worked on the ability to create Items via POST /entities/items (phab:T342990)
We worked on making it possible to modify the data of a Property via PATCH /entities/properties/{property_id} (phab:T347394)
EntitySchemas: We are getting back to making architecture improvements to enable the new datatype to link to EntitySchemas in statements.
IP masking: We wrapped up the work on adapting Wikibase for the upcoming IP masking changes.
Query Service: We added the ability to stop a query in the UI in order to not have to wait for a broken query to finish or run into a time-out. You can find the stop button below the run button in the query UI now.
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 12:45, 26 March 2024 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 12:49, 26 March 2024 (UTC)
Updates on designing a new Community Wishlist Survey
Hello everyone, there is new information concerning the redesign of the Community Wishlist Survey.
Firstly, in case you missed them earlier, the updates we have provided so far are:
Update 4: Since we are planning on keeping the Wishlist open all year and also looking at how the community have participated in vetting/refining wishes, should wishes be editable?
Please have a look at any of the updates that interest you, particularly Update 4 and 5.
To keep the conversation in one place, please leave your feedback on the central talk page for all the updates (preferably). However, you can leave comments under this post too.
For position (Q4164871) when I look for it to use in instance_of, it no longer appears as an option as I type in the name, what changed to have it so far down the list of options? RAN (talk) 16:28, 21 March 2024 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 16:52, 28 March 2024 (UTC)
Why is only 1 wikipedia reference from an item possible
I'm trying to register Surinam plantations and some are now a village. The village has been described as a former plantation, howeven I can't link the wikipedia article as the village item already links to the article. Why only a 1:1 relationship? Nico5038 (talk) 07:37, 24 March 2024 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 16:52, 28 March 2024 (UTC)
I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. RVA2869 (talk) 16:53, 28 March 2024 (UTC)
Referencing information confirmed via email
I am working on updating the Wikidata items for Indigenous performing artists, part of which is asking them to confirm that information about their Indigenous identity listed in publicly-available bios is accurate and is something they want to have attached to their Wikidata item. Sometimes, when contacted, artists will share details that are relevant and useful inclusions, but are not present in the biographical pages I've found.
In citing this information with Reference, I've been using Type of Reference - Correspondence, along with a quotation from the email. I also make an entry on the talk page for the item I'm creating, noting that I am using information received directly from the artist. I would like to direct people to the talk-page comment directly from the reference using Reference URL, but doing so gets flagged because the field doesn't accept Wikidata links. For an example of what I've been doing, you can look at the place of birth or affiliation on Julian Taylor's page.
Is there a different property I should be using to link contextualizing information on the talk page? Should I be formatting references to information from private correspondence in a different way entirely? Any guidance would be great. MichifDorian (talk) 13:31, 18 March 2024 (UTC)
@MichifDorian, When using external references, anyone (or at least some other people) can access such a reference and verify that a Wikidata item or Wikipedia article provides accurate information. When you want to refer to information that is obtained via email and is available nowhere else, no one else is able to verify its accuracy. That is why, referring to the personal correspondence is unacceptable. Michgrig (talk) 21:14, 18 March 2024 (UTC)
Wikimedia Commons has a system in place called VRT, where emails from creators are stored that give permissions for Commons housing their images. Perhaps we need something similar for storing emails containing biographical data. --RAN (talk) 18:03, 23 March 2024 (UTC)
Cannot find the way to add a label in other languages
I'm allowed to edit labels in English and French, but cannot find the way to enter a label in a new language, say, Hebrew. I cannot find where in the interface I'm allowed to do this. David.Monniaux (talk) 16:30, 24 March 2024 (UTC)
I believe it lets you add languages that you indicate that you know on the babel box on your user page. You have added the boxes for fr/en so that's what it shows you. BrokenSegue (talk) 17:03, 24 March 2024 (UTC)
(after edit conflict) @David.Monniaux: There are a few ways to do this. First, if you add Babel boxes to your User page, the interface will offer you those label languages to edit. Another way is to activate the labelLister gadget in your preferences, which adds a new "labels list" menu item. I think there are some other methods as well, but I don't use them. From Hill To Shore (talk) 17:07, 24 March 2024 (UTC)
I didn't know about the Babel boxes and haven't checked if that works, although without them the edit page seems to display the three most recent languages chosen in settings. Right-clicking on the "edit" button to the right of the labels is one method, if occasionally adding labels, but the gadget can be more useful particularly if adding labels in several languages to an item. Peter James (talk) 23:26, 24 March 2024 (UTC)
Wikidata weekly Summary #621
Here's your quick overview of what has been happening around Wikidata over the last week. This is the Wikidata summary of the week before 2024-03-25.
2024 Women's Wiki Month Datathon (Indonesian) Between 24 - 26 March 2024, Wikimedia Indonesia is hosting a competitive Data-thon with Prizes available for editing and enriching Wikidata items on the topic of Women's Health. Full details on the project page.
Levelling Up Days 2024 is an online event designed to welcome new people to Wikidata, upgrade their editing skills and find out about useful tools and apps. Join us for a series of videos and discussions regarding Wikidata and its uses in and outside of the Wikimedia projects.
Program - the live portions of the event take place across the weekends 5-7 and 12-14 April
Ontology Explorer, also an Observable notebook, explore the subclasses in front or behind a Wikidata Item.
Tool of the week
Members of European Parliament by pac02. Explore visualisations of MEP's in the European Parliament. Country of origin, gender, political alignment and other properties from Wikidata can be explored.
Wikimedia Foundation Research Award of the Year 2023 - We encourage you to nominate and vote for research papers published in 2023 that focused on or benefited the Wikimedia projects (we're not biased but hope you vote for Wikidata-based Papers). Please submit your nominations by April 18, 2024.
In preparation for the Research Track of the 11th Wiki Workshop, there is now a Call for Contributions for the event scheduled virtually on June 20, 2024 (tentatively 12:00-19:00 UTC). The Wiki Workshop aims to bring together researchers who study all aspects of the Wikimedia Projects. We hope Wikidata and Wikibase are well represented! Submission deadline: April 22, 2024 (23:59 AoE)
Showcase Lexemes: अग्नि -the Sanskrit word for 'fire' and also the Hindu God of fire 'Agni'
Development
Wikibase REST API:
We worked on the ability to create Items via POST /entities/items (phab:T342990)
We worked on making it possible to modify the data of a Property via PATCH /entities/properties/{property_id} (phab:T347394)
EntitySchemas: We are getting back to making architecture improvements to enable the new datatype to link to EntitySchemas in statements.
IP masking: We wrapped up the work on adapting Wikibase for the upcoming IP masking changes.
Query Service: We added the ability to stop a query in the UI in order to not have to wait for a broken query to finish or run into a time-out. You can find the stop button below the run button in the query UI now.
Are we allowed to post the real life name or town of residency of this person or does that violate Wikidata:Living people? For source, look at the references i just added to the item Trade (talk) 03:27, 23 March 2024 (UTC)
@Trade: if the person is notable and there exists serious sources with such data (real life name, town of residency), then I don't see problem here. But if such data are unreferenced, you can delete them Estopedist1 (talk) 08:00, 23 March 2024 (UTC)
As i already stated, the data is referenced in the source added to the item. I am asking whether or not it's sufficient. Trade (talk) 15:53, 23 March 2024 (UTC)
the references are probably high enough quality to include the name / gender / country though I will say that really this discussion should happen in project chat (it's not really an admin question). BrokenSegue (talk) 05:59, 25 March 2024 (UTC)
If you want to create a new item for a hamlet (Q5084), it's easiest to model it after an existing item: list of hamlets (press play button). Not sure if I understood you correctly, if you prefer to ask in german, there is quite a few germans on Wikidata. Infrastruktur (talk) 13:18, 26 March 2024 (UTC)
Please protect Tornado Cash from scammers
After sanctions were imposed by the US Treasury Department on the Tornado Cash cryptocurrency mixer, the tornado.cash domain was blocked. Despite the sanctions, people are trying to find access to this mixer and scammers are using it. Scammers embed their sites on wikipedia pages. They use the WIKIDATA for this. Therefore, I ask the administrator to take action and prohibit editing the Tornado Cash page on the WIKIDATA. If you do not do this, people will give their cryptocurrency to scammers. Vladlen Terezhe (talk) 03:31, 25 March 2024 (UTC)
but wikipedia has an article and includes a link? i'm not sure we have a policy around not linking to websites sanctioned by the treasury. BrokenSegue (talk) 03:58, 25 March 2024 (UTC)
Linking to the former official tornado.cash is not a problem, that one's dead anyway. I think what @Vladlen Terezhe is saying is that 1) there are other people who want to profit from the site being down by funneling users to their own copycat website and 2) they use Wikipedia/Wikidata to make their site(s) appear "legitimate". El Grafo (talk) 08:42, 25 March 2024 (UTC)
I've hidden the dubious links. The registration for the original site runs out 2024-07-17, so other people might be able to register it after the usual grace period. I changed the claim for the website so that it says for which period the website was historically valid, this is sufficient IMO. Infrastruktur (talk) 10:33, 25 March 2024 (UTC)
The "torrent.cash" site appears legit up until it was taken down two days after the sanction was announced. Same goes for the original github repos. Bloomberg has some paywalled articles raising concern about the remnants of the site possibly being under the control of bad actors. Although generally considered a solid source I don't know which sources they used so I will take this with a grain of salt as I tend to do with most crypto currency-related claims. The nail in the coffin as far as I'm concerned is this. One of the developers behind Torrent Cash, Alexey Pertsev, was arrested six days after the sanction in the Netherlands. I find it unlikely that he's editing from jail, so actions will be taken on the account PertsevAlex. Infrastruktur (talk) 09:45, 25 March 2024 (UTC)
It seems likely there will be conditions attached to his his bail and he's still quite literally under surveilance until the the trial comes up. More involvement in the sanctioned crypto anonymizer is taking a huge legal risk.
The accounts blocked were _only_ adding external links to the Tornado Cash item, nothing else. Using several accounts associated with a real identity. And why use two different accounts?
The source code for the site was open so it's easy for anyone to claim claim to be one of the original site's developers, set up a new site, and run away with the money without anyone being able to know who they are.
If one of the accounts can prove their identity I would be willing to unblock. But I would like to hear if they are the ones who registered the other account or not. I would also like to know why he's reopened a service where their customers are likely to have their crypto-assets siezed as a result of U.S. sanctions. The latter seems like a good reason to refuse the posting of any new links regardless of their source.
True, but I reacted to your statement that editing from jail was unlikely. He is not anymore in jail and supervision/surveillance in NL mainly is electronic tagging (Q17011956). However the points you make above of course are indeed reasons to act as you did. In your block reason you point to this discussion and I wanted to make clear that A.P. is not in fact in prison anymore. --Lymantria (talk) 13:58, 27 March 2024 (UTC)
Please undelete Eachy (Q5324889) (a long-standing Wikipedia hoax)
O.K., done. The item is almost empty and its edit history is not very significant -- I do not know if it is important in this case. --Wolverène (talk) 09:35, 27 March 2024 (UTC)
Should it be modelled as just a Wikipedia hoax, with most other statements set to deprecated due to being a hoax? Or as a fictional entity? In which case the other statements should probably be set to normal rank like on items for other fictional/mythical creatures. --2A02:810B:580:11D4:D503:C53A:9905:C38312:22, 27 March 2024 (UTC)
My wikidata id ( Q124839975 )
You deleted it for more accounts, but I think it deserves to stay. Can we somehow make it undeleted again?
please response
thank you Sherbazbichu (talk) 23:10, 20 March 2024 (UTC)
Wikidata is not a place that exists to advertise yourself. In general, making requests without giving any arguments for why a deletion decision was wrong is a sign that the deletion decision was right. ChristianKl ❪✉❫ 23:57, 20 March 2024 (UTC)
The references above do not convince me of notability, mostly advertorial like. Apart from that the creating user and item subject seem to be the same person, combined with the deleted twiwiki-links I judge the items as promotional. --Lymantria (talk) 07:56, 21 March 2024 (UTC)
The fact that a website lists him as a doctor working at a hospital is not evidence for him being famous. If he's famous, where are the independent sources that speak about him? ChristianKl ❪✉❫ 02:42, 22 March 2024 (UTC)
I already looked at those. Two mention Bombay Hospital, Mumbai, and are probably Shrirang Bichu; the COVID one seems to have no information beyond the name but United Arab Emirates is not one of the countries mentioned. Peter James (talk) 18:52, 22 March 2024 (UTC)
MV Dali (involved in the Baltimore bridge collision)
This has two Wikidata items. The main one is Dali (Q83641522), but there's also one at Category:Dali (ship, 2015) (Q84184607). There is no WP category associated with this, and this duplication is breaking the (useful) linkage from the en:WP article to the Commons category, as the wikidata 'category' record (I've yet to see any useful value to these records) is grabbing the link instead. Andy Dingley (talk) 12:56, 26 March 2024 (UTC)
It's how Commons categorises ships, probably because it's easier to maintain both by connecting the categories for different names of the same ship and avoiding having to recategorise images for every name change (and keep track of every change). Sometimes ships are renamed just before they are scrapped so with all images in one name category would it always be appropriate to use the most recent name? This one currently has one name and relatively few images but there are some such as Commons:Category:IMO 9106297 with hundreds of images and five name categories, each with at least one subcategory. Peter James (talk) 18:35, 27 March 2024 (UTC)
Does somebody know what did I do wrong here when I added a new complex constraint to reason for preferred rank (P7452)? The report has not updated in nearly three weeks even though it used to update almost daily before I added the new one. The constraint query seems to work well in WDQS but there must be something wrong with it. Many thanks in advance. –Samoasambia✎21:46, 26 March 2024 (UTC)
Here at Wikidata we reflect claims made by all valid sources (and we have quite broad definitions about what we accept as "valid") and can include multiple different claims. Where a statement can only have one true claim (such as date of birth) we record all the valid claims and mark one as preferred. Reusers of our data who want the "best" answer can retrieve just the preferred claim. Reusers of our data who want to study all claims (perhaps to see if a date of birth is disputed) can retrieve all claims. From Hill To Shore (talk) 15:33, 28 March 2024 (UTC)
Just to make that answer slightly more specific, the year-only claim has a reference and that source only gives the year. We can't put that reference on the specific date because the source doesn't support it. If we didn't have a claim for just the year, then we would have nowhere to add that reference.
The data on the website seems to be Attribution 3.0 (CC BY 3.0) license based on the Google Database search. I Have never imported data before, so any guidance would be appreciated. Wallacegromit1 (talk) 09:16, 28 March 2024 (UTC)
Look for Quick Statements, which allows you to import a text file where you've prepared the repeated elements. Vicarage (talk) 04:32, 30 March 2024 (UTC)
@Wallacegromit1: One thing to be very careful about is when you see external authorities give separate IDs to the same subject. It normally means there is some subtle difference that we need to reflect here. When doing your import, create separate items for each set of IDs and we can consider if it is appropriate to merge them afterwards. In this case, you have conflated two organisations in Box Hill Historical Society (Q125123995), which should have been created separately. Due to the merging of two local government areas, the "city" society effectively ceased to exist in 1994 but immediately reformed to cover the larger geographical area without "city" in the name. We should follow the external authorities' lead and model these as successor organisations. From Hill To Shore (talk) 08:14, 30 March 2024 (UTC)
@From Hill To Shore Thanks for creating the new item, and will be mindful of this going forward.
As for the second point, In terms of the registered business and society being seperate items, they can remain the same, as they are the same 'current entity'. However, if you or others feel differently, I would be open to splitting the item further. Wallacegromit1 (talk) 13:50, 30 March 2024 (UTC)
@Vicarage Thanks for the tip. I am a user of both OpenRefine and QuickStatements, however, I am looking for a workflow that will help me get information from a wedsite and add it in a openrefine/quickstatement format, in a more convinient or semi-autonomous manner. The Data:Import help page says to use the 'BOT', but with little instructions on how to use the bot. Any advise for this? Wallacegromit1 (talk) 13:56, 30 March 2024 (UTC)
Cannot add Bangla label of Q695838
internationalism (Q695838) is called "আন্তর্জাতিকতাবাদ" in Bangla. However, I cannot add the Bangla level in internationalism (Q695838), because it shows the following error: "Could not save due to an error. Label and description for language code bn can not have the same value." However, I have given different values for label and description for Bangla, in contrary to the error. Sbb1413 (he) (talk • contribs) 05:17, 29 March 2024 (UTC)
I think the error means that there is another item with an identical label and description in that language. If you can rephrase the description so it is even slightly different, the edit should go through. From Hill To Shore (talk) 07:14, 29 March 2024 (UTC)
If it was because the other item had the same label and description, the error would have been: "Could not save due to an error. Item Q7457921 already has label "আন্তর্জাতিকতাবাদ" associated with language code bn, using the same description text." Q7457921 had a different description; the error was because Q695838 had the description "আন্তর্জাতিকতাবাদ"[12] and when the item was edited the change of label and description were processed as separate changes, with the label change processed before the new description was applied. Changing the description first then the label would have worked. It is possible to make changes as one edit (such as by using "labelLister") but not in the Wikidata user interface. Peter James (talk) 08:44, 30 March 2024 (UTC)
I see @William Graham used has effect (P1542). This property on its own does not distinguish between unintended side-effect and method of operation, however.
All of them look to be different people, or at least I am unable to find any commonality between the articles/papers they have authored. William Graham (talk) 00:34, 27 March 2024 (UTC)
The French page as well as the Spanish one have been recently moved to "Cyanobacteriota". As a result, they are not linked any more to the wiki pages in other languages (and there are many, about 80 of them), they are only linked to each other. In fact the redirect page still links to the 80 other pages, but this is rather impractical. I tried to solve the problem, being reverted each time. The two instances should be merged as they are perfectly synonymous, Cyanobacteriota being only a recent renaming of the phylum Cyanobacteria.