Open main menu

Wikidata:Project chat/Archive/2019/08

This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion.



Could anyone eliminate this. I rushed over and created an existing page for the same link. Conde Edmond Dantès (talk) 00:40, 1 August 2019 (UTC)

Elimination likely isn't the answer. It sounds to me like you created a dublicated item. In that case you have to merge both items that are about the same thing. ChristianKl❫ 11:12, 1 August 2019 (UTC)

Discussing a potential merge

I think the items comptroller (Q673633) and Q12306661 should be merged. As far as I can tell, they are both about the occupation, called "comptroller" or "controller". Some of the links should arguably be in the item auditor (Q10949665) instead. The thing is, I'm not completely certain to what extent these concepts overlap in all of the different languages. Where should I bring this potential merge up for discussion? Gabbe (talk) 04:50, 1 August 2019 (UTC)

No languages

"⧼valueview-expertextender-languageselector-languagetemplate⧽" instead of languages for name (P2561) is showing in the list of languages. Eurohunter (talk) 16:16, 2 August 2019 (UTC)

Hello, this could be related to Wikidata:Project_chat#Unit_box_not_appearing?. --M2k~dewiki (talk) 16:27, 2 August 2019 (UTC)
Yup (not certain if actually related but we’re tracking it in the same task), I’m on it right now --Lucas Werkmeister (WMDE) (talk) 16:34, 2 August 2019 (UTC)
Seems to be fixed now. --Lucas Werkmeister (WMDE) (talk) 17:55, 2 August 2019 (UTC)

Hi! I wonder how the value of ⧼valueview-expertextender-languageselector-languagetemplate⧽ can be queried in a multi-property query. Please see Yiddish list. Thanks for your feedback in advance! Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 07:53, 3 August 2019 (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. Premeditated (talk) 10:29, 7 August 2019 (UTC)

request for merge

Hi! I created Ragnheiður Steindórsdóttir (Q66057526) because of a misspelling on a related webpage but today trying to write an article I could find the already existing Ragnheiður Steindórsdóttir (Q16424080) which is already linked to iswiki. Please merge the items. Thanks in advance!
I assume that Wikidata:Project chat ahouls have a high level selector "'Request merge". Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 10:21, 3 August 2019 (UTC)

  Merged. Please read Help:Merge for instructions on how you can do it yourself. - Premeditated (talk) 10:41, 3 August 2019 (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. Premeditated (talk) 10:29, 7 August 2019 (UTC)

import from WP infobox

Is there a tool that imports infobox field entries from WP for existing WD items? People make updates in WP which don't get reflected here. --SCIdude (talk) 08:43, 5 August 2019 (UTC)

I will recommend using harvesttemplates for an easy interface, and the pywikibot script for more advance usage. DISCLAIMER: Be careful with what you import as the tool it not very smart and just include everything that has a WP-link. - Premeditated (talk) 08:52, 5 August 2019 (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. Premeditated (talk) 10:28, 7 August 2019 (UTC)

edit check

I'm wondering what I did wrong to trigger a revert - see this If I'm messing up - let me know now before I continue too much further. Reverts usually signal something really wrong. TY for your time Ched (talk) 15:31, 5 August 2019 (UTC)

  • The field is called "description" so adding "description" in there isn't really helpful. Try to write a sentence "<heat lightening> <is (a)> <text from description>"
BTW, English labels and descriptions should usually start with lowercase letters. No period at the end. Help:Description might help.
"lightning not accompanied by thunder sound" might do. --- Jura 15:43, 5 August 2019 (UTC)
good point on the word "description" - trying to remember lower case and no period - thanks for your time Ched (talk) 15:54, 5 August 2019 (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. Matěj Suchánek (talk) 09:59, 7 August 2019 (UTC)

Petscan ?

Petscan is down for me since yesterday. Am I alone ? I've try Firefox and Chrome, same thing. I've try to find a bug on Phabricator, but I see nothing. I though to enter an issue, but Magnus is away, so someone/somewhere else to contact ? Simon Villeneuve (talk) 10:53, 6 August 2019 (UTC)

@Simon Villeneuve: PetScan is down for me as well. I've reported it to Phabricator (phab:T229921) since it might be something out of his control. Jc86035 (talk) 11:47, 6 August 2019 (UTC)
This is likely related to a planned maintenance downtime of the server [1]. --Pasleim (talk) 15:53, 6 August 2019 (UTC)
Magnus restarted Petscan this morning. — Envlh (talk) 06:47, 7 August 2019 (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. Matěj Suchánek (talk) 09:59, 7 August 2019 (UTC)

Would it be safe

to set the sex or gender (P21) for those 63700 items to male?

select ?item where {
  ?item wdt:P31 wd:Q5 .
  ?item wdt:P735 ?voornaam .
  ?voornaam wdt:P31 wd:Q12308941 .      
  optional {?item wdt:P21 ?geslacht} .
  filter (!bound(?geslacht))}

Try it!  – The preceding unsigned comment was added by Edoderoo (talk • contribs) at 19:25, 30 July 2019‎ (UTC).

It's possible that there would be some errors, since it's not unheard of for females to be given supposedly male given names, including in middle names, and vice versa (e.g., Willy Kanis (Q388542). There are also Wikidata oddities, such as "Marian" having different items for "male" and "female" versions: Marian (Q545486) and Marian (Q20155725). There's no guarantee that the right one will be used: Marian Hedwig Mülberger (Q55906557). Ghouston (talk) 04:09, 31 July 2019 (UTC)
I don't think that's a good idea. ChristianKl❫ 13:53, 31 July 2019 (UTC)
Imho it also sounds a bit risky. Given name may not unfrequently shifts gender accross countries or time period (or get mixed up with false friends) and I'm not certain Wikidata currently reflects all theses subtleties. Alexander Doria (talk) 13:59, 31 July 2019 (UTC)
Good example of this problem: Maria (Q25413386). - Jmabel (talk) 15:44, 31 July 2019 (UTC)
It sound more than "a bit" risky, I guess it would possible to cross this with other data to be sure it' actually men? For instance if it's a priest and with a male given name, then it's very likely a man (just to illustrate my point, it's not a very good example, there is only 8 priests in these results) Cdlt, VIGNERON (talk) 19:45, 31 July 2019 (UTC)
I actually see exactly the same issues if I fill in the gender by hand, so I should refrain by editing someones gender, until I had a chance to peek into their pants? Another question, why do we register gender, first of all, especially if it is so tricky to fill it in. Edoderoo (talk) 07:42, 1 August 2019 (UTC)
There can be references, typically where somebody is referred to with a gendered pronoun. Ghouston (talk) 09:15, 1 August 2019 (UTC)
Why would a gendered pronoun be more a more useful reference then the property on the wikidata item of the first name? Edoderoo (talk) 10:16, 1 August 2019 (UTC)
Because the gendered pronoun directly implies which gender the person has while first names are in many countries relatively unregulated and contain many expections. ChristianKl❫ 11:21, 1 August 2019 (UTC)
If you look more carefully to the SparQL, you will see that the gender is defined in the linked first name. For the names that have no specific gender, like Mischa (Q18229299) the query has no outcome. For firstnames that can be both male or female, like Marian, there are two wikidata-items. I assume that they are not randomly given to peoples wikidata-item. The above script should not set male to females, unless someone attached a male firstname to a female. Then the error is already there. Edoderoo (talk) 11:29, 1 August 2019 (UTC)
I've assumed that "male given name" just means the name is usually assigned to males, and there's nothing wrong with using that item for females too. Having two items for the same name would seem silly, because they are the same name after all. Ghouston (talk) 01:53, 2 August 2019 (UTC)
Personally I wouldn't name it silly, as it just depends how far do you want to specify a first name. You can easily defend that Maria, Marja, Mara and Marian have the same origin, and are therefor the same first name. About assumptions you are right. My assumption is that we can freely copy the assumed gender from the first name in case it is missing. When entering it manually, we do exactly the same thing after all. Edoderoo (talk) 08:17, 2 August 2019 (UTC)
Should we have male and female versions even of "unisex" names like Adrian (Q372250)? What would be the advantage? Every name is potentially unisex. I think a better idea would be to change the descriptions "male given name" or "female given name" to just "given name". I was momentarily confused by Adrian (Q372250) because the description in Dutch is "mannelijke voornaam", or "male forename", so I wondered if there was a female version, but there doesn't seem to be. In cases where there are two versions of a name, we also haven't considered which one to use for people of unknown sex or non-binary sex. Ghouston (talk) 08:52, 2 August 2019 (UTC)
@Edoderoo: We register gender because it's information that people care about. It allows automated scripts to find the right pronoun. It always people to answer questions such as "Which are the 5 biggest cities that are ruled by female majors" via SPARQL that are hard to answer without Wikidata. ChristianKl❫ 11:21, 1 August 2019 (UTC)
In Sweden we have the tradition that we add the name of the hamlet somebody comes from into the name. Then the name would still be female even if it is used on a male person. (Or the other way around.) IP (Hej!) 09:02, 2 August 2019 (UTC)

soweego 1.0 release

(Please disregard this message if you have already read it in the Wikidata mailing list, and apologies for the distraction)

  • TL;DR: soweego version 1 is out!
  • Like it? Star it!

Hi everyone,

The soweego team is delighted to announce the release of version 1! If you like it, why don't you click on the Star button?

soweego links Wikidata to large catalogs through machine learning. It partners with Mix'n'match, which mainly deals with small catalogs.

Soweego bot (talkcontribslogs) is currently uploading 255 k confident links to Wikidata: see it in action! 126 k medium-confident links are instead getting into Mix'n'match for curation: see the current catalogs [2], [3], [4], [5], [6], [7], [8], [9], [10].

The soweego team has also worked hard to address the following community requests:

  1. sync Wikidata to external catalogs & check them to spot inconsistencies in Wikidata;
  2. import new catalogs with reasonable effort.

Thinking of the best way to contribute? Try to import a new catalog.


Hjfocs (talk) 08:59, 31 July 2019 (UTC)

Have thought about this possible application of machine learning in the past, very cool stuff, will be keeping an eye on this! --SilentSpike (talk) 13:07, 31 July 2019 (UTC)
@Hjfocs: Just curious, could you please give a definition of „large catalogs“? What means that? A dataset with more than X records? --Succu (talk) 19:28, 31 July 2019 (UTC)
@SilentSpike: thanks for the kind words and the star on GitHub! Stay tuned for version 1.1 proposal: m:Grants:Project/Rapid/Hjfocs/soweego_1.1.
@Succu: thanks for the nice question. The order of magnitude is millions of records. For instance, the latest Internet Movie Database (Q37312) name dump (people) [11] contains 9.5 million entries roughly.
Cheers, Hjfocs (talk) 16:55, 2 August 2019 (UTC)

Searching only aliases and descriptions

list of most viewed online videos in first 24 hours (Q22954400), gl (Q29733870), and 3 branches of goverment (Q56490649) all contain "none" in either description or alias fields. Is there a way to search for these errors without also picking up every item that contains the word "none" somewhere else? WhatamIdoing (talk) 02:52, 2 August 2019 (UTC)

  • [12]--GZWDer (talk) 02:54, 2 August 2019 (UTC)
    • Thanks, GZWDer. I think I've cleaned them all out now, and I left messages for the two bots involved. (The rest were either IPs or people who use enwiki's "Shortdesc" script.) WhatamIdoing (talk) 03:13, 2 August 2019 (UTC)

sporadicall error - unable to achieve clickable link at qualifier point in time (P585)

Hi! Please look at Abraham Sutzkever (Q330652). I was not able to achieve a clickable link for the value added at point in time (P585). It is the first time that I can not get such a link. Not shure what interface language I used at the other contributions. Would be happy about any suggestion and fixing. Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 08:16, 2 August 2019 (UTC)

Merge request

Hi, could someone merge human being (Q28937368) with human (Q5). It is obviouly a duplication. thx, --Yanik B 08:10, 8 August 2019 (UTC)

  • I merged the items. It seems that the linked article on the French Wikipedia is the one for human.ChristianKl❫ 08:25, 8 August 2019 (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. Matěj Suchánek (talk) 12:19, 8 August 2019 (UTC)

Moving statements from one property to another

Is there a way to easily move statements from quote (P1683) to quotation or excerpt (P7081)? The former property should only be used in references to include supporting text from a quotation (and is apparently usable for qualifiers but I'm not sure if that's an appropriate use case), so it would be nice to have something similar to {{Autofix}} for those. Right now there are a few dozen constraint violations which would be resolved by changing to the newer property. Jc86035 (talk) 16:55, 29 July 2019 (UTC)

@Pasleim:.--GZWDer (talk) 17:00, 29 July 2019 (UTC)
Write mw.loader.load( '//ěj_Suchánek/moveClaim.js&action=raw&ctype=text/javascript', 'text/javascript' ); // [[User:Matěj Suchánek/moveClaim.js]] to Special:MyPage/common.js. Then reload the property page P1683, and you'll see move/copy buttons on each claim. —MisterSynergy (talk) 17:08, 29 July 2019 (UTC)
@MisterSynergy: I already had that installed. Did you copy the wrong JS? The script is only for copying statements between different items, not for changing the property while retaining the statements on the same item. Jc86035 (talk) 17:19, 29 July 2019 (UTC)
Oh I see, I misunderstood your request. You don't want to move claims between property pages, you want to move claims in items from one property to another. Yeah, that needs bot code, not a user script.
Fortunately, I have such code as well, adapted from Pasleim's version a while ago. If you describe your task in more detail, I can have a look whether I can help you. Pasleim seems to be quite busy offwiki these days. —MisterSynergy (talk) 17:52, 29 July 2019 (UTC)
@MisterSynergy: There are only 41 affected items, though I wouldn't really want to replace them all manually. The lexemes should probably be using usage example (P5831) instead, and I think most of the items should be using quotation or excerpt (P7081) (or the statements should be removed). As an example, on Article 1 of the Basic Law for the Federal Republic of Germany (Q712652), I would want the statements to essentially be unchanged except for the property number (the datatypes are the same). Jc86035 (talk) 16:00, 2 August 2019 (UTC)
Article 1 of the Basic Law for the Federal Republic of Germany (Q712652) is done per Special:Diff/991089102 (except for the copyright status (P6216) qualifier which is now being suggested). I just need QIDs to perform those moves, and it is fairly simple for me to do that. So if you sort that out, I'd do the moves. Lexemes are not supported by the move script anyways, thus you need to fix them manually. —MisterSynergy (talk) 19:07, 2 August 2019 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────@MisterSynergy: Having checked all of the statements on items, it seems most of the statements are problematic in other ways, though I would still move most of them to quotation or excerpt (P7081). I think a more in-depth discussion about the acceptable uses of quotation-related properties (e.g. through proposing a property for quotations about things, or proposing to widen the scope of quotation or excerpt (P7081)) would be necessary to determine whether we should keep this data.

In summary, at most four of these would work without any constraint violations (if none of the P7081 constraints are changed). Jc86035 (talk) 17:22, 3 August 2019 (UTC)

I just moved statements in 12 items, as indicated above. Let me know if you need more moves… —MisterSynergy (talk) 18:11, 3 August 2019 (UTC)

Unit box not appearing?

Trying to add area (P2046) to Toddbrook Reservoir (Q7348277), I can add the value, but the box to select the unit doesn't appear - and I can save the value without a unit. Is anyone else having this problem? Thanks. Mike Peel (talk) 17:58, 1 August 2019 (UTC)

Same issue with monolingual text. Ayack (talk) 18:10, 1 August 2019 (UTC)
Phabricator task T229604 seems related... ArthurPSmith (talk) 19:07, 1 August 2019 (UTC)
Is working again now. - Premeditated (talk) 20:13, 1 August 2019 (UTC)
It is, thanks to whoever fixed it. :-) Mike Peel (talk) 20:31, 1 August 2019 (UTC)
Many thanks to him or her! Nomen ad hoc (talk) 20:57, 1 August 2019 (UTC).
That was a very temporary fix, it just stopped working again :( Moebeus (talk) 22:13, 1 August 2019 (UTC)

sporadical error - unable to add qualifier title (P1476)

Hi! I added Nikolaus Lenau „Der Postillion" as YouTube video ID (P1651) to Fritz Stavenhagen (Q1440135). This is the third item where I can not add the qualifier title (P1476). Normally a second input field should open automatically requiring to add the mandatory addition about the language of the title / video content. Without the language code I can not publish my additions.The other cases I could not add title are: St. George's Cathedral (Q1062101) and Nikolaus Lenau (Q84280 . Please confirm and if possible let me know how to avoid this error. Thanks in advance! Regarcs
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 23:41, 1 August 2019 (UTC)

This is the same error as above and as reported here: and here: Very unfortunate bug, affects dates , titles, subtitles, units, etc. Moebeus (talk) 00:14, 2 August 2019 (UTC)
Thanks Moebeus for the feedback! Is there a possibility to add a trace qualifier to be able to fix the title at a later stage. I will probably find hundred othes as Alter Kacyzne (Q302639). Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 05:08, 2 August 2019 (UTC)
Please also see Wikidata:Project_chat#No_languages. --M2k~dewiki (talk) 16:28, 2 August 2019 (UTC)
Thanks M2k~dewiki! I added the missing titles including the language. Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 07:45, 3 August 2019 (UTC)

Son of the cousin

Hello. Do we have item for "the son of the cousin"? Something like niece (Q3403377). I found it as "first cousins once removed" in English. Xaris333 (talk) 01:59, 3 August 2019 (UTC)

Hans-Georg Münzberg

How do I add DNB ID 122752996 to his entry Q66103475? + there must be other authority control numbers. --Gerda Arendt (talk) 13:43, 3 August 2019 (UTC)

@Gerda Arendt: the easiest way: in your Preferences > Gadgets, turn on the "Authority Control" gadget. On the person's item page, click "Authority Control" in the left menu and + to add a whole set of identifiers at once. Watch out for duplicate names, and reload the page to see the entries. - PKM (talk) 20:17, 3 August 2019 (UTC)
Thank you. --Gerda Arendt (talk) 20:20, 3 August 2019 (UTC)
The "Identifiers" header does not need to be created manually. The statements are automatically moved into that "section" when they are created, so you can also use the link right above the "identifiers" header to add these properties without needing to scroll all the way down. Circeus (talk) 20:32, 3 August 2019 (UTC)

Merge request

Can someone merge and, preferably keeping the entry as Q302... ? Thank you!  – The preceding unsigned comment was added by (talk • contribs).

Hi dear IP,
you can do it yourself: please see Special:MergeItems.
Best, Nomen ad hoc (talk) 21:54, 8 August 2019 (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. Nomen ad hoc (talk) 09:28, 9 August 2019 (UTC)

reverted contribution

Hi! Q2438|oldid=prev&diff=990675452 is about a revert by Tacsipacsi concerning December 18 (Q2438). I might understand that the killing of inocent people during the start of the Romanian revolution is of litte interest to some contributors.But the addition is a valuable testimony especially for people from Timisoara / Romania. I would like to here some other opinions about this. Thanks in advance! Best regarcs
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 23:52, 1 August 2019 (UTC)

I don't think this would be a valid use of YouTube video ID (P1651), since the statement would violate the property's type constraint. Jc86035 (talk) 05:25, 2 August 2019 (UTC)
I agree. That property is for copies on Youtube of the audiovisual work described in an item (films, trailers, music videos, etc.). It's not intended as a way to add links to Youtube videos merely containing some information related in some way to the topic. --Kam Solusar (talk) 14:43, 4 August 2019 (UTC)

similar items

sound recording (Q5057302) and audio recording (Q3302947)

While searching I found sound recording (Q5057302) and audio recording (Q3302947). Aren't they pretty much the same? --D-Kuru (talk) 09:41, 2 August 2019 (UTC)

On a second thought after think about why one is linked from the other. One seems to be the action of recording, the other seems to be the result of that action. If so, we should update the description to be more clear. --D-Kuru (talk) 09:46, 2 August 2019 (UTC)
Or maybe audio recording (Q3302947) could be called audio record? The alias of sound recording (Q5057302) is audio recording which does not really help a lot here. --D-Kuru (talk) 09:56, 2 August 2019 (UTC)

central processing unit (Q5300) and processor (Q1466064)

As above, central processing unit (Q5300) and processor (Q1466064) look pretty much the same. Is there any page where such similar items can be discussed? --D-Kuru (talk) 11:20, 4 August 2019 (UTC)

Ineligible for copyright

There are a number of items representing concepts which could form and/or be described as creative works, but are nevertheless not copyrightable due to not being complex enough (e.g. C diminished seventh (Q58233626)). The Commons analogues are c:Template:PD-ineligible and c:Template:PD-shape. Is there an item which can be used as a determination method (P459) qualifier for copyright status (P6216) statements, or does one need to be created? Jc86035 (talk) 17:29, 3 August 2019 (UTC)

Common needs these templates to mark files, which could (in theory) be copyrighted, hence the need for the template. I don't think there is a need for something like that here. Circeus (talk) 20:35, 3 August 2019 (UTC)
Your example can be marked up like this; which is factually correct, but perhaps does not serve any useful purpose. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 08:58, 4 August 2019 (UTC)

same label and description

I can't understand why we can't have two items with the same label and description. Georgios Papadopoulos (Q53996986) Georgios Papadopoulos (Q66108465) Xaris333 (talk) 20:01, 3 August 2019 (UTC)

because the description are supposed to be disambiguating and if they are not, they're no more useful than there being no description at all? Circeus (talk) 20:40, 3 August 2019 (UTC)
The other reason, I think, is that if there are two items with the same label and description, it is quite likely that they're duplicates. So the prohibition is a nice reminder to double-check for that -- and if indeed they're not duplicates, the requirement to make the description distinct (to get past the prohibition) also serves to make the description more specific and more useful as a discriminator. —Scs (talk) 13:52, 4 August 2019 (UTC)

Movement Strategy online surveys - opportunity to share your thoughts about reworking movement structures

Community conversations are an integral part of movement strategy “Wikimedia 2030”. They have been ongoing in multiple formats and in numerous languages over the last 2.5 years. Now it is possible to also contribute to the development of recommendations on structural change via an online survey. We are keeping the survey open for additional 2 weeks and post it to wikis to provide wider opportunities to participate for people interested in it.

The survey is available in 8 languages: Arabic, English, French, German, Hindi, Portuguese, Simplified Chinese, and Spanish. They contain designated questions about each of the nine thematic areas that the working groups are analyzing and drafting recommendations for. You can freely choose the thematic areas you want to contribute and respond to. The survey questions have been created and designed by the members of the working groups.

Here is the link to the survey.

Here you can find more information about the survey.

With any questions, please contact me on my meta user talk page.

Thank you for your kind attention! --KVaidla (WMF) (talk) 14:41, 4 August 2019 (UTC)

p.s. If this is not the right place to post such message on your wiki, I apologize. Feel free to move it where appropriate according to your guidelines. Thank you!

Wikidata items for references/citations?

Can Wikidata be used to store references for Wikipedia? I mean, today a Wikipedia article uses a book as reference. My question is can we make an item on Wikidata for the book and then use the Wikidata item as a reference in an article? Like this, references could be tracked easily, being shared among multiple articles etc.--Kozuch (talk) 11:26, 4 August 2019 (UTC)

PS: In Wikimedia_Foundation_Annual_Plan/2018-2019/Final I can see following: "High quality citations support verifiability. To build a more reliable internet based in facts, we need a more robust network of references on and beyond Wikipedia." This may point to using Wikidata for citations.--Kozuch (talk) 11:30, 4 August 2019 (UTC)

  • Wikidata stores most of the fields needed for w:Template:Citation. I think there is some development ongoing to store it more systematically (cite extension?).
Also, Wikidata stores references for specific statements that can be displayed on Wikipedia (e.g. date of birth). Not that this is really new. --- Jura 11:43, 4 August 2019 (UTC)

That is a very interesting topic that you have brought up and will most likely be a hot topic for Wikidata in the time yet to come. I think that Wikidata Bridge will be the project to really make this change come alive, and connect Wikidata and Wikipedia like Wikipedia and Commons are linked today. If you are looking for an existing way of using Wikidata entries with w:Template:Citation. Will I recommend using w:Template:Wikidata. - Premeditated (talk) 13:25, 4 August 2019 (UTC)

  • I don't quite see the relation, but w:Template:Cite Q might be closer to what Kozuch has in mind. --- Jura 13:33, 4 August 2019 (UTC)
    Cite Q documentation has the following warning at the beginning:
Warning: From the TfD closure: "[...] works-in-progress should rarely be used in the article space until they are 99% finished/accurate/etc. Until the matter of transcluding Wikidata on Wikipedia is resolved (most likely with a huge and contentious RFC) usage of this template should be extremely vetted to ensure that all of the transcluded information is accurate."
 – The preceding unsigned comment was added by Jc3s5h (talk • contribs).
  • Well that's just enwiki and it's not even red there. Other Wikipedias might not require their volunteers to re-copy references for each statement. --- Jura 17:38, 4 August 2019 (UTC)
  •   Comment When you talk about making a Wikidata item for a book, you need to be careful, because a book (in Wikidata terms) should never be cited. A "book" can mean a particular work, or a particular edition of that work, or a particular physical copy of that edition. For that reason Wikiproject:Books has rejected the use of book (Q571). Instead you should cite an edition of a book. In other words, you need one data item for the work and another data item for the edition (specific year, specific publisher, specific ISBN, etc.), and you should cite that edition of the work. Different editions will have different publication details, possibly different page numbers, and even different text, so it is important to cite the particular edition used as a source. --EncycloPetey (talk) 21:03, 4 August 2019 (UTC)

Checking if a concept exist in Wikidata

In Dutch there is the railway term 'wisselstraat' for a combination of Double slips (Q14905968) crossing diagonaly the other tracks. In de Commons:Category:Railway points in Belgium there are many examples of this. Is there a Wikidata item for this? It is a combination of Q14905968. In Dutch there is only a redirect nl:Wisselstraat (spoorweg) is 'wisselstraat' kan also be a ordinary streetname. ('straat' is 'street' in English) As in Wisselstraat (Q29939452).Smiley.toerist (talk) 12:47, 31 July 2019 (UTC)

@Smiley.toerist: If you're referring to the concept of multiple double slips one after the other (as in the photo), I don't think this concept exists in most languages so it's probably safe to assume that a Wikidata item doesn't exist. If there isn't an existing sitelink then I probably wouldn't create an item for it, although I guess you could create a lexeme for "wisselstraat" and then link a new item for the concept to the lexeme. Jc86035 (talk) 16:09, 2 August 2019 (UTC)
I am thinking of using 'multiple double slips lane' for a Commons category. This is then a start. Can later be added to Wikidata.Smiley.toerist (talk) 14:00, 5 August 2019 (UTC)

"Error in Wikidata"

Can someone point me to or perhaps explain how to fix issues arising from "Error in Wikidata". In particular at Ovis aries. DonSpencer1 (talk) 03:41, 5 August 2019 (UTC)

I think the problem is that the VN template used by the Commons gallery is intended to be linked with a species item on Wikidata, but the gallery has been linked with a common name item instead. It should be linked with Ovis aries (Q29350771), not sheep (Q7368). Ghouston (talk) 06:45, 5 August 2019 (UTC)

Wikidata weekly summary #376

dedicated to / named after / titular

Hello! As you may know, churches, chapels and altars are named after a saint or another religious figure (God, Trinity), which is called "titular [saint]"; for example, the titular of a "Saint Paul church" is saint Paul. As far as I know, currently there are two properties used for it, property:P825 ("dedicated to") and property:P138 ("named after"); both are somehow correct, but a new "titular" property would be far more appropriate. Is it possible to create it? --Syrio posso aiutare? 12:02, 5 August 2019 (UTC)

Also commemorates (P547). What issue would another property address? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:37, 5 August 2019 (UTC)
And also patron saint (P417). --Matěj Suchánek (talk) 16:28, 5 August 2019 (UTC)
I'll be honest and say if something like that was proposed, I would very much oppose it. named after (P138) is the perfect property for this. Circeus (talk) 18:38, 5 August 2019 (UTC)
xkcd = 927 Jheald (talk) 19:48, 5 August 2019 (UTC)

Modeling the Survey of Scottish Witchcraft database on Wikidata - which property proposals to go forward with

Hi, currently have a property proposal in place for 'manner of torture/punishment' in place but looking at other interesting data in the Survey of Scottish witchcraft database and looking at where there maybe a good case to argue for inclusion on Wikidata and a new property proposal:

  • Date of torture (add as qualifier to manner of torture/punishment using Point in Time?)
  • Date of confession
  • Location of confession
  • Date of arrest
  • marital status - assume we would just link to spouses, children, siblings, relatives using those existing properties? However, this is sparse.
  • accused of - there is rich information in terms of the different things the accused witches are actually accused of e.g. devil's mark, shapechanging (to a cat, a dog, a hare, an apparition).
  • Ordeal type - technically distinct from the extrajudicial torture that took place although many ordeals involved painful procedures, this was a test conducted in order to let nature or God reveal the truth. Values would include: Bierricht (corpse bleeds), Ducking (otherwise known as the water test, Pricking (the body of the suspect witch was pricked with pins in order to find a Devil’s mark). I imagine we could characterise under manner of torture/punishment if having a separate property for ordeal proved inappropriate. Although this test would have formed part of the trial process.
  • Ordeal date.
  • Cause of malice - why the accused witches were accused of acting as they did (values such as revenge, debt, slander, grudge, failed business). Could Has Cause (P828) be used here? It is more of an accusation/complaint than a direct cause though.

Any thoughts on how these could work? Stinglehammer (talk) 11:15, 23 July 2019 (UTC)

@Stinglehammer: Try to use as much as possible significant event (P793):this reduces the need of specialized properties.
Ex: significant event (P793):confession (Q1219996)
point in time (P585):XX.XX.XXXX
Or significant event (P793):arrest (Q1403016)
point in time (P585):XX.XX.XXXX
location (P276):Y
Snipre (talk) 22:30, 23 July 2019 (UTC)
@Stinglehammer: There is the property convicted of (P1399) too. Snipre (talk) 06:48, 24 July 2019 (UTC)
Many thanks Snipre, this is very useful and a good workaround which encompasses much of what we're looking to achieve. We had planned on using significant event (P793) to link the accused witch to the item on their witch trial. I would just question how useful it is to have confession, arrest, trial and more all under the significant event (P793) umbrella from a modelling and querying point of view. If the general consensus is this is best practice then can certainly import with that method in mind. Any thoughts on the other property types I mentioned? Stinglehammer (talk) 11:45, 24 July 2019 (UTC)
I think an "accused of" property could be valuable and even extend easily to modern persons/companies (where subject item of this property (P1629) is accused (Q28056401)). The "cause of malice" should probably be specified as a qualifier of some sort, but I'm unsure what's best. I think all of the "date of" and "location of" can be done using significant event (P793). --SilentSpike (talk) 08:45, 29 July 2019 (UTC)
Can definitely look to put forward a property proposal for accused of if this would be useful elsewhere too. We've been working with adding date and location information via the significant event (P793) property as suggested above for confessions etc. One thing to note is that we have items of data for each instance of Scottish witch trial with trial dates now added. However the database also includes case dates whereby the case denotes the whole period of investigation the accused witch is subjected to so it could cover a ten year period with one or more or multiple instances of witch trials in this period. Therefore we were wondering whether significant event (P793) could be used again with a new item created for witchcraft investigation/investigated for witchcraft and then dates added as qualifiers. OR whether we might need to create a new item for EACH case and then have those link to the trials.e.g. Case of Isobel Gowdie. Thoughts on this would be welcome. Stinglehammer (talk) 14:58, 31 July 2019 (UTC)

Since so many of these properties involve dates, I should interject that I have emailed Prof. Julian Goodare, one of the editors of the database. The dates in the database use the calendar that was in effect at the time and place of the event described, which is nearly always the Julian calendar, since England and Scotland didn't convert to Gregorian until 14 September 1752. Also, Scotland considered the year to begin on 25 March until 1600, and England did the same until 1752, but the database always treats 1 January as the first day of the year. So dates from the database should be mapped to proleptic Julian calendar (Q1985786), not proleptic Gregorian calendar (Q1985727). Jc3s5h (talk) 14:04, 2 August 2019 (UTC)

Many thanks Jc3s5h, Prof Julian Goodare is aware of the project and meeting with us next week to discuss the work to date with a view to quality assuring. The dates seem to have defaulted to Gregorian on import so we are making changes accordingly. We now have items of data for (1) accused witches in the database (biographical, geographical data added) and these are linked to (2) the witch trials they underwent and (3) the people associated with those witch trials (sherriffs, lairds, judges etc.). Returning to my other question, we also have dates of 'cases' which are defined as "The witchcraft case follows the series of events and specific accusations that emerged from a denunciation of a person accused for witchcraft that was pursued by either ecclesiastical and/or civil authorities. The case encompasses the beginning (i.e. initial denunciation or supposed denunciation), middle (i.e. the investigation, arrest and trial of an accused witch), and end (i.e. the way the case ended in dropped charges, a verdict of innocence, or execution) of a witchcraft accusation. The ‘case’ does not only refer to a trial or trial process but also includes references to cultural beliefs about witchcraft practice that were found in the documents. It should also be pointed out that ‘case’ refers to one individual who was accused of witchcraft. The term does not refer to a series of trials in a large witch-hunt." e.g. .e.g. Case of Isobel Gowdie - would it be appropriate to have this modelled as significant event (P793) and linked to a newly created item for 'investigated for witchcraft which would be an instance of inquiry (Q21004260) and/or a subclass of witch hunt (Q188494) as a way of including the dates of the case/period of investigation? We have case dates for almost all 3,219 accused witches but trial dates for only 600 so modelling case dates would be very useful for querying/visualising purposes. Thoughts welcome on this. Stinglehammer (talk) 15:40, 6 August 2019 (UTC)

New data type for EntitySchemas

Hello all,

Based on community request, a new data type is about to be created, in order to link to Schemas from statements. You can already have a look at it on Beta (a test property has been created). The value type is string, so you will enter for example "E1" as a value.

You can learn more in the ticket. This data type will be available on August 7th. Cheers, Lea Lacroix (WMDE) (talk) 15:01, 5 August 2019 (UTC)

Thanks, but why string? Wouldn't it have been wiser to make them entities just as items, properties, and lexemes?
  • One could search for EntitySchemas by label, instead of having to type the E-ID.
  • Use in WDQS would have been more comfortable.
  • It would create backlinks.
    • That would make it easier to validate all statements with this data type via the pagelinks table at any time. Now the value is apparently validated during input (i.e. needs to start with "E" and then followed by numbers; E-ID needs to exist). However, if the EntitySchema is deleted or merged, it will be difficult to update this in statements.
    • Even more in general if everything goes smoothly, it would be interesting to easily find items which link to an EntitySchema.
--MisterSynergy (talk) 15:40, 5 August 2019 (UTC)
@MisterSynergy: With this datatype, it should be possible to unlock Wikidata:Property proposal/Shape Expression for class, that in turn will allow us to write human (Q5) : Shape Expression for class = E10
Presumably this will then add a line wd:Q5 wdt:Pxxx wd:E10 to WDQS (though @Lea Lacroix (WMDE): it would be good to confirm that this will indeed be the case).
You're right that, regarding backlinks, there does appear to be a problem -- eg shows nothing, whereas it ought to show as linking there. Jheald (talk) 19:05, 5 August 2019 (UTC)
In WDQS it will look like wd:Q5 wdt:Pxxx 'E10'. Those are strings, not IRIs, pretty much as with Commons category (P373). —MisterSynergy (talk) 19:15, 5 August 2019 (UTC)
Well that would just be a complete pain. I have added comment phab:T214884#5394245 to the Phabricator thread, that the datatype is not ready to go if this is what would be the case.
EntitySchema-valued properties ought to work in exactly the same way as Lexeme-valued properties, as per eg this query: Jheald (talk) 19:32, 5 August 2019 (UTC)

Hello all,

Thanks for your quick feedback. It seems that we have indeed more things to discuss and to investigate here, in order to find more sustainable solutions. I don't have a precise date but since part of the team will be busy with Wikimania, it will be postponed at least to the end of August. Lea Lacroix (WMDE) (talk) 07:36, 6 August 2019 (UTC)

Thanks, glad to read this. Some additional thoughts: if you make EntitySchemas entities rather than string, ...
  • ... they would come with a canonical URI (which we don't have yet, but ShEx validation needs this as well to reliably resolve relative URIs to absolute ones); see phab:T225778
  • ... their labels etc. could be incorporated in WDQS
  • ... the actual ShEx code could be linked in WDQS as well, e.g. wd:E10 foo:shex <> (with foo:shex being some quickly invented placeholder here, as I am not sure which predicate one would really want to use). We do something similar with images which are not by themselves in WDQS, but there are links to image files at Commons via wdt:P18
--MisterSynergy (talk) 08:02, 6 August 2019 (UTC)

New gadget: related items

Wikidata is a strongly interlinked database with many items having ingoing and outgoing links. Unfortunately, the user interface does not display any ingoing links ('inverse properties'). For example, it does not display which notable people are or were member of Footlights (Q857679).

I've created a gadget which enhances the user interface by exactly this kind of information. You can activate it under Special:Preferences#mw-prefsection-gadgets by setting a tick at relateditems. This will add you a new button to the bottom of each item with which you can load the incoming links.

The scope of service can be expanded by any user. Just add to a property page with item datatype a new statement inverse label item (P7087) linking the property to an item containg the inverse labels. --Pasleim (talk) 19:05, 5 August 2019 (UTC)

@Pasleim: very nice! Thank you for that! − Pintoch (talk) 08:18, 6 August 2019 (UTC)
That sounds an excellent idea! Many thanks. Nomen ad hoc (talk) 09:37, 6 August 2019 (UTC).
@Pasleim: I tried it on human (Q5)     and it seems it takes a long… time. Is there any risk it will fail on an out of memory error or something like that ? What could be cool is a limit on the number of loaded statements and a link to the equivalent sparql query for more results. author  TomT0m / talk page 10:40, 6 August 2019 (UTC)
Thanks for the tip. I will improve the performance of the query once I'm interface administrator. --Pasleim (talk) 15:49, 6 August 2019 (UTC)
@Pasleim: Thanks for turning this into a gadget. Does this work with statements on properties (or lexemes, etc.)? I tried to use this on inverse label item of (Q66205187) (which is, of course, now used in a inverse label item (P7087) statement on inverse label item (P7087)) and the inverse statement doesn't show up. Jc86035 (talk) 10:58, 6 August 2019 (UTC)
The gadget only works on item pages but it shows incoming links from items and properties. So if you go to anthem of (Q65933124) the gadget will show you inverse label item of (Q66205187)=anthem (P85). --Pasleim (talk) 15:49, 6 August 2019 (UTC)

Move Q24003762 to Q1496690

Can someone redirect Q24003762 to Q1496690? Thanks. --Almorca (talk) 10:52, 12 August 2019 (UTC)

Hi Almorca, you can do it yourself: just have a look at Help:Merge. Best, Nomen ad hoc (talk) 11:25, 12 August 2019 (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. Nomen ad hoc (talk) 16:13, 12 August 2019 (UTC)

Cleaning up the class tree of « Wiktionary linguistic edition »

It seems the superclass tree of Wiktionary language edition (Q22001389)     is really wrong. Can someone help cleaning this up ?

A linguistic edition of Wiktionary should never became … an industry according to Wikidata, for example. A linguistic edition should not became an instance of Wiktionary, as wiktionary is a concrete project and an instance of a project should not become an instance of another one. It should be a part of the bigger project, not an instance of it.

@Jarekt: who found this thanks to an error in a template in the talk page.

Note : Query to find the class tree author  TomT0m / talk page 13:23, 31 July 2019 (UTC)

  • Amongst the problems :
    • relationship between a mediawiki website and a mediawiki project. It seems to me a mediawiki project is something different than its numerical support, the website, hence I removed the claim that a mediawiki project is a website to assert that a mediawiki project has a website.


community project (Q5155040) is supposed to be

< community project (Q5155040)     > of (P642)   < community >

. Don’t we need a better property to denote that a group of person runs/contributes to some project ?

Is a computer service a human action ?


Currently, a database or a computer service are subclasses of « human action ». Should we have two types or services, one for automated one and one for human driven one to prevent one ? Currently we have


. This seems wrong. (@Adelsheim: who added this statement)

Difference between dataset and database

  • Currently database is a subclass of datasets … datasets are supposed to be collections of datas. Databases are also collection of datas. What’s the real difference ? These are actually the same concepts imho.

Ruud Koot
Daniel Mietchen
Tinker Bell
Jasc PL
Tris T7
Peb Aryan
  Notified participants of WikiProject Informatics and

In any case in the class tree there seem to be a big confusion between the database engines that manages the databases and the collection of data themselves. Should be cleaned up.

Database and models

It seems that currently, databases are subclasses of models. Is this correct ? Note that a database as a model, the database schema, but the relationship is more like an instanciation of the model that a subclassification. On the other hand, a database models something in the real world.

author  TomT0m / talk page 16:04, 2 August 2019 (UTC)

Redundancies : "A subclass of B subclass of C" and "A subclass of C"

shown by this query The "A subclass of C" statement is probably redundant in those cases, or an indication of a mess. author  TomT0m / talk page 09:25, 5 August 2019 (UTC)

"A subclass of B subclass of C subclass of D" and … "A subclass of D"
select ?class ?classLabel ?intermediaire ?intermediaireLabel ?intermediaire2 ?intermediaire2Label ?parent ?parentLabel  {
  ?class (^wdt:P279)* wd:Q22001389 ; 
         wdt:P279 ?parent .
  ?class wdt:P279 ?intermediaire . ?intermediaire wdt:P279 ?intermediaire2 . ?intermediaire2 wdt:P279 ?parent .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en". }

Try it! author  TomT0m / talk page 09:43, 7 August 2019 (UTC)

Zero-length value

Is it technically possible to set a zero-length value for an identifier statement? !!! (Q371) has the ID "" for artist ID (P7109) (link). Using just an underscore works as well, but it appears that their canonical database value is the empty string. This is probably because their website is sort of falling apart at the edges, but I think the empty string would still be more technically correct. Jc86035 (talk) 11:58, 6 August 2019 (UTC)

If you can't do it with the regular user interface you could try through the API (or maybe with Quickstatements)? ArthurPSmith (talk) 17:46, 6 August 2019 (UTC)
@ArthurPSmith: QuickStatements allows me to import the value and send it, but performing the action results in an error, so I don't think it's possible. Jc86035 (talk) 09:26, 7 August 2019 (UTC)
The interface just wraps around the API, just like QS... --Matěj Suchánek (talk) 09:59, 7 August 2019 (UTC)

What are some data cleanup methodologies

What could be the best technologies .  – The preceding unsigned comment was added by 2001:420:5441:1252:190e:26c6:6a1d:514b (talk • contribs) at 09:14, 7 August 2019‎ (UTC).

For manual work, Help:Merge and gadgets/tools, for semi-manual Help:QuickStatements. --SCIdude (talk) 09:39, 7 August 2019 (UTC)
In terms of which objects might need some cleanup there a lists with Constraint violations: Category:Constraint_violation_reports. --M2k~dewiki (talk) 10:14, 7 August 2019 (UTC)
  • Sounds like a topic for nice help page. --- Jura 12:09, 7 August 2019 (UTC)
The answer depend on the type of cleaning you want to do: bad modeled items, duplicates, wrong data, outdated data,.... Then which steps of the cleaning do you want to perform: detection of the errors or correction of the errors ?
To correct something you first need to find the error. To find the error, you need a reference and a comparison tool. Snipre (talk) 13:31, 7 August 2019 (UTC)

itWikiCon 2019 - call for submissions

(sorry for crosslang crossposting) Hi, I would like to point out that the submissions are open for the sessions of the ItWikiCon 2019 that will take place in Rome from 15 to 17 November. The reference page is as usual the one of the proposals, in which you can either add ideas on what you would like to see in this edition, or make a proposal of a presentation/workshop/seminar/working group/etc. that you want to make, following the procedure through the inputbox or using directly the proposal model that we have prepared.

The deadline is 13 October 2019, the date from which the program committee will evaluate all proposals received, defining, in the days immediately following, the official program of ItWikiCon 2019.

For any clarification or suggestion please write in the proposal talk page or send an email to and if you plan to participate, don't forget to sign in the participants page.

On behalf of the programme committee, I would like to thank all those who would like to contribute to making ItWikiCon 2019 rich and diversified--Ferdi2005 (Posta) 13:38, 7 August 2019 (UTC)

Which websites uses information from Wikidata?

I've heard Quora (Q51711) were using information from Wikidata but i can't see anything on Quora (Q51711)'s website to suggest that. --Trade (talk) 16:02, 4 August 2019 (UTC)

  • The delights of CC-0. They can take anything from Wikidata, up to and including reproducing every item on the site, and are under no obligation to acknowledge us.- Jmabel (talk) 16:17, 4 August 2019 (UTC)
  • There used to be a reference on manage pages like that talked about links to Wikidata and it seems to me it's not there anymore. Maybe they removed the interlinking when they didn't get the links to their new languages? ChristianKl❫ 06:48, 5 August 2019 (UTC)
  • Did you WMF/WMDE provide any support to this website? @Lydia Pintscher (WMDE): --- Jura 10:07, 5 August 2019 (UTC)
    • Yes I had a few discussions with them and I believe that because of that and other people's work they're actually one of the nicer consumers of our data. They wrote blog posts about their work. They fixed data issues directly on Wikidata. They encourage their users to directly edit Wikidata if there is a mistake. They show the Wikidata logo in a comparatively prominent place for such a large website as Andy said. --Lydia Pintscher (WMDE) (talk) 07:57, 6 August 2019 (UTC)
      • @Lydia Pintscher (WMDE): Are they still encouraging users to directly edit Wikidata? I haven't found it anymore in the interface. ChristianKl❫ 20:51, 6 August 2019 (UTC)
        • Hmmmm you're right. They seem to have changed things. They do show the Wikidata connection at the top of the right sidebar of the manage pages for topics still. I'll see if I can find out more. --Lydia Pintscher (WMDE) (talk) 19:26, 7 August 2019 (UTC)

Item search: Please show thumbnails

Choosing an item: much better with thumbnails
  • Problem: It is hard to guess what item is the one I meant, especially when my locale (not English) has very few Wikidata labels and even less descriptions. Imagine the screenshot at the right with no descriptions nor thumbnails.
  • Solution: Display thumbnails based on each item's image (P18).

This could be used in the "Search Wikidata" field, and also in the edition dialogs, such as when selecting an item for instance of (P31).

Why is this important? Because people sometimes select the wrong item, generating much confusion.

Thanks! :-) Syced (talk) 05:46, 2 August 2019 (UTC)

Good point - Salgo60 (talk) 13:03, 2 August 2019 (UTC)
  • Yes, (as shown) it's already happening in mobile view ..--- Jura 17:26, 2 August 2019 (UTC)
  • I agree that images would be usefull When available I think the thumbnails should be based on icon (P2910) and image (P18) should be used as a fallback. ChristianKl❫ 14:52, 5 August 2019 (UTC)
  • Would you like this to happen only for the search box in the top right corner of the page or also when selecting an Item to link to in a statement? --Lydia Pintscher (WMDE) (talk) 08:06, 6 August 2019 (UTC)
    • I'd say the search box is the more important, but once the technology is implemented it should be pretty easy to let users configure this, no? - Jmabel (talk) 15:19, 6 August 2019 (UTC)
      • Heh it might actually be hard to prevent it since they share code ;-) But let's start with what you'd like and then see if/how we can make it possible. --Lydia Pintscher (WMDE) (talk) 15:47, 6 August 2019 (UTC)
    • Having it at both places would be valuable. Later one when statements can be created from Wikipedia that search interface likely should likely also have it. Item search as well as property search should have it. I'm not certain about lexemes. ChristianKl❫ 18:54, 7 August 2019 (UTC)
    • It would also be nice to display the images along with the description when hovering over an item. ChristianKl❫ 07:36, 8 August 2019 (UTC)

How a human becomes faultly an artificial entity

--Micru (talk) 21:46, 24 August 2014 (UTC) Tobias1984 (talk) TomT0m (talk) Genewiki123 (talk) Emw (talk) 03:09, 9 September 2014 (UTC) —Ruud 16:15, 9 December 2014 (UTC) Emitraka (talk) 14:32, 14 October 2015 (UTC) Bovlb (talk) 19:10, 21 October 2015 (UTC) Peter F. Patel-Schneider (talk) 22:21, 23 October 2015 (UTC) ArthurPSmith (talk) 15:51, 5 November 2015 (UTC) --Daniel Mietchen (talk) 20:53, 3 January 2016 (UTC) --Harmonia Amanda (talk) 22:00, 27 February 2016 (UTC) --Lechatpito (talk) --Andrawaag (talk) 14:42, 13 April 2016 (UTC) --ChristianKl (talk) 16:22, 6 July 2016 (UTC) --Cmungall Cmungall (talk) 13:49, 8 July 2016 (UTC) Cord Wiljes (talk) 16:53, 28 September 2016 (UTC) DavRosen (talk) 23:07, 15 February 2017 (UTC) Vladimir Alexiev (talk) 07:01, 24 February 2017 (UTC) Pintoch (talk) 22:42, 5 March 2017 (UTC) Fuzheado (talk) 14:43, 15 May 2017 (UTC) YULdigitalpreservation (talk) 14:37, 14 June 2017 (UTC) PKM (talk) 00:24, 17 June 2017 (UTC) Fractaler (talk) 14:42, 17 June 2017 (UTC) Andreasmperu Diana de la Iglesia Jsamwrites (talk) Finn Årup Nielsen (fnielsen) (talk) 12:39, 24 August 2017 (UTC) Alessandro Piscopo (talk) 17:02, 4 September 2017 (UTC) Ptolusque (.-- .. -.- ..) 01:47, 14 September 2017 (UTC) Gamaliel (talk) --Horcrux92 (talk) 11:19, 12 November 2017 (UTC) MartinPoulter (talk) Bamyers99 (talk) 16:47, 18 March 2018 (UTC) Malore (talk) Wurstbruch (talk) 22:59, 4 April 2018 (UTC) Dcflyer (talk) 07:50, 9 September 2018 (UTC) Ettorerizza (talk) 11:00, 26 September 2018 (UTC) Ninokeys (talk) 00:05, 5 October 2018 (UTC) Buccalon (talk) 14:08, 10 October 2018 (UTC) Jneubert (talk) 06:02, 21 October 2018 (UTC) Yair rand (talk) 00:16, 24 October 2018 (UTC) Tris T7 (talk) ElanHR (talk) 22:05, 26 December 2018 (UTC) linuxo Gq86 Gabrielaltay Liamjamesperritt (talk) 08:44, 21 June 2019 (UTC) ZI Jony Ivanhercaz (Talk) 11:07, 15 July 2019 (UTC) Gaurav (talk) 22:39, 24 August 2019 (UTC) Meejies (talk) 04:38, 29 August 2019 (UTC) SilentSpike (talk)

  Notified participants of WikiProject Ontology

per this query A human is a person, which is a individual (Q795052) which is a unit of analysis (Q7887142)     which is a unit of analysis (Q7887142) which is a unit which is an abstract object which is a work which is a product which is an artificial entity.

There is definitely something wrong in subclassing unit with human. author  TomT0m / talk page 13:33, 7 August 2019 (UTC)

I solved this by replacing a «subclass of» unit of analysis with an « instance of » claim. It seem more correct to say that a « unit of analysis » is a type of things that is taken to be measured in a scientific study than to say that Einstein is a unit of analysis. @Fractaler:. Or when metamodelling is useful … author  TomT0m / talk page 13:57, 7 August 2019 (UTC)
Would it be better to break it at unit (Q2198779) and abstract object? Couldn't a unit be, say, a specific car? That is not an abstract object. — Finn Årup Nielsen (fnielsen) (talk) 16:40, 7 August 2019 (UTC)
I readded omnivore (Q164509) to regain the link to organism (Q7239) however this may have (arguably) introduced some additional noise. The reason for this seems to be that component (Q1310239) is a conflation between an "abstract piece of something" and a "type of manufactured good (Q22811462)". I'm not sure how the latter is intended to be used but maybe deep down we're all just type of manufactured good (Q22811462)s and I'm just too prideful to admit it. :P --- ElanHR (talk) 05:08, 8 August 2019 (UTC)
Biological components shouldn't be type of manufactured good (Q22811462). I introduced the new nonbiological component (Q66310125) for things that are type of manufactured good (Q22811462). ChristianKl❫ 07:24, 8 August 2019 (UTC)

Chaib aisa musico

Chaib aisa, es el apellido de un músico y compositor canario ha colaborado con importantes bandas compositores como sergio Pizarro Jack Bruce así con innumerables músicos.  – The preceding unsigned comment was added by (talk • contribs) at 5 August 2019‎ 07:39 (UTC).

Probably better to write to spanish Wikidata:Café. --Infovarius (talk) 14:17, 13 August 2019 (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. Nomen ad hoc (talk) 20:25, 14 August 2019 (UTC)

Company Name imports from GRID

Hi Wikidata folks. At the moment, I believe Wikidata downloads the entire company name from the GRID Database for entities, e.g., which may cause some confusion with international brands, institutes, etc. and also multiply entries unnecessarily, e.g. and Is there some way to exclude parenthetical information from the pulled GRID Company Name for future entity references, and does anyone have any ideas about fixing the multiples problem for existing content? Thank you.  – The preceding unsigned comment was added by (talk • contribs).

Thanks for bringing this to our attention. It appears that these two items were created by different means, and only the former comes (directly) from GRID. This question might be better addressed to the maintainer of @APSbot, @ArthurPSmith. Bovlb (talk) 00:27, 9 August 2019 (UTC)
It looks like it's been merged now. I'll look at leaving out the parenthetical country components of GRID names in the future - it's also a good idea to be a little more careful when creating new entities to match existing things - a simple search in the UI would have found the old entry in this case for example. ArthurPSmith (talk) 02:14, 9 August 2019 (UTC)

Question regarding Module:Wikidata in Wikipedia

Hi, I have a question regarding some usage of the module, I hope someone can help me with this.

In the article id:Volgograd I'm trying to show the "instance of (P31)" of the geographical entity of "Volgograd (Q914)", but it shows several lines. The code that was used to invoke it was: {{#invoke:Wikidata |claim |formatting=table |property=P31 |rowformat=$0}}. How can I modify it so that it only shows the best statement (getBestStatements) of the property? Thanks. Bennylin (talk) 23:23, 8 August 2019 (UTC)

I figured it out by adding "| list=firstrank ". Bennylin (talk) 00:14, 10 August 2019 (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. --SilentSpike (talk) 10:35, 14 August 2019 (UTC)

taxon name (P225) and taxon rank (P105) usage in instances of group of organisms known by one particular common name (Q55983715)

Entities using the taxon rank (P105) property should be instances of taxon (Q16521) (or of a subclass of it), but pig (Q787) currently isn't.

I got the warning above when I added taxon name (P225) and taxon rank (P105) properties in an item that is an instance of "group of organisms known by one particular common name (Q55983715)"

In my opinion, the warning should not be displayed for instances of Q55983715, because it is possible for a group of organism to be in the same taxon rank as in this example (pig - genus Sus), or dog, or cat, etc. Bennylin (talk) 05:43, 7 August 2019 (UTC)

The Wikidata:WikiProject Taxonomy probably answers this frequently. --SCIdude (talk) 09:36, 7 August 2019 (UTC)
It seems pig (Q787) and Q57817848 could be merged based on the British English language labels and enwiki text, but not based on the wikidata item (which has instance of (P31) group of organisms known by one particular common name (Q55983715) of (P642) Q57817848 and Sus scrofa (Q58697).
Since the later statement is unsourced, and doesn't seem to match any of the wikipedias which I understand, I will merge the two. NisJørgensen (talk) 14:27, 7 August 2019 (UTC)
When you do based on what you do not understand, chances are that you get it horribly wrong.. Please don't. Thanks, GerardM (talk) 09:09, 10 August 2019 (UTC)

VIAFbot added unnecessary commas

User:Maximilianklein's User:VIAFbot included commas when it added Chinese labels and aliases. I recently found two: special:diff/991812104 special:diff/994268733. A bot should probably check its contribs to remove all these commas.--Roy17 (talk) 19:10, 9 August 2019 (UTC)

Adding transliteration in native label (P1705)

How can I add the native label (P1705) in transliteration form? For example I added Gangneung (Q42056) the Korean (Hangul), but how can I add the Hanja, the Revised Romanization, and McCune–Reischauer Romanization? It seems that the language option for native label (P1705) was for the language, while I think it should be for the writing system instead. Bennylin (talk) 04:30, 10 August 2019 (UTC)

I hope this is correct [17] Bennylin (talk) 19:25, 17 August 2019 (UTC)

Merging two items

Hi all, I have done some merges before, but always with a degree of worry! I've noticed two items for the same Library in Skerries, Dublin, but the Merge tool doesn't want to merge them: Q55049516 and Q56260806. Thanks! Smirkybec (talk) 19:45, 16 August 2019 (UTC)

@Smirkybec: And good thing it wasn't merged. Buildings and their occupants should be different items. This is the same reason we keep Q694178 and Google (Q95) as separate items. Circeus (talk) 20:01, 16 August 2019 (UTC)
Ah, fair enough! Smirkybec (talk) 20:04, 16 August 2019 (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. Nomen ad hoc (talk) 20:06, 16 August 2019 (UTC)

Multi-language default editor for labels

Hi, I have some technical question

Right now when I'm viewing items, the default language is shown in two row, Indonesian (my default language), and English (I guess site default? I never set it manually). But now what should I do to add more languages in the default label editor (for example Javanese - jv). I tried to change my preference -> editing -> translation option, from "en" to "en, jv", but it still doesn't show it for me. Any help would be appreciated. Thanks. Bennylin (talk) 16:09, 9 August 2019 (UTC)

@Bennylin: You need to edit your babel template. There is no option' to set up it in the settings which is the biggest missing basic feature you can only imagine. There should be also option to add more than 20/30 languages. Eurohunter (talk) 16:14, 9 August 2019 (UTC)
Got it. Thanks! Bennylin (talk) 16:37, 9 August 2019 (UTC)
Both of you might want to try the "labelLister" gadget from Special:Preferences#mw-prefsection-gadgets. Activate it, and you'll see a new "labels list" tab over item pages that allows you to modify (add, change, remove) labels, descriptions, and aliases in any supported language. —MisterSynergy (talk) 17:28, 9 August 2019 (UTC)
@MisterSynergy: I have it but I want to have listed more than 20/30 languages at once. It impossible to edit when you have to add additional 50 wiki codes to be able to add labels. Eurohunter (talk) 17:47, 9 August 2019 (UTC)
Seems that I do not exactly understand what your workflow is. It sounds that User:Jitrixis/nameGuzzler.js could help you, but it is already imported in User:Eurohunter/common.js. Does it work, and did you try it? --MisterSynergy (talk) 20:33, 9 August 2019 (UTC)
@MisterSynergy: Good point. I already forgot about it. It never worked for me. Nothing pop ups. Eurohunter (talk) 20:36, 9 August 2019 (UTC)
It should appear unter "Tools" section in the left menu bar, with the title "VIP's labels". If it does not work, edit User:Eurohunter/common.js and replace importScript('User:Jitrixis/nameGuzzler.js'); by mw.loader.load( '//', 'text/javascript' ); // [[User:Jitrixis/nameGuzzler.js]]. Does it work then? —MisterSynergy (talk) 20:43, 9 August 2019 (UTC)
@MisterSynergy: I didn't knew that there is "VIP's labels", I usually not loking there. Thank you. Eurohunter (talk) 20:55, 9 August 2019 (UTC)
A related question, how can I add a secondary script, in this case Javanese script, in the Javanese label? Did I do it correctly here [18]? Thanks. Bennylin (talk) 00:34, 10 August 2019 (UTC)
@Bennylin: I think this would definitely be a valid alias. Perhaps it would be better to add the Javanese-script name as a statement, but I'm not sure how this would work (maybe you could add writing system (P282) as a qualifier to a official name (P1448)/native label (P1705) statement). Jc86035 (talk) 11:52, 11 August 2019 (UTC)

Name of Wikidata properties in a dynamic Infobox

Do we need more labels for Wikidata properties when presenting it in a different context like Infoboxes? can we use the name property name (P2561) or name of issued by (P2378) ?

  • when we are in edit mode in Wikidata we would like to see a label telling "Find A Grave memorial ID"
  • when presenting external identifiers in a dynamic infobox we maybe should have another label "Find a grave:"'

Examples of "Infoboxes" displaying all labels

  • Wikidocumentary George Washington see Facts
  • user Larske did a test Template in sv:Wikipedia playing around see link
    • we tested using different labels (Larske correct me if I am wrong)

@Susannaanas, André Costa (WMSE), Larske: - Salgo60 (talk) 10:48, 10 August 2019 (UTC)

@Salgo60: The test function "auktoritetsdatatest1" can now be called with an optional parameter shortnames. shortnames=ja gives the result that the short name (P1813), if it exists, of the external identifier is used as (link) text in the box. See the sandbox. Note than only a few external identifiers do have a short name (P1813) defined, se this list, but you can try to add some more short names for identifiers with very long labels. It can be a challenge to define a short name (P1813) that is short but still unique and possible to understand. --Larske (talk) 08:26, 11 August 2019 (UTC)
Larske thanks I am on my way bicycle to Uppsala to take some gravestones photos so I will do a try later tonight for dear August Strindberg (Q7724) properties...
OT: regarding the "show all template in sv:Wikipedia" I right now feel a Wikidocumentary solution maybe is step 1 (ex. en,sv) and if other people think "show all" is the way forward we can have a new discussion in sv:Wikipedia what to do. Right now I feel the biggest problem is that we have a weak understanding of what Authority data is which is a step in the wrong direction... - Salgo60 (talk) 08:37, 11 August 2019 (UTC)

P180 on abstract or diagrammatic images

How would one use depicts (P180) to describe something like an abstract drawing or a simple geometric shape? I've been thinking about using structured data for the BSicon (Q21152830) image set on Commons, and this would probably be an issue; the individual images are often almost meaningless on their own, but may represent specific things and can be used to draw complicated diagrams. Jc86035 (talk) 07:12, 11 August 2019 (UTC)

  • That's something you probably want to discuss on Commons. I doubt the item BSicon (Q21152830) itself should have P180. --- Jura 11:59, 11 August 2019 (UTC)

You had one job... Face-smile.svg

Fixing the interwiki confusion that was monoicous (monoicous (Q150053)), vs monoecious (monoecious (Q66368485)).

I have created the latter item (seems we still can't link to redirects), and added a couple of "known" wikipedia articles.

It will need some polygloty as well as botany to get all the wikipedia links to the correct place.

Thanks in advance for helping with this.

All the best: Rich Farmbrough10:30, 11 August 2019 (UTC).

@Rich Farmbrough: You can link to redirects, but it's necessary to blank the page first (or otherwise remove the redirection). After linking you can restore the redirect.
Also, if the two concepts are very similar to the point that the different sitelinks mostly cover the same topics, you can use Template:Interwiki extra (Q21286810) to add the correct interwikis to the respective articles without having to modify any redirects. Jc86035 (talk) 11:39, 11 August 2019 (UTC)
Thanks! I performed the hack to add the en entry. Good luck with the rest. All the best: Rich Farmbrough11:44, 11 August 2019 (UTC).
You may find this page helps

Add statement on the top

Does anyone have a skin, which would place Add statement link on the top of the page. Its loss of time, to scroll down every time. --Juandev (talk) 13:28, 10 August 2019 (UTC)

@Juandev: I use the keyboard shortcut gadget and press "A", that works pretty well? See Keyboard_shortcuts Moebeus (talk) 12:42, 11 August 2019 (UTC)
The link goes to unrelated discussion. The gadget opens second add statement, I need the first. An it opens it on the lower botton of a vissible screen, which does not scroll doen when there is a drop down window, so its usability is hard.--Juandev (talk) 14:41, 11 August 2019 (UTC)
Ah, it had been archived, here you go: Wikidata talk:Tools/Archive 1#Keyboard shortcuts Moebeus (talk) 21:11, 11 August 2019 (UTC)
You can try Wikidata:Recoin, it is a good tool for existing items (not for new one).--Jklamo (talk) 16:54, 10 August 2019 (UTC)
I will write it in Czech, because, you have probably misunderstood, what I am talking about.//Píšu česky protože jsi asi nepochopil o co mi jde. Tahle fičura je sice hezká, ale vůbec neřeší můj problém. Tím problémem je to, že pokud chci k položce přidat obrázek, musím scrollovat až dolu. To se stává problémem v momentě, kdy tento krok dělám stokrát, protože ty sekundy scrollování jsou z nich rázem minuty a možné hodiny ztracené času. Takže hledám někoho, kdo si to třeba šoupnul nahoru přes CSS. Mě se ten blok nepodařilo najít ve zdrojovém kodu stránky.--Juandev (talk) 15:06, 11 August 2019 (UTC)
Juandev, you're not alone: Wikidata:Project_chat/Archive/2019/01#add_statement_button_at_the_start. Time and time again wikidata cannot get the most basic problems resolved but instead tells users to use all sorts of hacks (which wastes a lot more time).--Roy17 (talk) 00:06, 12 August 2019 (UTC)

Vote for Pasleim ;)

Pasleim needs votes for being an interface admin and making the awesome new relateditems work better. The page gets little traffic and thus we only have 6 of the 8 needed votes with an additional day to go: ChristianKl❫ 18:55, 11 August 2019 (UTC)

Uhm, we should review Wikidata:Requests for permissions/Interface administrator/Header and align it with Wikidata:Interface administrators. According to the former, already-admins need 75%+ support and the candidacy needs to be open for 3+ days. However, the policy does not support such a path. —MisterSynergy (talk) 19:04, 11 August 2019 (UTC)

Results not showing up in Mix'n'Match?

Yesterday i tried to search after Avenged Sevenfold (Q165193) but the website said that 'No matches found'. I tried again the next day and now it suddenly works. Despite that, i'm still getting the same message when searching after Panic! at the Disco (Q277551). --Trade (talk) 20:58, 11 August 2019 (UTC)

Wikidata weekly summary #377

Item record

Russia (Q159) have 289 entries on Wikipedia. Is there other item that have higher number of entries? -- 16:27, 12 August 2019 (UTC)

Depends what you are counting, but Category:User en block (Q5626526) has 501 sitelinks. ArthurPSmith (talk) 17:47, 12 August 2019 (UTC)
I count only Wikipedia links. —- 17:52, 12 August 2019 (UTC)
Finland (Q33) Eurohunter (talk) 17:55, 12 August 2019 (UTC)
So is Finland (Q33)... Thanks Eurohunter!!! - 17:59, 12 August 2019 (UTC)
Wikidata:Database_reports/Most_sitelinked_items --- Jura 18:07, 12 August 2019 (UTC)


Hi, can you move Q66123117 to the title "Typhoon Lekima" from "Tropical Storm Lekima"? I have no idea on how does pages work on Wikidata. --Znotch190711 (talk) 10:58, 14 August 2019 (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. Matěj Suchánek (talk) 13:59, 18 August 2019 (UTC)

Emperor of China / Chinese Empire

Emperor of China (Q268218) - Covers two concepts that should be differentiated. Could someone look into this? --Beat Estermann (talk) 12:18, 12 August 2019 (UTC)

I think Chinese Empire (Q12060881) is the other item. Ghouston (talk) 03:20, 13 August 2019 (UTC)

Edit request for a protected page

A simple edit request:

Also, the only indication that the page is protected is the lock mark on the top right corner of the page. You can't even click on it in order to get proper information on why you can't edit the page. Please fix this; it's confusing as it doesn't give advices on what to do when a page is protected. 18:06, 12 August 2019 (UTC)

No, in cases like this {{Edit request}} should be used directly on the item’s talk page, as far as I’m aware. No need to put it on the noticeboard too. —Galaktos (talk) 21:52, 12 August 2019 (UTC)
It's not clear why the item is protected. Some leftover from Abian's. --- Jura 22:01, 12 August 2019 (UTC)
    • The administrator noticeboard is a place where the request to protect or unprotect an item can be made. It's not a place for edit request that can be done by every autoconfirmed user. ChristianKl❫ 09:41, 13 August 2019 (UTC)
  Done by DannyS712Galaktos (talk) 23:20, 12 August 2019 (UTC)
Well, we don't want autoconfirmed users having to deal with items where it's unclear why they had been protected in the first time. --- Jura 13:10, 13 August 2019 (UTC)
I don't see a reason for that. I think requests should be posted where there's an audience that can help fulfill the request. In a case like the above I don't see how it's important to know why the item was protected to move the link. Even if it would be important it's not necessary to be an admin to look at the history page and see " Protected "Ferdinand (Q3276175)": High risk, one of the 3000 most used Wikidata Items on Wikimedia projects". ChristianKl❫ 14:20, 13 August 2019 (UTC)
There is nothing in our protection policy that authorises this. Please avoid having additional volunteers do redundant work when it could have been done by one person. --- Jura 15:15, 13 August 2019 (UTC)
Protecting items does create additional work but it also helps with vandalism. It's a tradeoff. We had an RfC about whether we want to protect high use items and found a community consensus whether or not you like the consensus that was found. It costs energy as a community when people don't accept the consensus of the RfC's we have and try to fight the decisions. ChristianKl❫ 15:58, 13 August 2019 (UTC)
As seen in the subsequent discussion, the RFC didn't authorize the protection of all items Abian protected, but apparently Abian didn't accept that, didn't unprotected items protected in violation of protection policy nor attempt to make steps to implement the RFC in a reasonable way. So users like me end up having to deal with the fall out. --- Jura 16:05, 13 August 2019 (UTC)
Abian seem to think that all of those 3000 most used items have >500 references. Are you saying that you know that some of those don't have that many references? ChristianKl❫ 16:50, 13 August 2019 (UTC)

New proposed data model: Post-conviction relief

I'd like to propose a new data model, but I can't find a relevant WikiProject through which to propose it. It would include a bunch of new properties. This is an extension of my proposal to add an "exonerated of" property. The purpose of this proposal is to better represent the complexity of legal systems, especially the ways in which a person may be found to be innocent (exoneration) or less culpable (commutation) after an initial conviction. (If there's a better place to start this conversation, please let me know.)

Data Model for "Post-Conviction Relief"

What is it: the category of legal actions taken after someone is convicted of a crime that can lead to reduced or eliminated sentences. This model covers pardons, exonerations, and commutations. In general, I propose that these be represented by pairs of properties -- one would be a "post-conviction relief applied" qualifier to the "convicted of" property, and the other would be a standalone property in its own right ("exonerated of", "pardoned of", "received commutation for")

New qualifier

Convicted of: <crime>

  • Point in time
  • Etc…
  • NEW: Post-conviction relief applied: <pardon|exoneration|commutation>  -- constraint: must have corresponding property, e.g.: exonerated of

New properties

-Kenirwin (talk) 20:03, 12 August 2019 (UTC)

The model here is that you normally are commuted after 2/3 of the time. You have to do something extra ordinary to do full time. IP (Hej!) 07:34, 13 August 2019 (UTC)
Similarly in the UK. But one might want to record released = date, and tie it to a particular prison term. Jheald (talk) 12:09, 13 August 2019 (UTC)

Stock code properties discussion

Note that there is stock code/ticker related property discussion at Wikidata:Property proposal/Bursa Malaysia stock code. As result, there may be 200+ possible new properties.--Jklamo (talk) 15:17, 13 August 2019 (UTC)

Discussing Inclusion criteria for lists like television series episode (Q21191270)

For well known series like Star Trek episode (Q61220733), the included episodes on Wikidata is quite exhaustive. (700+ items) Similarly for The Simpsons (Q886) where it has about 600+ items.

But for a smaller show like Suits (Q370185), even when at least some episodes are on some language wikipedia, there are no records found.

I think for the sake of consistency, it makes most sense that to modify the inclusion criterion to something along the lines of

That creates completeness, rather than half data for series that are not quite as popular but still popular enough to warrant some Wikidata entries. And I believe on Wikidata we should strive towards complete data as much as possible

Soni (talk) 20:41, 7 August 2019 (UTC)

Where is this inclusion criteria that you would have amended? Wikidata:Notability/Inclusion criteria is not it. Can't find it within Wikidata:WikiProject Movies. Nor is it covered AFAICS at Wikidata:Notability; but that page makes the point that items are notable if they meet some structural need, which episodes of a series surely do. As to the proposed change: should is inappropriate for an inclusion criteria, in the sense that the criteria for item A should not be 'oh go on then, create also items B - Z'. Could is appropriate for an inclusion criteria, but the structural need notability criteria seems to have that angle covered already? --Tagishsimon (talk) 21:00, 7 August 2019 (UTC)
  • If not all episodes of "Suits" are here it's just that nobody bothered including them. --- Jura 21:04, 7 August 2019 (UTC)
  • With the current standards there's nothing stopping someone from creating items for all episodes of Suits. We don't need to change any policy for this case. There might be cases where a YouTube user has a "TV series" with 1000 episodes and one of those episodes is a source for a Wikidata claim. In a case like that I woulnd't want a policy statement that says we automatically should add items for all the episodes. ChristianKl❫ 07:12, 8 August 2019 (UTC)
    • This. Nothing is witholding you to document every episode of a notable television series. Just takes some time. Sjoerd de Bruin (talk) 09:07, 8 August 2019 (UTC)
      • Thank you all for this clarification. I did not know if documenting every episode was a good idea or not, and did not want to waste time doing so, only to be deleted. Hence thought it's better to clarify policy/ask directly for the policy to be changed before I start doing so. Soni (talk) 19:12, 13 August 2019 (UTC)

Looking to create a dataset for our website

Our site appears in Wikidata, but I believe that there isn't an actual dataset for our movies/tv-shows on your website.

The idea is to export all the URLs from our website with a small description, so links will appear at the bottom of each movie/tv-show in relations to data that appears on our end and link the two.

What is the best practice for it, and how can I get it done?  – The preceding unsigned comment was added by ‎Geverem (talk • contribs).

What "our site" and "our movies/tv-shows" specifically? ChristianKl❫ 20:48, 13 August 2019 (UTC)


Meladeia (Q6810943) "This property's value should not be present on any other item, but is also present on P6810943 (Deleted Property) and Meladeia (Q6810943)". But is the same item. What is the problem. Xaris333 (talk) 23:07, 15 August 2019 (UTC)

I had something similar on Tomb Chest, 6 Metres South Of Chancel Of All Saints Church (Q26522409); it looks like it's related to a bug that was reported: ( Peter James (talk) 23:13, 15 August 2019 (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. Matěj Suchánek (talk) 07:08, 19 August 2019 (UTC)

What are P44395 and P103510?

Hi. I just noticed that there are some weird items (not properties) that start with P (instead of Q), for instance when searching for items that have a conversion to standard unit (P2442):

select distinct * where {
  ?item wdt:P2442 [] .
  filter(contains(str(?item), "P"))

Try it!

When clicking on one of the results it reads "This entity does not exist.". Toni 001 (talk) 14:02, 16 August 2019 (UTC)

Please note Wikidata:Contact_the_development_team#query.wikidata_returns_a_Property_instead_of_an_entity. --Succu (talk) 13:34, 16 August 2019 (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. Matěj Suchánek (talk) 07:08, 19 August 2019 (UTC)

References to non-existing properties

On pages like

there a references to non-existing properties (e.g. P7727, P27411, P102448, P258379, ...). Where do they come from? --M2k~dewiki (talk) 11:39, 17 August 2019 (UTC)

I've seen it a couple of times in film/tv series items in the past few days, where at least IMDb ID (P345) and Netflix ID (P1874) would show such constraint violations, claiming that the value was also already used on a property (where the property number = the Q number of the item) as well as on the item itself. But looking through the items I've edited, I can't find any of these error messages anymore. So maybe it's already fixed? --Kam Solusar (talk) 12:30, 17 August 2019 (UTC)
Hello, thanks for your reply. I dont think this has been fixed. Initially I found the above lists by getting two entries in a query like
SELECT ?item  WHERE { ?item wdt:P3077 "12880" . }

Try it!

which currently returns two entries: wd:P27411 (the non-existing-property ?!?) and wd:Q27411 (the actual item). --M2k~dewiki (talk) 12:59, 17 August 2019 (UTC)

Oh,you're right. After a bit more digging, I've found these two examples: Invader Zim: Enter the Florpus! (Q60741104) (shows constraint violations only for IMDb and Netflix IDs) and Dolemite Is My Name (Q54959170) (constraint violations for all external IDs). Purging the pages doesn't help. --Kam Solusar (talk) 13:42, 17 August 2019 (UTC)
I think it's this bug: phab:T230588. --Kam Solusar (talk) 03:25, 18 August 2019 (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. Matěj Suchánek (talk) 07:08, 19 August 2019 (UTC)

Property proposal : creation of an Identifier for the database RPG Geek

I would like to ask the creation of an Identifier for the database RPG Geek, a subset but independant database of Boardgame Geek. Here are the informations I already collected for this project :

- Database for the property: RPG Geek (Q64848366)

- Represents: RPG Geek (Q64848366)

- Associated item: RPG Geek (Q64848366)

- Data type: External identifier

- Domain : tabletop role-playing games (Q1643932)

- Source website for the property :

- Allowed values : [1-9]\d{0,5}

- Example : Call of Cthulhu (Q1027121) -->

- Formatter URL :$1

- See also : BoardGameGeek designer ID (P3505), Luding game ID (P3528), BoardGameGeek game publisher ID (P6160)

Thanks for your help! Pmartinolli,20190817

You need to use the form on Wikidata:Property proposal/Creative work. --Matěj Suchánek (talk) 13:39, 18 August 2019 (UTC)

--Thanks, it's done. Pmartinolli,20190818

I think that this discussion is resolved and can be archived. If you disagree, don't hesitate to replace this template with your comment. Matěj Suchánek (talk) 07:05, 19 August 2019 (UTC)

Preferred status for population

@Martin Urbanec: is about to remove preferred status from the population property ([19]). He explained his reason here. I would rather see a preferred rank bank in this property. Is it the right way to record the population that way? Many features rely on the preferred status. Now you can see the strange effect in the infoboxes here: voy:de:Heilbronn and here: commons:Category:Heilbronn. If somebody wants to look for the most recent entry he can run through the result set of getAllStatements without being interferred by a possible prefferred rank, right? And besides, the bot PreferentialBot will put the rank back, if I understand it right. -- DerFussi 05:13, 9 August 2019 (UTC)

It is a result of a discussion on cswiki (w:cs:Wikipedie:Pod_lípou_(technika)#Počet_obyvatel_z_Wikidat), as for some municipalities 30+ years old data were marked as preferred. Removal of preferred rank for most recent data seems to be error in script to me. Anyway, PreferentialBot will fix all soon. Not limiting P1082 output is more likely design flaw of the template.--Jklamo (talk) 08:57, 9 August 2019 (UTC)
Maybe the edit summary should explain that (after an onwiki discussion). It's a feature (or known problem) that PreferentialBot doesn't adjust preferred ranks once new data is added. You'd have to ask @Laboramus: explicitly to do it. --- Jura 16:35, 9 August 2019 (UTC)
Edits made by Martin Urbanec break a lot of articles in ruwiki, one of the most integrated project with wikidata. Changes that big must be well announcement before execution, so we can prepare our code or at least expect consequences. --Serhio Magpie (talk) 21:30, 10 August 2019 (UTC)
@Serhio Magpie Could you please explain how that breaks ruwiki infoboxes? I would not expect any Wikipedia would want very dated population numbers to be "preferred rank". Please could you elaborate how ruwiki can expect that? --Vojtěch Dostál (talk) 10:40, 11 August 2019 (UTC)
Query and infoboxes generally expect just one or a few values. Usually there are just 1 or a few values with preferred rank. Unless this is fixed by next Tuesday, I will undo the changes by Martin. --- Jura 12:24, 11 August 2019 (UTC)
Jura already answered. I agree, it's our local design flaw, that we are dependent only on a preferred rank not on a fresh date mark. Because of that infoboxes rendering whole list of population values or throw template argument size error. I notified our lua guys about it, but for this time they did not solve a problem. Thanks. --Serhio Magpie (talk) 01:17, 12 August 2019 (UTC)
Update. Ghuron is fixed infoboxes on our local side, so i don't have any complains now. --Serhio Magpie (talk) 17:00, 13 August 2019 (UTC)
@Serhio Magpie But on second thought, it makes sense _not_ to remove preferred rank from most recent population numbers and I did not realized Martin had not skipped those edits. I am happy you were able to solve the problem anyways --Vojtěch Dostál (talk) 15:40, 14 August 2019 (UTC)

Names in different alphabets

Some of European people has their articles in Japanese or Thai version of Wikipedia saved in their alphabets (names and pseudonyms). Are they transliterated or transcripted? Can I add all diffferent alphabet variations to birth name (P1477) with language and writing system (P282) indication? Eurohunter (talk) 18:17, 12 August 2019 (UTC)

Can only speak for Thai, there names are usually transcribed, but using mai than tha khaat (Q4466856) to keep spelling close to original one. But IMHO, we should keep birth name and similar properties to those in the original script of the subject, only for a bi-national both scripts might be worth adding. Ahoerstemeier (talk) 13:11, 14 August 2019 (UTC)

RfC about model for document description

Dear all,

I started a RfC to improve our model for document description. As I am not all-knowing, I did a proposition of model and I expect some modifications before we can launch a decision phase. The principle of that 2 phases RfC is to avoid to start the decision before having input from a large number of contributors mentioning particular cases which can change the original proposition. This is always frustrating to give its opinion about a proposition which changes several times between the start and the end of the decision phase.

Please have a look at Wikidata:Requests for comment/Improved model for document description. Snipre (talk) 13:31, 14 August 2019 (UTC)

“Wikidata item” link to be moved in the menu column on Wikimedia projects

Hello all,

Currently and since the Wikimedia projects pages and Wikidata items have been connected, the link "Wikidata item" appears on the menu column of the Wikimedia wikis, in the section "Tools".

However, many editors from various projects told us that it would make more sense to have the link in the "In other projects" section, since Wikidata is one of the Wikimedia projects and the Wikidata item link doesn’t really belong to the list of special pages.

This is why we are going to change the position of the link, on August 22nd for all Wikipedias and August 21st for all the other projects. After this date, you will find the "Wikidata item" link in the "In other projects" section.

In some cases, for example on help and meta pages, the section may contain two links to Wikidata, for example on en:Help:Contents where there will be the "Wikidata" link (linking to d:Help:Contents) and the "Wikidata item" link (linking to d:Q914807).

If you want to know more about the previous discussions or mention a bug or an issue, please add a comment to the related Phabricator task.

Thanks for your understanding, Lea Lacroix (WMDE) (talk) 09:45, 8 August 2019 (UTC)

  • Can you hold implementing this until the problem discussed at Wikidata:Contact_the_development_team#phabricator:T66315 is resolved? --- Jura 11:24, 8 August 2019 (UTC)
  • The solution of having a "Wikidata" and a "Wikidata item" link in cases were seems to me like a good solution to the problem. I think it's helpful to list is in the list that way to introduce us as being a proper site in the Wikimedia ecosystem. I can't see of a reason why it would make the enwiki people mad, so I'm in favor. It feels to me that if there are no objections coming up here the time till the 21st is enough time so that further delay isn't needed if the hotkey still works. ChristianKl❫ 14:39, 8 August 2019 (UTC)

Sounds a positive change. Thanks! Nomen ad hoc (talk) 20:26, 14 August 2019 (UTC).

Lizard mess

I was trying to find the page for the concept "lizard" and discovered a bit of a mess. lizard (Q15629245) is supposed to be "lizard" but it includes Sauria (Q2254408) which according to is a "crowned-group of all modern reptiles (including birds)". The page "lizard" links to Lacertilia (Q15879) even though "lizard" is apparently broader than Lacertilia (Q15879). I don't have the time or inclination to sort all this out, but maybe one of you does.... Calliopejen1 (talk) 01:05, 15 August 2019 (UTC)

named as (P1810)

Should I recognise original CAPITALICS from source or nOt? Eurohunter (talk) 09:55, 4 August 2019 (UTC)

  • What are "CAPITALICS"? Capital letters? Italics? Specifically capitals in italics? Something else? Is it a word always written entirely in uppercase? - Jmabel (talk) 16:13, 4 August 2019 (UTC)
    • @Jmabel: Yes just capital letters (schould be included or not). "Is it a word always written entirely in uppercase" - no there is no reason for it but capital letters are used sometimes in the titles of articles or for marketing purposes and question is if we should recognise it (original records) or not. Eurohunter (talk) 20:33, 9 August 2019 (UTC)
      • To my knowledge there is no definite rule regarding capitalization. I personally prefer to "normalize" all-capital letter strings to a more readable form, and it depends on the language what this exactly means. Mind that quite often names are only *displayed* with capitals only, but not stored in that form in the source. —MisterSynergy (talk) 20:40, 9 August 2019 (UTC)
        • @MisterSynergy: How to know how they are stored? What if capital letters are needed in some cases to destinglish one name from another and not in the others in one item? Eurohunter (talk) 17:53, 15 August 2019 (UTC)
          • You know how they are stored if you get them from a web page that is using an HTML small caps style via font-variant (or a now-deprecated tag), but the underlying content uses lowercase or mixed case. - Jmabel (talk) 23:40, 15 August 2019 (UTC)
            • @Eurohunter: I think that P1810 is not used "to destinglish one name from another". It's not a search string, but information about the subject. My interpretion is that the format is free, as already mentioned, but the intention for this property is to be accurate rather than to align with Wikipedia naming standards. If a list is in all caps I would probably normalize it, but if particular names are stylized, such as "A$AP Rocky", I would keep that format in P1810. The main use is probably not to show formatting, but to indicate where he's known as "Rakim" or some other "alias".

Biggest missing feature

I mentioned it few times already but it's the most important missing and basic feature of Wikidata. There need to be option to set up languages to edit labels and desriptions in languages which you want instaed of using babel template. There is also needed option to add more than these 20/30 languages to babel template. Eurohunter (talk) 16:28, 9 August 2019 (UTC)

I urged once too: Wikidata:Project_chat/Archive/2019/04#Let_me_pick_a_language_in_which_I_want_to_add_labels.--Roy17 (talk) 19:10, 9 August 2019 (UTC)
Try Preferences -> Gadgets -> labelLister. Thanks. Mike Peel (talk) 19:17, 9 August 2019 (UTC)
@Mike Peel: Unfortunattelly no. I have it but I want to have listed more than 20/30 languages at once. It impossible to edit when you have to add additional 50 wiki codes to be able to add labels. Why so simple thing can't be resolved for seven years? Eurohunter (talk) 20:27, 9 August 2019 (UTC)
What are you trying to do? What's wrong with clicking on "All entered languages" for your use-case? Why do you feel it's impossible to use the babel template? ChristianKl❫ 11:08, 10 August 2019 (UTC)
@ChristianKl: I think Eurohunter's use case is specifically that he adds labels in all (or most) Latin-script languages at the same time, since this is possible/appropriate for items for which the label doesn't need to be translated. Using "All entered languages" doesn't work for items where this is necessary, since there has to be data for a language to be included in that list (if not indicated in the user's Babel boxes). Jc86035 (talk) 11:18, 10 August 2019 (UTC)
It seems to me like it would be a lot better to have a mul-Latn fallback language then to have a feature that allows you to list the 40 languages that use latin labels for that problem. ChristianKl❫ 19:16, 11 August 2019 (UTC)
User:Jitrixis/nameGuzzler.js AKA VIP's Labels is the solution for this problem. - PKM (talk) 19:00, 15 August 2019 (UTC)

P1957 (Wikisource index page) should be data type "interwiki link" instead of "url"

All valid statements with this property are interwiki links, but the data type is URL. Can this be fixed? Should this be fixed? Beleg Tâl (talk) 19:50, 9 August 2019 (UTC)

  • I don't really use it so I have no idea.
In general, you could propose a new property with datatype "item" and then add the data with that property. Once done, the above could be listed for deletion. --- Jura 12:24, 10 August 2019 (UTC)
It's possible to have a property's data type changed, see the discussion page of Irish Grid Reference (P4091). You need to establish consensus and then a Phabricator task can be created. --SilentSpike (talk) 12:48, 11 August 2019 (UTC)
That's between string and external-id. --- Jura 12:56, 11 August 2019 (UTC)
  • This sounds sensible, but I notice that Commons category (P373) (a similarly-purposed property) is of type 'string'... wouldn't it also be of type 'interwiki link'? Maybe there's a reason for that. Sam Wilson 00:21, 16 August 2019 (UTC)

retrieved (P813) as qualifier

Shouldn't be retrieved (P813) added to id (as qualifier) instead of reference (as part of reference which de facto doesn't exists)? Additionally it is inconsistent in case if we have "real refererence" under id. Eurohunter (talk) 14:27, 10 August 2019 (UTC)

@Eurohunter: I've requested some queries to see how common this is. I don't know how common this is or whether this is accepted practice, and most of the uses I've seen were added more or less entirely by Moebeus. Jc86035 (talk) 11:46, 11 August 2019 (UTC)
@Jc86035 : This is a habit I've had since is I first started editing, it's more or less on autopilot. I was led to believe this was "best practice" and I find it useful to quickly glance to see if a statement is "fresh" or maybe might be outdated, before I investigate further. If it's the wrong thing to do I'm happy to stop the practise, but some kind of timestamp seems useful to me. Moebeus (talk) 12:49, 11 August 2019 (UTC)
  • What is this edit meant to mean? If it's to say that on that point in time the identifier was that, it should use point in time (P585). OTH, if it's an identifier, it shouldn't really be different. --- Jura 12:30, 11 August 2019 (UTC)
    • @Jura1: Wouldn't it be missleading? In Wikipedia we use data as publication/release date and access date to determine when site was avaiable. Eurohunter (talk) 13:22, 11 August 2019 (UTC)
      • @Jura1: So why "access date" can by only used with references? From my perspective "access date" fits perfectly everywhere. Eurohunter (talk) 17:28, 15 August 2019 (UTC)

Help with given names and surnames

Hi! I was trying to add the given name and surname for the Brazilian actress Maytê Piragibe, but I don't know how.

Her full name is Maytê Bernardes Rodrigues Piragibe.

  • Maytê = given name
  • Bernardes = mother surname (1)
  • Rodrigues = mother surname (2)
  • Piragibe = father surname

Minerva97 (talk) 17:43, 12 August 2019 (UTC)

I believe Scandinavian middle family name (P6978) was designed for this, though somebody changed the English label to be specific to the "Scandinavian" case. ArthurPSmith (talk) 17:48, 12 August 2019 (UTC)
OK! Thanks for the help. I already added the given name. I'll try the surnames now. Thanks. Minerva97 (talk) 18:08, 12 August 2019 (UTC)
  • There is also second family name in Spanish name (P1950), if applicable, although I don't know how different Portuguese and Spanish naming customs are. -Animalparty (talk) 04:45, 13 August 2019 (UTC)
    • In Spanish custom, the mother's family name always comes after the father's, though, which seems not to be the case here. - Jmabel (talk) 06:14, 13 August 2019 (UTC)
Portuguese naming is opposite of Spanish: given name/mother's lname/father's lname (Spanish is given name/father's lname/mother's lname). So while Spanish people might omit the second last name, the Portuguese will typically omit the first last name. Scandinavian middle names are typically the same as the Portuguese "mother's last name".. 🤷‍♂️ Moebeus (talk) 23:33, 14 August 2019 (UTC)
So it sounds more like Scandinavian middle family name (P6978) than second family name in Spanish name (P1950); either the latter should be renamed (in English) to not be specific to Scandinavia, or we should have another property just like Scandinavian middle family name (P6978) for Portuguese usage. - Jmabel (talk) 23:44, 15 August 2019 (UTC)

What to do when two items share the same identifier

Hello, I was wondering what should be done when two items share the same identifier (with the property having a distinct value constraint)? This relates to UEFA player ID (P2276) as part of UEFA's player database. UEFA currently have a mistake with the following people who have the same name:

I would think both items should have the identifier. Is there a way this can be marked so the statements are not listed as having an 'issue'? Otherwise the identifier will continue to be removed from either one of the items. Thanks, S.A. Julio (talk) 05:19, 15 August 2019 (UTC)

An adhoc way of marking it would be to add a together with (P1706) qualifier and link the other item. Ghouston (talk) 06:51, 15 August 2019 (UTC)
What about identifier shared with (P4070)? It seems a little more accurate. Nomen ad hoc (talk) 06:59, 15 August 2019 (UTC).
Ideal. Ghouston (talk) 07:18, 15 August 2019 (UTC)
Great thanks, just what I was looking for! S.A. Julio (talk) 08:33, 15 August 2019 (UTC)
When I'm looking at the list I'm not sure that a single value constraint is appropriate. It seems to me that Ermin Bičakčić (Q75993) has two values by virtue of being in the tournament both in 2016 and 2020 and that looks like it's intentional. ChristianKl❫ 11:27, 15 August 2019 (UTC)
@ChristianKl: I don't think it is intentional, UEFA use the same player ID for people in their match reports throughout their careers, except in these few instances. With the Ermin Bičakčić (Q75993) example, you can see the top bar of UEFA's website appears a bit different in the two links. This is because UEFA keeps a "cache" of player profiles to display. Occasionally the UEFA ID of a player will be changed, as in this case. While both links appear to be valid, the reason is a bit clearer looking at UEFA's competition database: the ID 250025190 is now invalid, and has been replaced by 250008197. However, because UEFA keeps this "cache" page, the profile of the invalid ID still is accessible on their site. S.A. Julio (talk) 23:06, 15 August 2019 (UTC)

Source item showcase

Lately I've been working to adequately source some of my statements because Wikidata as a whole is pretty lacking in that regard. However, there are a few hurdles I've encountered in the creation of source items which I'm only just now starting to understand.

I think to help others like myself and even just as a reference for the more experienced, it could be very valuable to establish some showcase source items (à la Wikidata:Showcase_items). Ideally, an item for each imaginable type of source (i.e. a book with multiple edition items; a periodical with multiple volume, issue and even article items; an e-book; an audio book, etc.) and an example of their use as a source. Perhaps it's best to create a separate area for this purpose for discoverability and since there's a slightly different priority to the item showcase.

Would anyone else be interested in something like this? We could set up a WikiProject and establish some goals. Admittedly I still have a few questions regarding source items, but part of the motivation here is to have those answered with future reference available. --SilentSpike (talk) 13:59, 15 August 2019 (UTC)

  •   Support This sounds like a good idea to me! ArthurPSmith (talk) 15:10, 15 August 2019 (UTC)
  •   Support I’d be happy to help with this. - PKM (talk) 18:01, 15 August 2019 (UTC)
  •   Oppose A Showcase item should have references for all important statements. --Succu (talk) 18:06, 15 August 2019 (UTC)
    • I'm not proposing we treat these as showcase items - perhaps I should've avoided that terminology as I just meant we could have a page/area with similar function but specifically to demonstrate how to structure and use different types of sources. --SilentSpike (talk) 19:12, 15 August 2019 (UTC)
      • So do you have something in mind like Species Plantarum (Q849308)? The first edition (=Species Plantarum. 1st Edition (Q21856050)) is at least heavly used. --Succu (talk) 20:34, 15 August 2019 (UTC)
        • Yeah that is a good example of a book source item because it shows the correct work + edition structure with multiple editions. Basically I'd like to have some discover-able resource of different types of source items and their use that people can refer to if they need to create new such items to add as sources to statements. See the question I link to in my original post for an example of a situation where I have been unable to identify the correct source structuring in practise --SilentSpike (talk) 21:13, 15 August 2019 (UTC)
  •   Support; I assume that it would be useful for the newbies (I remember that for me, it was hard to learn how to add a citation). Nomen ad hoc (talk) 18:47, 15 August 2019 (UTC).
  •   Oppose I would prefer to improve Help:sources first before starting to spread the information over several pages. People don't have the time to read to read several pages to find the information they are looking for and for me showcases can't provide a rule: there are always plenty of special cases and a showcase can't provide a good overview of what is important to respect. Snipre (talk) 20:04, 15 August 2019 (UTC)
    • I agree, especially when it comes to Help:Sources/Items not needing sources. --Succu (talk) 20:27, 15 August 2019 (UTC)
    • You raise a good point and updating the help pages could also be a priority for a WikiProject in this vein. I personally find it easy to learn from examples which is why I think it's valuable to have items for reference somewhere too (no need to do only one or the other). If we create a WikiProject we could even just use that as the repository of example items and include a link to the project from the help pages with the suggestion that sourcing questions be directed there. Any questions that do come up can be clarified on the help page and we can make sure an example is present that covers the situation (like testing in a development environment, you encounter an issue, you write a test for it so that future instances are caught). --SilentSpike (talk) 21:13, 15 August 2019 (UTC)
      @SilentSpike: The definition of a data model is often an iterative process with improvement of the initial model with the study of special cases which can't be covered by the inital formulation of the model. And this is why showcase can't be a good way to represent all particular details requiring additional properties/qualifiers. Often showcases represents the simple case in an item category (perhaps 80% of all items) and then to cover the remaining 20%, you need 2-4 additional showcases. So at the end you don't have 1 showcase by category but 3-5 showcases. And without any explanation describing what are the differences between the showcases: people have to analyze the difference shocases to understand what is the characteristics of each item and then establish themselves the general rule leading to the different models.
      Finally, do we really need a new Wikiproject for that ? Currently, when people have problem with describing sources they use the talk's page of Help:Sources. From my experience of Wikiproject, very few persons are involved in Wikiproject and it is difficult for someone from outside WD to find the correct Wikiproject where to present the problem. Most problems are solved directly in the Project Chat but then we lose the solution when the discussion is archived.
      My recommandation is to use Help:Sources as place for discussion and data model presentation. Snipre (talk) 21:54, 15 August 2019 (UTC)

Add exact time of death

Zhao Tieqiao (Q10311642) died at 21:00 Shanghai local time (which should be 13:00 UTC), 1930-07-24. (referece: ) Could you please help key in this?--Roy17 (talk) 13:22, 10 August 2019 (UTC)

  Done I let you insert the reference. --- Jura 13:38, 10 August 2019 (UTC)
Special:ListDatatypes says time is formatted as +2013-01-01T00:00:00Z blah blah blah, but it's actually not? such that users have to use qualifers to circumvent?
And reference could only use reference URL? Is it possible to add a textual reference like what the template:cite web takes on wikipedias (in this case "申報" 1930-07-25...)?--Roy17 (talk) 00:06, 12 August 2019 (UTC)
For instance, date of birth, place of birth and citizenship of Tao Ho (Q4833229) could be found in page 362 of the book 9781349041848. How to present this reference properly?--Roy17 (talk) 17:55, 16 August 2019 (UTC)
Ypu should have a look at Help:Sources#Books. Best, Nomen ad hoc (talk) 18:07, 16 August 2019 (UTC).

Work in progress : a gadget to visualize class trees and subclass-of paths between two classes

Motivation : there is often problems with parent classes of some items − see for example the section #How a human becomes faultly an artificial entity above. It’s then useful to visualise the superclasses of a class to see which other classes their instances are also instances of.

So I created a gadget that allows to embed the graph visualisation of WDQS in a Wikidata item page. If a weird superclass is found


The few next steps :

  • add translations in other languages
      Done translation possible, translated if french 15:07, 16 August 2019 (UTC)
  • add a feature, for items who are instances of some class, to show to the user some of the other class he his instance of to spot some weird stuff, to help finding bugs in the class tree. For example if « Albert Einstein is an Abstract object » is shown this should raise suspicions. Maybe add a button to leave a message with the oddity on some page for other people to try to correct the bug.
      Done 15:07, 16 August 2019 (UTC)
    also   Done, the ability to report a « bug » in the class tree, if it occurs something weird, such as human becoming an instance of law or something like that. 15:14, 16 August 2019 (UTC)
  • TODO: Other idea : show the metaclass (for example if an item is an instance of a class that is itself an instance of « car model », show that its class is a car model. 15:07, 16 August 2019 (UTC)
  • See if there is other ideas/suggestions/bugs after this post :)

author  TomT0m / talk page 15:03, 10 August 2019 (UTC)

well this proves what I knew - “clothing” upper tree is a mess. (“Headdress” is an “economic sector”?) - PKM (talk) 22:30, 10 August 2019 (UTC)
Thanks so much! I've been unrolling the instance of (P31) hierarchy mainly using SPARQL but this is so much more convenient. :) ElanHR (talk) 06:08, 11 August 2019 (UTC)
@TomT0m: You can also do this using {{Item documentation}} on talk pages, if I'm interpreting the purpose of the user script correctly. Jc86035 (talk) 11:48, 11 August 2019 (UTC)
@Jc86035: I know I'm a contributor to this template :) I added the template to the talk page quite a few times. But it's less flexible, we need to go to the talk page, add the template if it's missing, and there is no dynamic abilities such as showing the subclass chain between the class and one of it's parent (useful if the class tree is missing). author  TomT0m / talk page 12:02, 11 August 2019 (UTC)

Making progresses

I made progresses and implemented something for showing the classes an item is an instance of on a sandbox for this gadget User:TomT0m/classification/sandbox.js. It’s now possible to report weird instances to WikiProject Ontology by clicking a link of the gadget, resulting in [22] this kind of messages. (the weirdness comes from the fact that this item is labelled « heterotroph » in english and in french it’s more like « heterotrophia », which is a process)

author  TomT0m / talk page 17:10, 12 August 2019 (UTC)

Researcher and scientific articles

Trygve Ottersen (Q66424846) is a researcher and have published notable work (P800) A Typology of Scientific Advisory Committees (Q56797691) as author (P50) and other works as contributor(s) to the creative work or subject (P767). Is it wanted that all (about 40 ) Works and "co-Works" shall be listed as notable work (P800) for this person? Pmt (talk) 19:18, 15 August 2019 (UTC)

No IMHO; only the notable ones. Nomen ad hoc (talk) 19:45, 15 August 2019 (UTC).
+1 to only listing notable ones. If someone wants all works by a person they can query the opposite direction and look for all items that have this person as author (P50). ElanHR (talk) 20:33, 15 August 2019 (UTC)
notable work (P800) includes some POV and needs a good reference. --Succu (talk) 20:39, 15 August 2019 (UTC)
They can be shown through "Derived statements" too ;). Nomen ad hoc (talk) 20:44, 15 August 2019 (UTC).
@Nomen ad hoc: Do you refer to inferred from (P3452)? --Succu (talk) 21:11, 15 August 2019 (UTC)
No; I speak about the "Relateditems" tool, that you can activate on your preferences, and which displays reverse properties at the bottom of any item. Nomen ad hoc (talk) 21:16, 15 August 2019 (UTC).
OK, you have a conflict. How dose the gadged resovlse this? --Succu (talk) 21:35, 15 August 2019 (UTC)
I have no conflict... I only reverted a claim that seems inaccurate (notability of Q56797691 is not proven). Nomen ad hoc (talk) 21:41, 15 August 2019 (UTC).
@Nomen ad hoc:, @Succu: and @ElanHR:. I do think the same as you they should not be listed for the researcher, and have had an answer. @ElanHR: Mainly such scientific articles do not have author (P50) , only author name string (P2093) so there is no "connection" to Trygve Ottersen (Q66424846) as author (P50) just author name string (P2093) as Trygve Ottersen. @Nomen ad hoc: You are reverting my edit by asking me, to what extent exactly it is notable? together With your comment No IMHO, only the notable ones. I can only answer to what extent exactly it is NOT notable? Being a scientific article, published in a well known scientific Magazine. Probably evaluated by other fellows. Recogniced for a publication by a Publishing commity. The main thing is shall all the scientific articles be listed as notable work (P800) for a researcher. Probably not. But there must be a better argument for not listen the article than just some dull not notable and at the same time the articles must be connected or related to the Author in some way. And I can not see that author name string (P2093) will cover this as long as it is a text string. The same for short name (P1813). Queries may be ok, but for daily use thay may be to "Heavy" I am just asking Pmt (talk) 23:57, 15 August 2019 (UTC)
Hi. I've replaced all claims of Ottersen as author name string (P2093) by author (P50) ones. Pmt, just publish a work in a well-know journal doesn't make it notable; it should have been cited ("evaluated by other fellows", in your words) a certain number of times, for instance, or, say, have received an award. So the burden of the proof of its "notability" lies with you. For me, the better solution to your concerns is, for each author, to replace all claims of P2093 by P50, as done in this case. Cheers, Nomen ad hoc (talk) 07:12, 16 August 2019 (UTC).
@Nomen ad hoc: Thank you for your answer. And thanks for changing from author name string (P2093) to author (P50). Is there any tool for doing this or is it done manually? As you say the burden will be on the person claiming the article being notable, and have also given some basic criterias such as cited for a certain number of times. Pmt (talk) 19:29, 16 August 2019 (UTC)
My pleasure. But sorry, I don't know any tool for this... Nomen ad hoc (talk) 19:34, 16 August 2019 (UTC).

──────────────────────────────────────────────────────────────────────────────────────────────────── Author Disambiguator is the tool you want. - PKM (talk) 19:45, 16 August 2019 (UTC)


Is Property:P1281 giving real information? When I click on the entry for Westbrookville, New York it shows the images that are clearly not from Westbrokkville, New York and then when I click on the entry for New York City given as an example on the property page, I see the same three images. Clearly the Flickr property is not really used in any images, and the data they present is deceptive filler, that can lead people astray. --RAN (talk) 14:33, 16 August 2019 (UTC)

The problem is clearly with how Flickr handle there being no actual material for the place in question. Circeus (talk) 19:58, 16 August 2019 (UTC)

how to merge 2 items

These two items represent the same thing. How do I merge them. I filed request for deletion of one of them but it did not work.

merge: Q66066995 and Atherosclerosis Risk in Communities (Q20707753) EncycloABC (talk) 02:46, 22 August 2019 (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. Matěj Suchánek (talk) 16:51, 22 August 2019 (UTC)


List of ministers.

Hello. Should I add all the ministers in Ministry of Foreign Affairs (Cyprus) (Q3315347) using chairperson (P488) (like United States Department of State (Q789915)) or should I add them to Cyprus Minister of the Interior (Q25309937) with officeholder (P1308)? (Of course with qualifiers start time (P580) and end time (P582) for both cases). Xaris333 (talk) 04:32, 17 August 2019 (UTC)

Well, a minister isn't a "chairperson" of its ministry, is it? So I would consider the second option. Nomen ad hoc (talk) 06:52, 17 August 2019 (UTC).
So the list in United States Department of State (Q789915) is wrong and we should move it to United States Secretary of State (Q14213)? Xaris333 (talk) 07:19, 17 August 2019 (UTC)
I think it is more likely that chairperson (P488) has a bad English-language label and should be "presiding member". - Jmabel (talk) 15:35, 17 August 2019 (UTC)
Changing property labels is easily the most fractious thing I've ever seen attempted on Wikidata. Good luck with that. Circeus (talk) 20:34, 17 August 2019 (UTC)
:D. Nomen ad hoc (talk) 20:53, 17 August 2019 (UTC).

Image number of values

Hey, why there is no warning, when somebody add more images into this value. I have seen it elsewhere and recently I was warned that this property should have just one value. --Juandev (talk) 16:24, 15 August 2019 (UTC)

It doesn't have single value constraint (Q19474404). --Matěj Suchánek (talk) 13:58, 18 August 2019 (UTC)

Qualifiers for identifiers

Which qualifiers should be included in identifiers? Schould they be inluded in references or schould they have references independly anyway (expect cases reference is needed to confirm that certain id is official)? Eurohunter (talk) 17:56, 15 August 2019 (UTC)

Qualifiers are usually used to differenciate several values, and as identifier are by principle unique, no qualifier should be used with identifier. Snipre (talk) 21:59, 15 August 2019 (UTC)
There are some exceptions to that rule, identifier shared with comes to mind? Moebeus (talk)
@Snipre: Shouldn't it have atleast named as (P1810) (any expections?) or identifier shared with (P4070) in few cases (very rare I think)? Eurohunter (talk) 23:29, 15 August 2019 (UTC)
I don't think there are any that should, but there are plenty that can. Some examples:
Q7169723#P3615 has a good example of how qualifiers on an identifier can be appropriate - it uses subject has role (P2868) to clarify that it refers specifically to a certain type of administrative area, and has start/end dates because the admin area only existed in that period. Start/end dates can also be used where you have two identifiers that apply at different times (say an official ID scheme changed and they issued new ones for every item, or the external source breaks things down more granularly than we do).
named as (P1810) means that you can say what name the identifier uses, which might be very useful if, say, you have an author who writes under two different pseudonyms, and there are identifiers for each of them separately. Andrew Gray (talk) 19:57, 16 August 2019 (UTC)
@Andrew Gray: When I'm adding stated in (P248) as qualifier it shows me "The source property should not be used in this location (as a qualifier). The only one valid for this property is as the source". Eurohunter (talk) 21:07, 16 August 2019 (UTC)
I think this is because that particular property should only be used in references - it doesn't matter if it's for a normal property or an identifier. Andrew Gray (talk) 21:33, 16 August 2019 (UTC)
@Andrew Gray: Most of identifiers can't have references because it would be nonsense and would conflict with "real references" under identifiers that requires it. So what to do? Eurohunter (talk) 08:55, 17 August 2019 (UTC)
@Eurohunter: I'm not quite sure what you're trying to do here - is there a more appropriate qualifier you could use than stated in (P248)? If it's only for references, and you aren't using references, it's probably not the right one. Andrew Gray (talk) 09:28, 17 August 2019 (UTC)
@Andrew Gray: So why is bot adding it to references if it's not real reference? Eurohunter (talk) 10:54, 17 August 2019 (UTC)
I don't understand what you're asking. Could you show me an example? Andrew Gray (talk) 13:05, 17 August 2019 (UTC)
@Andrew Gray: [23] in reverse. Eurohunter (talk) 13:24, 17 August 2019 (UTC)
@Eurohunter: The problem here is that stated in (P248) should only be used as a reference, not as a qualifier. So the bot moves it into the references section whenever it's added as a qualifier. In this case, though, I don't think it makes sense to add as a qualifier or as a reference - what is it you were trying to say by adding this qualifier? Andrew Gray (talk) 16:46, 17 August 2019 (UTC)
@Andrew Gray: So what is the way to describe "id provider"? Eurohunter (talk) 19:04, 17 August 2019 (UTC)
@Eurohunter: In this case, I don't think we need to. "VIAF ID provided by VIAF" doesn't seem like something we need to state at all? But if we were saying "this ID was found in a different database", then a reference with stated in (P248) would be sensible. Andrew Gray (talk) 09:39, 18 August 2019 (UTC)

Participation in an event

If a person is set as a participant in an event entry I do not see that the event shows up as participated in from the person entry. I would think that logically the relationship should be established between the two records, but I am new to Wikidata. So for now, I should assume I always have to create a participant property in both the person and event records?

Thanks for any guidance...  – The preceding unsigned comment was added by Tdefarlo (talk • contribs).

IMHO participant of (P1344) in the participant's item would suffice.
Best, Nomen ad hoc (talk) 12:40, 18 August 2019 (UTC).

Wikidata weekly summary #378

Linking WD item to a section of a WP page

I am trying to link the Wikidata item to the en.WP page section,_Auckland#Isa_Outhwaite but I keep getting an error message. Any advice on how to do this? TIA. MurielMary (talk) 09:09, 17 August 2019 (UTC)

@MurielMary: Any WP page can only be linked to one Wikidata item - in this case, the family page is already linked to Outhwaite Family (Q7112076) so it won't allow a second link. The only way to link to a section of a page is to create a redirect on WP pointing to the section, and link the redirect to the item. Andrew Gray (talk) 09:23, 17 August 2019 (UTC)
Thanks, that's very helpful! MurielMary (talk) 10:45, 17 August 2019 (UTC)
@MurielMary: Do a pagelink to the redirect en:Isa Outhwaite. That will get you what you want. - Jmabel (talk) 15:37, 17 August 2019 (UTC)
You can use described at URL (P973) as in Q6895759#P973. Visite fortuitement prolongée (talk) 18:59, 18 August 2019 (UTC)


It will be not the first time when I'm asking about it. How to add urls to British or German iTunes? It links only to American iTunes. Eurohunter (talk) 20:52, 15 August 2019 (UTC)

The property doesn't exist for adding urls in the first place. It's an external ID property and British or German iTunes doesn't use a different ID. ChristianKl❫ 07:05, 19 August 2019 (UTC)

Where are items for dpcm and ppcm

We do have dots per inch (Q305896) and pixels per inch (Q7398951), but at least in Mainland China, we also do really have "Dots per centimetre (Q174728)" and "Pixels per centimetre (Q174728)", please tell me that which already-existing items are used for both, or can someone please create both? --Liuxinyu970226 (talk) 23:11, 15 August 2019 (UTC)

Oude beroepen op Wikidata?

I found this website: . Seems to me very nice data to incorporate on Wikidata. I do not have the time to do something with this, but perhaps there is someone else who has the time to make contact with the (elderly) author to ask permision. Another site of him: --> look at "Oude termen". Efidder (talk) 10:16, 19 August 2019 (UTC)

P803 as a qualifier?


shouldn't be professorship (P803) a qualifier of educated at (P69) employer (P108)? It is obviously intrinsically related to the place of work.

Nomen ad hoc (talk) 16:00, 16 August 2019 (UTC).

@Nomen ad hoc: educated at (P69) is not for the place where you teach, but where you studies. author  TomT0m / talk page 16:09, 16 August 2019 (UTC)
My bad, I meant P108. Nomen ad hoc (talk) 16:10, 16 August 2019 (UTC).
I woukd say no for two reasons. (1) According to the property definition, this should have a value of a named professorship like Regius Professor of History (Q7309487) which by its definition will be associated with a university. (2) This property should be qualified with start time, end time, replaces, replaced by - you can’t qualify a qualifier. That said, this seems to be a widely misused property. Values like “professor” and “2001” are incorrect. Perhaps the definition needs to be refined. And we need to add a lot of additional named professorships. - PKM (talk) 19:57, 16 August 2019 (UTC)
You're right. Yes, it really needs to be refined (what about chairs? And other academic positions?)... Don't hesitate, if you're motivated. (Furthermore, I've opened a — rather unsuccessful — property proposal for all kinds of academic positions/ranks.) Nomen ad hoc (talk) 20:00, 16 August 2019 (UTC).
I’m not really motivated to take this on. Sorry. - PKM (talk) 02:14, 20 August 2019 (UTC)

And what about academic degree (P512), according to you, PKM? Nomen ad hoc (talk) 18:55, 19 August 2019 (UTC).

@Nomen ad hoc: I use property educated at (P69) with qualifier academic degree (P512), which seems to be the most common solution (see discussion at Property talk:P512). I think a case could be made either way. - PKM (talk) 02:14, 20 August 2019 (UTC)
Yes, I think we should set it as a qualifier. Nomen ad hoc (talk) 06:35, 20 August 2019 (UTC).

Adding a label AND description using QuickStatements

Help:QuickStatements#Running QuickStatements through URL

Welp. It doesn't seem to work. %0A does nothing. I can add a label: don't run this or a description: Don't run this, but not both at once. Why not? Alexis Jazz please ping me if you reply 06:18, 20 August 2019 (UTC)

Two different infoboxes displaying same image in the same article

I recently added an image into an item (Q1088200). I saw two infoboxes of the Russian article ru:You Shook Me having the same image, but the infoboxes have different info. One is about Muddy Waters version; other the Led Zeppelin version. The Led Zeppelin image shouldn't have displayed a different release, so what can be done about it? George Ho (talk) 17:39, 17 August 2019 (UTC)

Hello George! Thank You for Your message. I fixed it in Russian Wikipedia, is it all right now? --Ksc~ruwiki (talk) 18:31, 17 August 2019 (UTC)
I see the edit now. Thanks, Ksc~ruwiki! Nonetheless, I would appreciate a hidden note like this: <!-- example -->. That way, most editors can understand why the edit is to be made. George Ho (talk) 18:54, 17 August 2019 (UTC)
That's a solution used only in Russian Wikipedia (I am not sure about other Wikipedias). The mechanism described here (in Russian: "Чтобы отключить свойство Викиданных, ничего не добавляя взамен, используйте в статье значение -. " --Ksc~ruwiki (talk) 19:27, 17 August 2019 (UTC)
  • In plwiki the word no (Q1814990) is used as argument for image file name to disable the image in those infoboxes, which uses some image property from Wikidata such as image (P18). Paweł Ziemian (talk) 17:44, 20 August 2019 (UTC)
  • Ideally the two infoboxes would have two different items associated with them. --- Jura 17:57, 20 August 2019 (UTC)

Obtain a label onwiki in a specific language

I simply want, for example, the French label for Q34679 to be returned. (sable) If there is no French label, I want nothing. Reading w:en:Module:Wd, my impression is that this is not possible without changing that module to return what I want instead of the language of the wiki. (and I don't speak Lua, so that's that) How do I do this? Alexis Jazz please ping me if you reply 23:40, 18 August 2019 (UTC)

@Alexis Jazz: {{#invoke:WikidataIB|getLabel|Q34679|lang=fr}} should do what you want - except the version of Module:WikidataIB here is a bit out of date, so you have to use it on enwp/commons/elsewhere with the latest version available. Thanks. Mike Peel (talk) 06:50, 19 August 2019 (UTC)
@Mike Peel: Thanks! I need it on Dutch Wiktionary, so I'll just import any required modules and templates. Alexis Jazz please ping me if you reply 08:33, 19 August 2019 (UTC)
@Mike Peel: The same thing doesn't work for descriptions, is that correct? They keep falling back to English. I needed this for wikt:nl:Sjabloon:vertalingen#Voorbeelden. Alexis Jazz please ping me if you reply 07:38, 20 August 2019 (UTC)
I think this needs @RexxS: to add lang as a parameter for the getDescription function. Thanks. Mike Peel (talk) 09:52, 20 August 2019 (UTC)
@Alexis Jazz, Mike Peel: It's not quite that simple, Mike. If you want the description in a particular language, you have to load the entire Wikidata entry, then look through the description table to find that language. Rather than have to import all of a modified WikidataIB and its dependencies into the Dutch Wiktionary, I've simply created nl:wikt:Module:Wdnl with the two requested functions. On its talk page, nl:wikt:Module talk:Wdnl, I've make some tests as a demo of how to call the functions and to show that they work in the way that I think was intended. if that doesn't do the job, then please let me know and I'll try again for you. Cheers --RexxS (talk) 16:09, 20 August 2019 (UTC)
@RexxS: Ah, sorry, I thought it would be straightforward. Is that because there's no description version of mw.wikibase.getLabelByLang? Thanks. Mike Peel (talk) 20:47, 20 August 2019 (UTC)
@Mike Peel: Principally that, but also determining whether the fallback languages are used or not. If I add the language parameter to WikidataIB|getDescription (which I can do), what do I do when someone else wants to be able to specify the language and have it fallback if that language has no description (which is how I treat getValue returns that are monolingual text)? I have all the code, so I suppose I could add a parameter |fallback=true or false for the call, and then the same for the labels. But it looks like work, so I think I'll wait for somebody to ask for the features. --RexxS (talk) 21:32, 20 August 2019 (UTC)
@RexxS: OK, I've submitted a phab ticket for this at phab:T230839. Thanks. Mike Peel (talk) 21:45, 20 August 2019 (UTC)

Change interwiki in Q116933

Hello, I'm kinda new to Wikidata, so, sorry if I'm asking in the wrong place. Can anyone edit the pt.wikipedia link in Q116933? The page has been moved from "VK (rede social)" to "VK" (proof). That means the item is linking to a redirect instead of the actual article. I can't do it myself because the item is protected. Thank you. GhostP. (talk) 22:51, 20 August 2019 (UTC)

@GhostP.: Looks like it's been taken care of by a bot. ArthurPSmith (talk) 00:29, 21 August 2019 (UTC)

Close to me: view Wikidata items on a map

Hey everyone, i made a new tool that lets you see Wikidata items with a location based on your current location. It's basically a wrapper around the 'map' view on the query service, but without the need to write a complex SPARQL query ;) Check out Close to me. Husky (talk) 08:37, 19 August 2019 (UTC)

Wow! Very interesting. Thanks! Nomen ad hoc (talk) 08:40, 19 August 2019 (UTC).
Honestly, there should be a well-maintained app that would have one's phone buzz anytime one is next to any of those spots and that would then have one take a picture that would immediately be uploaded to Commons and added as a relevant image (P18) here. Using the built-in clock and calendar of the phone (and possibly image recognition technology) could bring that app to use nighttime view (P3451) or winter view (P5252) instead. Thierry Caro (talk) 09:38, 19 August 2019 (UTC)
Magnus pointed me to WikiShootMe! which you can use to upload pictures. I would hope that in the future WD items could also be added to the already existing 'Places' tab on the Wikipedia app. Husky (talk) 10:55, 19 August 2019 (UTC)
There's also the Commons app, which also shows the nearby Wikidata items. But not so smart - it includes items which are no longer existing, e.g. around my home it shows elements of the city wall torn down more than 100 years ago. My suggestion of filtering items with dissolved, abolished or demolished (P576) hasn't made it into the app yet... Ahoerstemeier (talk) 08:25, 20 August 2019 (UTC)
It's too bad the Commons app doesn't work on iOS :(. I could imagine showing items that have dissolved, abolished or demolished (P576) might be interesting as well, a picture of the place where those city walls used to be might of use to an article as well... Husky (talk) 08:26, 21 August 2019 (UTC)
As a kind of thank you I uploaded the picture of the inn where I stay tonight. Edoderoo (talk) 20:39, 19 August 2019 (UTC)
\o/ Husky (talk) 08:26, 21 August 2019 (UTC)

fix redirects

There are some 100 items who should refer to Q1713320. I am waiting for the job that makes that fix.. How often does that job run? Thanks, GerardM (talk) 15:24, 19 August 2019 (UTC)

My bot will pick it up on the first Saturday at least week (7 days) after redirecting. --Matěj Suchánek (talk) 06:57, 21 August 2019 (UTC)
Thank you :) GerardM (talk) 07:16, 21 August 2019 (UTC)

New termbox on mobile is now live

Screenshot of the mobile view of the termbox (taken from the Earth item on beta)

Hello all,

For the last few months, the Wikidata team has been working on improving the termbox interface on mobile. The termbox is this area on Items where you can read or edit labels, descriptions and aliases in various languages. Our goal is to provide a nicer experience while reading and editing this area from the mobile web pages.

Thanks to the combined work of various teams at WMDE and WMF, to our UX researchers and the editors who gave feedback on the prototypes, the first version of the feature is now deployed on

What is included in this new termbox interface?
  • You can browse labels, descriptions and aliases on a new and more mobile-friendly interface, for all existing languages
  • You can edit existing labels, descriptions and aliases directly from the interface of the Item
  • The behaviour of the mobile termbox is very similar to the current termbox on desktop: preferred languages are sorted based on the Universal Language Selector extension, that is taking information from your Babel userbox (if you’re logged in)
  • The edits made through the interface will have the “mobile edit” tag, so you can track them and watch out for vandalism
What is not included yet, but may be in the future?
  • A dedicated termbox-mobile tag in the edit summary, to track edits more precisely
  • Adding labels, descriptions or aliases in languages that don’t contain anything for now (we’re working on it, in the meantime, you can do it with Special:SetLabelDescriptionAliases)
  • Various interface enhancements (floating buttons, colors, icons, text labels)
What was not included in the scope of this project
  • Ability to edit something else than the Item termbox (eg. statements, Lexemes, other entity types)
  • Any change on the desktop interface

How can you test it and give feedback?

The feature is now live on, which means you can access any Item from a mobile device and check it out. You can also see it from desktop by clicking on the "mobile view" link at the very bottom of a page, or adding &mobileaction=toggle_view_mobile at the end of the URL. Example: Q666

Please note that some pages are cached - if you don’t see the new termbox appearing on an Item, it will appear next time that someone edits this item.

If you have any suggestions or would like to share feedback with us, you can write on the talk page of the termbox project.

If you encounter any bug or serious technical issues, you can also create a task on Phabricator, using the tag Wikidata-Termbox.

Cheers, Lea Lacroix (WMDE) (talk) 09:50, 21 August 2019 (UTC)

Edit: the newly deployed feature seems to have created some errors on Special:NewItem on mobile. We're working on this and the issue should be fixed in the next 24 hours. Sorry for the inconvenience. Lea Lacroix (WMDE) (talk) 16:58, 21 August 2019 (UTC)

Modelling the wiktionary …

The wiktionary class tree is a mess. I think that this is partly because the item is actually merging different meaning : the community, the dictionary itself, the mediawiki website … I propose to sort that out by creating an item for the wiktionary as intellectual work (Q15621286)     and the wiktionary project as a Wikimedia project (Q14827288)    . I also created an item for the website wiktionary website (Q66542603)    . Some wikipedia defines the wiktionary as a dictionary, some other as the project, which make the current item a little blurry however.

There is relationships between all those items of course, I can think of

… (I don’t know if we have properties for everything)

The advantages are that the work item can clearly be subtyped as a kind of dictionary, and which, while the wiki itself is classified as a wiki … author  TomT0m / talk page 10:11, 18 August 2019 (UTC)

  Support This sounds like a reasonable proposal to me. I see this kind of issue a lot where editors have combined multiple distinct concepts onto the same item. --SilentSpike (talk) 12:17, 18 August 2019 (UTC)
  • It's unclear what it's going to fix and if isn't going to make a mess of every other WMF site entry. --- Jura 17:58, 20 August 2019 (UTC)
    • @Jura1: We usually differenciate the work of some author and the authors itself, the editor, and we also discriminate editions of some work. Comparatively … why should we merge such different concept as a work, a community of authors, in the same items, for a few projects ? author  TomT0m / talk page 08:21, 21 August 2019 (UTC)
  • Lots of things are like this. The United States is an area, an organization, a set of people, a set of laws, etc. Splitting things up like that doesn't help. Wiktionary is a community, a dictionary, a website, and a project. --Yair rand (talk) 07:53, 21 August 2019 (UTC)
I think it's worth considering that all the wikimedia project items should be modelled the same way. Although there is a Wikipedia community (Q876040) (and there should reasonably be equivalents for the other projects), there is no separation between wikipedia as content as wikipedia as a website. They are, by and large, one and the same, and edited not through any purported "wikimedia website", but rather through MediaWiki (Q83). Anyway, before you start unleashing this on other items, how about you start with wikipedia items, and remember that there is not so much a wiktionary website (Q66542603) as there are language editions (i.e. English Wiktionary (Q22001375)). Circeus (talk) 23:02, 21 August 2019 (UTC)


Is there a property to indicate that an item or URL is the draft version or manuscript of the item to which the property would be attached? Arlo Barnes (talk) 18:54, 21 August 2019 (UTC)

@Arlo Barnes: We have study or design for (P6606) for the item of the draft. Thierry Caro (talk) 20:28, 21 August 2019 (UTC)

Is Q55900412 the same as Q18202241

We have items for Hermann Buttersack (Q55900412) and Hermann Butterweck (Q18202241), two German actors with the same name, born the same month and died on the same day. They have different VIAF, GND and CERL IDs, but I suspect that they are the same person who had 2 entries in GND which were duplicated in VIAF and here. I am not sure if there is a way to establish if there were two such actors. Any ideas? --Jarekt (talk) 14:12, 21 August 2019 (UTC)

@Jarekt: The names are not the same. But in a case like this you should follow through to the original sources - VIAF etc. may have links to other information about them; if there's any news article or other work that lists them that might help. Searching Google books or might also turn something up. Presumably if they are indeed different people, the death date and perhaps birth date only really applies to one of them. ArthurPSmith (talk) 16:56, 21 August 2019 (UTC)
ArthurPSmith, thanks for pointing out difference in names, I did not noticed it. I did follow all the links I could find and I think they are all derivative of GND entry. Google search for term "Hermann Buttersack" finds no valid hits, nor does Google Books or Perhaps there are other places to search, but other than bare bones GND and mirrors I see no proof that Hermann Buttersack (Q55900412) ever existed. Most likely it is a version of Hermann Butterweck (Q18202241) with a typo. --Jarekt (talk) 19:36, 21 August 2019 (UTC)
That definitely sounds plausible. You didn't find the typo anywhere besides GND (and derivatives) then? ArthurPSmith (talk) 19:38, 21 August 2019 (UTC)
No the name is used only at GND and derivatives (VIAF, ISNI, CERL & Wikidata) and none of derivatives has anything about the person other than the bare bones that GND has: name, dates and occupation. --Jarekt (talk) 20:47, 21 August 2019 (UTC)
I think when this one first came up on Wikidata:Database reports/identical birth and death dates/1, we had a few similar ones, but for these others, VIAF/GND indicated that one name was an alias for the other.
It does happen with pairs that come up on Wikidata:Database reports/identical birth and death dates that names and dates get mixed up, without the two persons being the same.
For the pair discussed here, I had brought it up on Wikidata:Forum/Archiv/2018/11#Same_lifespan, but without much luck either. BTW, GND has some complicated way of indicating its source of their data. --- Jura 10:28, 22 August 2019 (UTC)
Somebody should probably add a note on the item talk pages for these items at least, linking to this and previous discussions! ArthurPSmith (talk) 18:25, 22 August 2019 (UTC)

New tools and IP masking

14:18, 21 August 2019 (UTC)

Johan, one tool I would find very useful for patroling changes, is a single click undo button to reverse bad (malicious or careless) item merges. I see quite a lot cases where some fresh user merges 2 items (possibly using some tool or "game") which is plainly wrong. It takes a lot of work to restore two separate items and it is quite possible that other items linking to the merged items are never corrected properly. I know such tool has nothing to do with IP masking, but if you are looking for potential tools for resent change monitoring, that would be one I could really use. --Jarekt (talk) 19:50, 21 August 2019 (UTC)
Seconded. --Daniel Mietchen (talk) 03:19, 22 August 2019 (UTC)
It's easy if you do it right away, by reverting the two items. It's harder to do it a couple of years and a few hundred edits later, but that would also be hard to automate. Ghouston (talk) 10:40, 22 August 2019 (UTC)
I agree that a one-click tool for undoing merges would be very useful for Wikidata. Especially a tool that also takes care about undoing the changed redirects that happened as a result of the merge. If we had such a tool we could make it easier to discover how to merge items for new users on Wikidata by not requiring to activate a gadget which would be a major step forward. Having a system for the redirects of merges that doesn't rely on a bot would be very helpful for other installations of Wikibase besides Wikidata where there's currently no automatic way to get redirects moved after a merge. ChristianKl❫ 13:25, 22 August 2019 (UTC)

One billion edits --GZWDer (talk) 20:15, 21 August 2019 (UTC)

So do we have now one billion unreliable edits? Or whats the point beyond counting edits? --Succu (talk) 20:28, 21 August 2019 (UTC)

Yeehaw!!! --Denny (talk) 20:35, 21 August 2019 (UTC)

I think it’s great! - PKM (talk) 21:11, 21 August 2019 (UTC)

🎊🎉🍾🥂 Exciting stuff! :) ElanHR (talk) 15:44, 22 August 2019 (UTC)

Change of datatype


Why I must give a DEFAULTSORT for persons when in wikidata is the family name? I think this is a work for a programme and not for the people they work here. Sometimes i have success and wikidata give the surname in the category and sometimes not and I must set a DEFAULTSORT. It must changed! I hate it. It is the baddest thing in this project. Adelfrank (talk) 09:06, 22 August 2019 (UTC)

  • It's something you have Wikipediathat isn't using Wikidata for this. Each Wikipedia is free to use Wikidata (or not). --- Jura 09:09, 22 August 2019 (UTC)

No, sorry - You don't understand me. It is the problem between Wikimedia commons and Wikidata!!! - Look here: [David Ameln] family name Ameln' and in Wikimedia commons he comes under A how Ameln [Category:Tenor opera vocalists] - But [David Pomeroy] family name Pomeroy comes in Wikimedia commons under D how David and I must set a DEFAULTSORT to receive P how Pomeroy. Where I must ask in which Wikipedia for this problem? It is a problem between Wikidata and Wikimedia commons and comes from Wikidata (the source). Adelfrank (talk) 09:38, 22 August 2019 (UTC)

    • I thought you were complaining about Wikipedia. Sorry about that. If it's about Commons, the right place is probably c:Template_talk:Wikidata_Infobox. --- Jura 09:42, 22 August 2019 (UTC)
      • @Adelfrank: If you set both family name (P734) and given name (P735), then you shouldn't need to manually define a DEFAULTSORT, the infobox will do it for you. In this case, I think it's a caching issue - you added the family name at about the same time as the infobox. In that case, try doing a null edit (edit the page, then save it without making any changes), or add "?action=purge" to the URL, in order to clear the cache. Thanks. Mike Peel (talk) 09:49, 22 August 2019 (UTC)
  • Yes, thank you Mike. This was the answer to my question. It's a time-problem. Greetings from Berlin Adelfrank (talk) 10:07, 22 August 2019 (UTC)

Let Abdur Rahman Bikash wiki stay

Abdur Rahman Bikash is a well known person and i strongly recommend you to make the page again, or undo your deletion. please

That won't be happening unless evidence is provided to demonstrate that you pass wikidata's notability requirements, which you'll find at WD:N. --Tagishsimon (talk) 15:00, 22 August 2019 (UTC)

Random lexeme to the sidebar

There is a feature request to include "Random lexeme" tool in the sidebar. Is the community fine with it? --Matěj Suchánek (talk) 11:38, 21 August 2019 (UTC)

  •   Support; lexicographical data activities were IMO pretty invisible recently, so let's push it a bit. --MisterSynergy (talk) 12:17, 21 August 2019 (UTC)
  •   Support Fine by me. Also worth noting Special:Random/L works too (as does Special:Random/P). --SilentSpike (talk) 12:18, 21 August 2019 (UTC)
  • I'm not sure about the usefulness of "random item" in the first place. Most items that come up aren't interesting and inviting to edit. Maybe we can exchange "random item" with "random content" and have random content lead to a page that lists 'random item', 'random lexeme', 'random wikiproject' and 'random property'? ChristianKl❫ 13:48, 21 August 2019 (UTC)
  • I would prefer to replace random item with random entity (item, property, lexeme and whatever stuff we make up in the future). Sidebar we can just change at MediaWiki:Sidebar, but I don't think that is needed. We just need to update $wgContentNamespaces to also include the other namespaces. Multichill (talk) 19:43, 22 August 2019 (UTC)
    • If we have random entity, we should have some kind of weighting in the algorithm so that the fact that we have a lot more items then other entities doesn't mean that practically it shows every time an item.  – The preceding unsigned comment was added by ChristianKl (talk • contribs) at 23. 8. 2019, 11:32‎ (UTC).

University presidents

I'm noticing a couple of different approaches for data about university presidents.

In some cases, an item is created for the specific position, such as President of Harvard University (Q19588706) and this is given as a position held (P39) statement on an item for a person (such as Drew Gilpin Faust (Q49128)). In other cases, the more general item university president (Q4376769) is used for position held, along with of (P642) to name the university (such as Edward John Ray (Q5343795).

Either approach seems valid, but is one preferred?

(I've also seen statements using occupation (P106) with either university president (Q4376769) or an item for a specific position, but that is flagged as a potential issue because university president (Q4376769) is not currently an instance of occupation (Q12737077).) Lrobare (talk) 21:36, 19 August 2019 (UTC)

Short answer (my view): probably the first of those - President of Harvard University (Q19588706) as a position held (P39) statement - since we tend towards increased granularity and because it supports easier queries. Ideally both approaches convey exactly the same information, although in fact the class trees for the two examples are currently worlds apart; Harvard President is a Rector which is a ..., and meanwhile a university president is something else altogether, the two meeting eventually at official (Q599151). So if subclass items such as President of Harvard University (Q19588706) are being relied on, the imperative is to square their class memberships with generic positions such as university president (Q4376769), else they become useless or positively harmful. --Tagishsimon (talk) 22:24, 19 August 2019 (UTC)
+1. Nomen ad hoc (talk) 22:27, 19 August 2019 (UTC).
I only create the entry President of Harvard University (Q19588706) if I am working on a list that has multiple entries, I do the same for Mayor of X and Commissioner of X and other positions. If I only have a single entry I use the second method with occupation (P106) filled in with the generic title. We could always have a bot create Mayor of X for every city in the USA, but almost everyone would have no entries. --RAN (talk) 03:16, 23 August 2019 (UTC)

Q334406 and Q65683402

Are these really not the same thing? One's an 'abbatial church' and the other is an abbey. Renard Migrant (talk) 13:17, 23 August 2019 (UTC)

No. One is a part of the other. It's fine & desirable to have distinct items for the abbey (organisation and/or set of buildings) and the abbey church (one of its buildings). --Tagishsimon (talk) 13:45, 23 August 2019 (UTC)

Lists and positions

Cyprus Minister of Communications and Works (Q38112367) is position (Q4164871). Is it correct to connect this kind of items with a list like en:List of Ministers of Communications and Works of Cyprus? Or is it better to have separate items like list of Ministers of Transport, Communications and Works (Q66583238)? I think the second option is better because en:List of Ministers of Communications and Works of Cyprus is Wikimedia list article (Q13406463) and not position (Q4164871). Xaris333 (talk) 05:05, 19 August 2019 (UTC)

A Wikimedia list and a position are different concepts, so they need different items — that can be linked to each other with is a list of (P360)/has list (P2354). Nomen ad hoc (talk) 06:09, 19 August 2019 (UTC).
Thanks. I have found many items that are wrong. Xaris333 (talk) 06:27, 19 August 2019 (UTC)
Please also link to/from Category:Cyprus Ministers of Communications and Works (Q8439036) using topic's main category (P910)/category's main topic (P301)/category related to list (P1754)/list related to category (P1753) as appropriate. :-) Thanks. Mike Peel (talk) 06:53, 19 August 2019 (UTC)

@Mike Peel: can you check list of Ministers of Defence of the Republic of Cyprus (Q1418781), Q65243720 and Category:Cyprus Ministers of Defence (Q8439042)? Are they ok? Xaris333 (talk) 06:07, 24 August 2019 (UTC)

@Xaris333: It seems that Minister of Defence (Q25907547) already existed, so I've merged Q65243720 into that. Otherwise all is good. :-) Thanks. Mike Peel (talk) 09:27, 24 August 2019 (UTC)

Which "Sun Yat-Sen/Zhongshan Park" the Q2622640 points to?

At least on zhwiki, we do have hundres of such named parks. --Liuxinyu970226 (talk) 10:19, 19 August 2019 (UTC)

The items seems to be a mess at the moment. The English item is a list of parks with that name. Likely, the item should be converted in a list item and I guess you also have an item that lists those parks on zhwiki. ChristianKl❫ 15:21, 19 August 2019 (UTC)
I don't speak it, but from the shape of the article, the linked page on zhwiki looks likely to be such a list - many different places and different images. Agree on converting this item to be a list. Andrew Gray (talk) 11:49, 20 August 2019 (UTC)
@Liuxinyu970226: given that you seem to be interested in the item I think it would be straightforward if you do the conversion to list article. Is there anything about that, that you don't feel comfortable with, so that you want that someone else helps you? ChristianKl❫ 16:47, 20 August 2019 (UTC)
@ChristianKl, Andrew Gray: why not call (and separate) that zhwiki article as Wikipedia article covering multiple topics (Q21484471) instead? There has also parks that are not met both zhwiki and Wikidata notability policies to have items (e.g. if your home has a garden that named "Zhongshan Park" by yourself). --Liuxinyu970226 (talk) 22:51, 22 August 2019 (UTC)
@Liuxinyu970226: Looking at them, I think all the articles are lists of parks - some only have a few entries but none of them seem to be about one specific place. So it should be fine to convert the item. Andrew Gray (talk) 10:12, 24 August 2019 (UTC)


This item, different meanings of the word for the direction left (Q18340392), is used to classify a specific set of disambiguation pages relating to word senses (not homographs or homophones). it should still have some properties, but I am utterly stumped as to which property+item combinations are proper. Circeus (talk) 19:15, 23 August 2019 (UTC)

IMHO, this item should not exist. It isn't linked to any category or wiki article; the concepts "left behind" and "leaving" have nothing to do with direction. If anything it belongs in lexemes. - PKM (talk) 21:30, 23 August 2019 (UTC)
Thank you for unhelpfully answering a questoin that was never asked. Circeus (talk) 02:29, 24 August 2019 (UTC)
As mentioned below, I tend to agree with PKM. --- Jura 11:45, 24 August 2019 (UTC)

Probable duplicate items

I'm pretty sure and represent the same thing and should be merged. I do know how to do merges, so no need to point me there, I just don't have enough knowledge about the royal family of Jordan to be absolutely certain they are actually not somehow different concepts that need to be separately maintained. CanadianCodhead (talk) 15:31, 23 August 2019 (UTC)

  Done the items are merged now. --Jarekt (talk) 22:08, 24 August 2019 (UTC)


Jasper Deng
Egon Willighagen
Denise Slenter
Daniel Mietchen
Andy Mabbett
Emily Temple-Wood
Pablo Busatto (Almondega)
Antony Williams (EPA)
Devon Fyson
Samuel Clark
Tris T7
  Notified participants of WikiProject Chemistry --Micru (talk) 21:46, 24 August 2014 (UTC) Tobias1984 (talk) TomT0m (talk) Genewiki123 (talk) Emw (talk) 03:09, 9 September 2014 (UTC) —Ruud 16:15, 9 December 2014 (UTC) Emitraka (talk) 14:32, 14 October 2015 (UTC) Bovlb (talk) 19:10, 21 October 2015 (UTC) Peter F. Patel-Schneider (talk) 22:21, 23 October 2015 (UTC) ArthurPSmith (talk) 15:51, 5 November 2015 (UTC) --Daniel Mietchen (talk) 20:53, 3 January 2016 (UTC) --Harmonia Amanda (talk) 22:00, 27 February 2016 (UTC) --Lechatpito (talk) --Andrawaag (talk) 14:42, 13 April 2016 (UTC) --ChristianKl (talk) 16:22, 6 July 2016 (UTC) --Cmungall Cmungall (talk) 13:49, 8 July 2016 (UTC) Cord Wiljes (talk) 16:53, 28 September 2016 (UTC) DavRosen (talk) 23:07, 15 February 2017 (UTC) Vladimir Alexiev (talk) 07:01, 24 February 2017 (UTC) Pintoch (talk) 22:42, 5 March 2017 (UTC) Fuzheado (talk) 14:43, 15 May 2017 (UTC) YULdigitalpreservation (talk) 14:37, 14 June 2017 (UTC) PKM (talk) 00:24, 17 June 2017 (UTC) Fractaler (talk) 14:42, 17 June 2017 (UTC) Andreasmperu Diana de la Iglesia Jsamwrites (talk) Finn Årup Nielsen (fnielsen) (talk) 12:39, 24 August 2017 (UTC) Alessandro Piscopo (talk) 17:02, 4 September 2017 (UTC) Ptolusque (.-- .. -.- ..) 01:47, 14 September 2017 (UTC) Gamaliel (talk) --Horcrux92 (talk) 11:19, 12 November 2017 (UTC) MartinPoulter (talk) Bamyers99 (talk) 16:47, 18 March 2018 (UTC) Malore (talk) Wurstbruch (talk) 22:59, 4 April 2018 (UTC) Dcflyer (talk) 07:50, 9 September 2018 (UTC) Ettorerizza (talk) 11:00, 26 September 2018 (UTC) Ninokeys (talk) 00:05, 5 October 2018 (UTC) Buccalon (talk) 14:08, 10 October 2018 (UTC) Jneubert (talk) 06:02, 21 October 2018 (UTC) Yair rand (talk) 00:16, 24 October 2018 (UTC) Tris T7 (talk) ElanHR (talk) 22:05, 26 December 2018 (UTC) linuxo Gq86 Gabrielaltay Liamjamesperritt (talk) 08:44, 21 June 2019 (UTC) ZI Jony Ivanhercaz (Talk) 11:07, 15 July 2019 (UTC) Gaurav (talk) 22:39, 24 August 2019 (UTC) Meejies (talk) 04:38, 29 August 2019 (UTC) SilentSpike (talk)

  Notified participants of WikiProject Ontology @Multichill, Jheald, Jane023, Spinster:

Our modeling of pigments is all over the place. I'd like to find a consensus that works for both chemistry and art materials communities of practice. Here are some current examples:

  • alizarin (Q267813) AKA Turkey red, pigment red 83 <instance of> "chemical compound" matched to AAT's "colorant for dye and pigment".
  • pigment blue 16 (Q414162) AKA phthalocyanine <instance of> "chemical compound" with no mention of pigment despite the labels.

I think we should have some agreement on these points:

  • The named color should be a separate item from the pigment (I think we've reached consensus on this before).
  • Should the chemical compound and pigment be the same item? If so, should the model be <instance of> "chemical compound" and <instance of> "pigment" or something else? (We might have <use>="pigment", but does this make querying for pigments counterintuitive?)
  • Should pigments be linked to their named colors with <named after> on the color and <color> on the pigment?

And lastly, is there enough interest in this topic to make a WikiProject (maybe dyes and pigments) where we can discuss and document best practices? - PKM (talk) 22:04, 23 August 2019 (UTC)

Please no new Wikiproject: already now too few contributors are involved in existing Wikiproject and we are missing an overview of the problems, partially due to the fact that each one is working in his corner.
To come back to the mentioned problem, my first question is: What is a pigment ? To solve ontological problem we need definitions.
If I take the English definition of pigment and chemical compound I have the following:
Can we link material and chemical substance ?
  • base material (Q214609): substance that can occur in different amounts, all with some similar [mixture of some] characteristics, and of which objects can be made up
  • chemical substance (Q79529): matter of constant composition best characterized by the entities (molecules, formula units, atoms) it is composed of
It seems we can find a link but we need to unify the upper classes of both items before to be able to merge anything. But I think a pigment is a chemical substance but not a pure chemical substance: some pigment are mixtures of several chemical compounds (chemical compound is a pure chemical substance). So if we want to model that case too we can't systematically merge pigment and chemical compound. I think we have a similar situation with drug: there are often one chemical active substance, but in really a drug is a mixture of several chemical compounds. Snipre (talk) 21:04, 24 August 2019 (UTC)

Are there a way to block scientific articles from showing up in search results?

It often makes it quite hard to find what i'm looking for. --Trade (talk) 03:01, 18 August 2019 (UTC)

Agree, or have an option to sort them to the list bottom. The usefulness of autocompletion completely vanishes if objects only show up if their labels have been written out. --SCIdude (talk) 06:19, 18 August 2019 (UTC)
See also Wikidata:Requests_for_permissions/Bot/LargeDatasetBot, which will import 12 million more such articles. − Pintoch (talk) 09:41, 18 August 2019 (UTC)
On Special:Search, you can append -haswbstatement:P31=Q13442814 to your searches to exclude all items with instance of (P31): scholarly article (Q13442814). That should already help a lot. Unfortunately, it does apparently not work if you type the search term directly into the search form on the upper right corner. --MisterSynergy (talk) 13:07, 18 August 2019 (UTC)
In other words, autocompletion should be configurable the way Special:Search is. --SCIdude (talk) 13:40, 18 August 2019 (UTC)
@MisterSynergy: Are you referring to the autocomplete dropdown, or to making a Special:Search query by typing text into the search box and pressing Enter? If the latter, you can additionally get item results only by prefixing the search query with ": " (Special:Search defaults to properties only for me, and I'm not sure how to change that; if this is happening for you then this would work). Jc86035 (talk) 07:55, 19 August 2019 (UTC)
Generally the search algortihm tries to rank scientific articles lower then other items. Can you give example of searchers you did where you think that the scientific articles get ranked to highly? ChristianKl❫ 19:12, 18 August 2019 (UTC)
@ChristianKl:*typing "neuropatholog" --> medical specialty at #30 after roughly 39 scientific articles but full "neuropathology" --> medical specialty at #1 --Trade (talk) 11:57, 25 August 2019 (UTC)
@ChristianKl: E.g. aiming at the object labelled "angiotensinogen", typing "angiot" gives a drug object at #1, the rest are all scientific articles. The target shows up with "angiotensino" but only at #5 after 4 articles. --SCIdude (talk) 06:09, 19 August 2019 (UTC)
From looking at that search, that does seem wrong. I'm not sure why the search behaves the way it does. It would be good if we would have >10 examples like this of where the search engine works poorly and give it to the developers to tune the search engine to give better results. ChristianKl❫ 07:02, 19 August 2019 (UTC)
@ChristianKl: As you wish (from my field):
  • typing "kissp" --> only scientific articles but full "kisspeptin" --> gene at #1
  • typing "cathel" --> only scientific articles but "cathelicidin" --> group of peptides at #1
  • typing "preproe" --> only scientific articles but "preproendothelin-1" --> non-article items at #1-4
  • typing "proglu" --> only scientific articles but "proglucagon" --> protein at #1
  • typing "preproi" --> protein at #3 after 2 scientific articles
  • typing "apeli" --> only scientific articles but "apelin" --> six non-article items <==== !!
  • typing "thymosin" --> only scientific articles but "thymosin beta" --> 3 non-article items, enough? --SCIdude (talk) 07:36, 19 August 2019 (UTC)
I think that's a good portion for the drug items. If there's a different items that have a similar problem it would be great to have example items of the problem as well @Lea_Lacroix_(WMDE): can you interface with the development team about tuning the search algorithm again? ChristianKl❫ 07:45, 19 August 2019 (UTC)
it should also help the search algorithm if wi link articles to their main topics. It’s a lot of work but as a very good hint of topic for the article is that the subject of research is present in the article title, it should be able to create a wikidata game or something like that to help to do that. If the topic become used in such statements, it should be higher ranked in the results, or am I being naive ? (of course it’s not an immediate answer to the problem). author  TomT0m / talk page 19:32, 18 August 2019 (UTC)
A simple query shows quite promising results. author  TomT0m / talk page 19:48, 18 August 2019 (UTC)
I'm uncertain whether we get millions of instances of people playing the game. It would be nice to have such a game but I don't know if it will move the needle. More direct Zotero integration would be great for having more main subject claims. ChristianKl❫ 07:02, 19 August 2019 (UTC)
I think it's pretty obvious from the apeli/apelin example alone that articles get ranked first and get only overtaken if the item is spelled out in full, which completely nulls the value of autocompletion. That's a bug in my dictionary. --SCIdude (talk) 07:47, 19 August 2019 (UTC)
As far as I remember we introduced a punishment for items types that come from big imports. Items about the proteins are supposed to rank less well then some other items. Songs and movie titles are behave in a similar way. It would be intersting to know if the problem appears the same way for normal items. ChristianKl❫ 09:35, 19 August 2019 (UTC)
I can understand that (1 million) proteins are ranked below less numerous objects. Scientific articles however are many millions, so by that logic should be ranked to the very bottom. --SCIdude (talk) 13:21, 19 August 2019 (UTC)
I do agree, my point was that it's useful to know whether the problem also exists for other items and that's why it would be good if the people who had problems at commons would also add a few examples, so that everybodies usecases are covered when the algorithm gets updated. ChristianKl❫ 15:30, 19 August 2019 (UTC)

Duplicate Item

Is there a way to duplicate an item from another item or a template? I'm just to lazy to fill in ~20 statements for >100 new items. --D-Kuru (talk) 22:21, 23 August 2019 (UTC)

@D-Kuru: There is User:Magnus_Manske/duplicate_item.js, which you can add to your common.js like so. Mahir256 (talk) 23:34, 23 August 2019 (UTC)
  • If they are all similar, you might want to use QuickStatements --- Jura 23:43, 23 August 2019 (UTC)
    Are there any plans to make QuickStatements more user friendly or less complicated? The fact that you need to own a license of Microsoft Excel in order to use it is kind of a let down. --Trade (talk) 00:31, 25 August 2019 (UTC)
That's a prize-winning fatuous statement, as if OpenOffice or LibraOffice or text editors or OpenRefine don't exist. Quickstatements is minimally complicated given the job it does. --Tagishsimon (talk) 00:33, 25 August 2019 (UTC)
QuickStatements just needs a CSV (or tab-delimited?) file as input, right? Those are industry-standard formats, with several million different tools which can generate them. They're totally not limited to Excel. —Scs (talk) 02:31, 25 August 2019 (UTC)
I quite like the way the WE-Framework gadget looks and works. If you gave it the ability to customize the properties available, save statements and run batches i could totally see it work. --Trade (talk) 03:34, 25 August 2019 (UTC)
Thanks for the info! --D-Kuru (talk) 11:49, 24 August 2019 (UTC)
Why not create an empty item, merge the original into it, and restore the original, as described in Help:Merge? --SCIdude (talk) 12:18, 24 August 2019 (UTC)
I think I saw some user do that and then edit only the English label/description. duplicate_item.js can lead to this as well.
BTW, User:Matěj_Suchánek/moveClaim.js can create a new item .. and allows to just copy statements one needs. --- Jura 12:34, 24 August 2019 (UTC)

Problems of bibliographic records

w.r.t Q66724823

  1. What's the difference between OCLC control number (P243) and OCLC work ID (P5331)? Please add this book's OCLC number 711513523.
  2. Is there a bot that would crawl worldcat and fill up info according to OCLC no. or ISBN?
  3. Searching an item by ISBN is too user-unfriendly. Take the example of ISBN-13 La Fraternité de l'Anneau (Q22137041), searching 9782267027006 would not find it. Seriously? Most lay users would not care about the dashes.--Roy17 (talk) 14:39, 24 August 2019 (UTC)
@Roy17: several things are very strange for this item. First is it a work or an edition? It seems to be an edition but which one? on Commons/Internet Archive, this is the 1988 edition with an ISBN-10 but you added an ISBN-13 (which is for editions after 2007, I corrected this ISBN to add the dashes but I can't find it in any database). For this identifier "711513523", it's a OCLC control number (P243) (for edition) and not OCLC work ID (P5331) (for work).
For the search, true it should be improved to give results even if you input without the dashes; that said, the ISBN numbers should always be stored with dashes (removing dashes is super easy, adding them can be quite tricky and almost impossible to do automatically). @Lea Lacroix (WMDE): could it be possible? (I vaguely remember that CirrusSearch has option for this but I can't find it…)
Cdlt, VIGNERON (talk) 12:33, 25 August 2019 (UTC)
@VIGNERON: thx a lot! I wanted to add the oclc no., but didnt know which to pick, so copied an ISBN. Since I intend to add a record from worldcat, oclc no. is reliable but isbn might not be, so I edited the item.--Roy17 (talk) 12:47, 25 August 2019 (UTC)
@Roy17: no problem, I added a instance of (P31) and the ISBN-10 (P957). PS: for the name of the Commons cat, should it be "sheng wu" or "shengwu"? Cheers, VIGNERON (talk) 13:59, 25 August 2019 (UTC)
@VIGNERON: the current name is in pinyin character by character, so it is sheng wu. How exactly should Chinese book titles be transliterated is not resolved yet on Commons: c:special:permalink/363319893#Categories_for_individual_Chinese_book. Does wikidata have something we can learn from?--Roy17 (talk) 14:06, 25 August 2019 (UTC)

How do we model pen-names?

We have an item Shōzō Numa (Q3959910) who is a pen name of Tetsuo Amano (Q3519206). Tho items have different VIAF numbers and some other identifiers. Should we use one or two items to model this? --Jarekt (talk) 22:28, 24 August 2019 (UTC)

List the alt identifiers under the real name with a stated as (P1932) statement and criterion used (P1013) pen name (Q127843)? (Typically this would be a reason for deprecation (P2241) statement instead of stated as (P1932), but I don't think that's appropriate here.) Circeus (talk) 05:32, 25 August 2019 (UTC)
Shōzō Numa (Q3959910) is shared pseudonyn of unknown authors. Some reference said Tetsuo Amano (Q3519206) is one of them.--Afaz (talk) 07:42, 25 August 2019 (UTC)
oh, then the details of who wrote what are on the individual works, and the pseudonym is a collective pseudonym (Q16017119). Circeus (talk) 16:45, 25 August 2019 (UTC)

"Organization of formation"

I noticed it is a constraint error that the organization NAFS(k) (Q10593756) has location of formation (P740) set to the school Nya Elementar (Q10604257). It's because the school is not a geographical object (Q618123), which makes sense, because a school could change its physical location and still continue as the same entity. So it would be correct to replace this with a geographical object where the school was located. But still, is there a valid way to store that the foundation was at that school? Pst (talk) 06:09, 25 August 2019 (UTC)

The proper way would probably be to create a new item for the school building and use that one as the value for location of formation (P740). --Kam Solusar (talk) 23:43, 25 August 2019 (UTC)


One more time. @Obsuser: vandalised names for ID's like Facebook, Instagram, IMDB or SoundCloud. I want to restore my corrections. Eurohunter (talk) 22:03, 23 August 2019 (UTC)

Eurohunter vandalised by removing translations completely, even those by long ago established users such as Rancher for Serbian (he translates on Translatewiki too).-- 19:02, 24 August 2019 (UTC)
@Eurohunter: Are you saying that he vandalized the archive link? If not, can you point to the divs that you are actually concerned about? ChristianKl❫ 09:52, 26 August 2019 (UTC)
@ChristianKl: There is thread about it. 18:02, 26 August 2019 (UTC)~

Merge Maple Lake Municipal Airport Error

When trying to merge Q1030802 and Q35278441 I get an error "Error while "Please wait...": A conflict detected on cebwiki: Q1030802 with cebwiki:Maple Lake Municipal Airport, Q35278441 with cebwiki:Mavencamp Airport"

Any ideas how to fix this?

cebwiki has two different articles. Wikidata items can only have one. So either one is wrong, or both are the same and need to be merged first. --SCIdude (talk) 05:31, 26 August 2019 (UTC)

In GNIS ID (Q19832959) (the source from where geonames imported, and then from where ceb imported) there are both entries, but notably the Mavencamp airport is listed there as historical. So maybe the airport was relocated? When checking in Google Maps, there's just an empty field at the historical location. Ahoerstemeier (talk) 07:47, 26 August 2019 (UTC)

Wikidata weekly summary #379

Category:In Progress datasets

Seems there isn't much going on there. Some users create pages there and then vanish (or maybe expect something to happen). Samples:

Shall we archive or delete pages that don't even link to a dataset? Pages have a lot of text, but most is template generated.

It seems to be me that project chat would be a better point of entry for people interested in doing an upload. --- Jura 18:38, 26 August 2019 (UTC)

Applications for the Wiki Techstorm 2019 in Amsterdam are now being accepted

Hello everybody,

From today, until 8 September, the applications for the Wiki Techstorm 2019 will be open.

Dutch Wikimedia volunteers are organizing together with Wikimedia Nederland, this time in Amsterdam on November 22nd and 23rd 2019, and we are looking for enthusiastic Wikimedians (and GLAM-mers! and technical non-Wikimedians!) who want to be there.
The Techstorm stems from an initiative of the Dutch Gender Gap group, to promote women's participation online in Wikimedia projects. Although this is still our foundation, we are now offering a low-threshold introduction to the world of Wikidata, Wikimedia Commons and linked open data to a wider audience.

Do you recognize yourself in the following categories?

  • Beginning or experienced developers with an interest in Wiki projects;
  • Experienced Wikimedian who wants to learn (more) about the technique behind Wiki projects;
  • Data specialist, information specialist and digital collection manager working in the heritage sector and/or library sector who would like to learn (more) about sharing collections via Wiki projects?

Then sign up for participation!

More information about the Techstorm can be found on information about last year's Techstorms can be found on Mediawiki.
There are several chapters who offer scholarships for travel and board.

On behalf of the organization a warm greeting,
Ciell (talk) 19:50, 26 August 2019 (UTC)

@Ciell: good news! You expect to accept a lot of them? :-) Multichill (talk) 19:56, 26 August 2019 (UTC)
Hi Multichill! Though we'd love to accept every application off course, we have to limit our number of participants to 60 people. This is now our third time organising a Techstorm and we think we are really developing a formula which creates retention and growth, as well in diversity as in numbers, within the communities, and we hope that we can inspire other chapters to adopt the concept.
Please get in touch if, after reading through our documents and recommendations on mediawiki, you would like to organise a Techstorm in your home country and have questions or need help.
Kind regards, Ciell (talk) 20:03, 26 August 2019 (UTC)

Requesting help

Dear fellow members of this wikiproject, I am new here (I am older in the English wikipedia), so I want to request your assistance. I have created 2 Items, the second is the one I need help. (here). That Greek company, FARCO, was brought by Chrysler and was renamed Chrysler Hellas S.A.. I chose the "replaced by" statement, I typed that but it said it is invalid. I chose the "unknown value" and "no value" options in the small box, but now it displays no place to type. Can someone help me out? Best regards, Enivak (talk) 21:24, 26 August 2019 (UTC)

@Enivak: What is the Wikidata {{Q}} code for the item that FARCO (Q66763562) was replaced by (P1366)? –MJLTalk 21:46, 26 August 2019 (UTC)
MJL You mean Chrysler Hellas S.A? If yes, I think it does not exist, this is generally the problem (has no code) Enivak (talk) 21:49, 26 August 2019 (UTC)
@Enivak: Ah, well then you'll likely need to make that item as well. Basically, the property replaced by (P1366) requires a second item in order to make a complete statement. Alternatively, you can make "Chrysler Hellas S.A." an alias like I previously did. Then you can add parent organization (P749) is Chrysler (Q181114). Does that help explain things? –MJLTalk 21:56, 26 August 2019 (UTC)
MJL I will try to see if that helps, if not I will post it here.Enivak (talk) 22:01, 26 August 2019 (UTC)
Update: I will do the the second thing you said, thank you very. much! Enivak (talk) 22:03, 26 August 2019 (UTC)

PHP script to run a Query gets 403 error

I'm taking my first stab at using a PHP script to read results of a query against Wikidata. When I use a GET URI in a browser window I have no trouble, but when I use the PHP script provided in the CODE link on the query page, I get a 403 http error. Is this a permissions errors (as 403 would suggest)? Do I need an API code or some such thing? A request for permissions? The script I'm working on is more interactive than what I'd call a bot, but maybe the bot rules apply and I need to apply for bot permissions? The fact that the query results page includes the PHP code makes me think it ought to "just" work. Any advice?

Thanks - Kenirwin (talk) 21:50, 26 August 2019 (UTC)

@Kenirwin: Not my area, but possibly a custom user agent issue: --Tagishsimon (talk) 22:36, 26 August 2019 (UTC)
@Tagishsimon: - thanks, that was helpful.

Deletion of birthdates

On Trygve Ottersen (Q66424846) the birthdate have/was been deleted with the reason please provide a reference for this date . Adding a birthdate to a person gives Suggestions citation needed constraint Statements for date of birth should have at least one reference. Is this an indication that whenever I come across an unrefferenced birthdate on Wikidata I shall delete it, or should I just await and have confidence that sooner or later an Reference will be added. My opinion is that even if no reference is provided and it is a suggestion someone have given this date because they hade found it somewhere?. imported from Wikimedia project (P143) can not be used here since the date first have been deleted on the wikipedia and then deleted in Wikidata. Pmt (talk) 12:55, 17 August 2019 (UTC)

Birthdates are very rarely incorrect in Wikipedia / Wikidata, because the subject of a biography (or a family member) typically corrects it if he or she sees that it is wrong. All statements should have a reference, but I wouldn't blindly delete a birthdate without a reference, unless I had specific reasons to believe that it was incorrect. Regards, Kjetil_r (talk) 13:44, 17 August 2019 (UTC)
The subject of a Wikidata item very rarely sees it and therefore is unlikely to correct an error - we should not rely on that happening. Nikkimaria (talk) 00:17, 18 August 2019 (UTC)
But they will see the connected wikipedia lemma, and correct it there. From there we could add it to a list where the dates between different wikipedias and wikidata (and other sources) differ. More sources means more checks so more chances to find errors, and improve the contents from there. Edoderoo (talk) 14:20, 18 August 2019 (UTC)
English Wikipedia does not allow the subject of an article to edit the article. In fact, it is not just "first person" banning, it is any personal involvement with the subject of an article. That means the offspring, or the media agents (for a celebrity), or even me, who was blocked for a COI when I tried to correct the spelling of my ex-girlfriend's son's name. Quakewoody (talk) 16:20, 18 August 2019 (UTC)
When we had the RfC about living persons the community decided against having a bot that goes and automatically removes privacy sensitive statements. This suggest that there's no desire to have people go out and simply remove all uncited privacy sensitive statements. ChristianKl❫ 17:51, 18 August 2019 (UTC)
@Pmt: (because of this edit) Deleting all birthdate statements that don't have at least one reference is one thing, removing a birthdate statement because one was not able to verify it with reliable sources another. Wikidata statements should generally be supported by referenceable sources (see Help:Sources). There are some exceptions, but I don't think that this birthdate is one of them. There are even stricter rules for biographies of living people (see Wikidata:Living_people). Please provide a reference if you want to keep this statement. - Valentina.Anitnelav (talk) 17:56, 20 August 2019 (UTC)
@Valentina.Anitnelav: Wikidata statements should generally be supported by referenceable Sources. Do yuo require this for all unsourced birthdates? Is this discussed within the sosiety. @ChristianKl: With respect to Your comment above, can you provide this discussion With some References? Pmt (talk) 18:04, 20 August 2019 (UTC)
@Valentina.Anitnelav: Can you please in this discussion describe and define the term reliable Soure, what a soure is and where I can find these definitions? Pmt (talk) 18:15, 20 August 2019 (UTC)
@Pmt: Help:Sources lists some kind of sources that may be appropriate references (books, scientific papers, databases, etc.) and in Wikidata:Verifiability you can find more information about the relevance of sources. Help:Sources explicitly excludes Wikipedia as an appropriate source ("Please note that while pages on Wikipedia (and other Wikimedia sites) should and can be added as sitelinks, they are not appropriate as sources for Wikidata statements."). If a source is a "reliable" or "referenceable" or "appropriate" or "authoritative" source does also depend on the context and may be subject of debate, but Bibsys seems quite appropriate to me for this statement.- Valentina.Anitnelav (talk) 18:42, 20 August 2019 (UTC)
@Pmt: Wikidata:Requests_for_comment/Privacy_and_Living_People was the RfC that lead to the current Living people policy and there you see the consensus against bots being used to enforce the policy automatically. ChristianKl❫ 11:34, 27 August 2019 (UTC)

Petscan help

A while ago someone who was more tech savvy than I am set up this pet scan for me, so that I could just enter category name from the English Wikipedia, and it would tell me corresponding articles on the English Wikiquote that were in the same category as defined by Wikidata. It no longer seems to be working and give an error: "manual wiki requested as output, but not set". I'm afraid I'm not pet-scan-inclined enough to know what this means or how to fix it. GMGtalk 12:56, 23 August 2019 (UTC)

When I have the interface preloaded through your link, I can see nothing that would be even close to what it used to do. Either you copied a wrong link or the query isn't working in the new version (although the opposite was suggested). --Matěj Suchánek (talk) 16:41, 23 August 2019 (UTC)
For the old version of PetScan, use So, this PetScan should work. - Premeditated (talk) 17:10, 23 August 2019 (UTC)
  • Thanks all. GMGtalk 18:07, 23 August 2019 (UTC)
@GreenMeansGo: It looks like there's some changes to the way it handles "manual sources" in the new version (you need to specify the wiki in a different place), which is why it's throwing an error. However, weirdly, the new one doesn't seem to like enwikiquote - you can do the query for commons or enwikisource, but not quote. I'll file it as a bug and let you know the URL for the updated PetScan when it's resolved :-). (The answers will be the same, but hopefully the new one will run a bit faster & more reliably for you.) Andrew Gray (talk) 20:21, 23 August 2019 (UTC)
Awesome. Thanks for the help. GMGtalk 20:23, 23 August 2019 (UTC)
@GreenMeansGo: Okay, bug fixed, and here's the completed version. (The old URLs should hopefully convert over now as well) Thanks to @Magnus Manske: for fixing the WQ issue (& for all of Petscan)!
To quickly walk through how this works so you can reconstruct it in future if you want:
  • tab 1, "Categories", tells it to find all pages, on enwiki, which are in "Category:American Nobel laureates". Depth = 0 means it will not include any subcategories.
  • tab 4, "other sources", tells it that we want the results to be given as enwikiquote pages - this is the "use wiki" option at the bottom. It will thus only give answers which have a valid enwikiquote link.
Everything else is left blank/default. Andrew Gray (talk) 12:24, 27 August 2019 (UTC)
  • Awesome. Thanks again. (My previous offer still stands. I'll donate $50 to the charity of your choice for anyone who can get this to work with cat-a-lot.) GMGtalk 12:28, 27 August 2019 (UTC)

Rejected property proposal

When rejected property proposal can be reapplied? Should I create new proposal page or update the old one? Eurohunter (talk) 10:10, 24 August 2019 (UTC)

  • It depends why it was reject. Some just get closed because they are incomplete, stale or none was interested. Unless there is a problem with the closure of the previous proposal, please create a new subpage. --- Jura 10:22, 24 August 2019 (UTC)
  • What do you mean with 'rejected because they are stale'? --Trade (talk) 00:28, 25 August 2019 (UTC)
  • @Trade: Just take a look at a bunch of close reasons. You will find that those words are sometimes used. ChristianKl❫ 08:21, 27 August 2019 (UTC)

Nth generation lineal descendant

Is there a way to model lineal descendants of the Nth generation from an ancestor in relative (P1038) statements? lineal descendant (Q19671714) can be used as a value of type of kinship (P1039) qualifier, but how to provide information for a specific generation? For the 3rd, 4th, 5th and 6th generation, we have specific items such as son's son (Q23684609), great-grandson (Q19682170), great-great-grandson (Q19682183), and great-great-great-grandchild (Q53277248), but how about other generations? Should we create items for each of them? For example, H. H. Kung (Q631416) is a 75th generation descendant of Confucius (Q4604). What would be the best way to model it? Moreover, how to specify that a lineal descendant (Q19671714) is from the male/female line? --Stevenliuyi (talk) 22:24, 24 August 2019 (UTC)


Are Grammies instance of (P31), subclass of (P279) or part of (P361) of Grammy Award (Q41254)? Because now there are all types... --Infovarius (talk) 09:21, 27 August 2019 (UTC)

I think subclass of (P279) should be used all the way through, except to use instance of (P31) for an item describing an individual award event (X person received Y award on this date) - which I don't think we typically add as items anyway, rather using statements with award received (P166) on the item for the person. So basically, it should probably be subclass of (P279) everywhere for awards. ArthurPSmith (talk) 13:48, 27 August 2019 (UTC)

Semi-protected item?

I am not sure what I need to do to qualify to edit Nepal. In the meantime, could someone change the official website to per enWP article? The current entry is a dead website, while my suggested update claims to be an official portal of Nepal. Thanks! Usedtobecool (talk) 09:29, 27 August 2019 (UTC)

Usedtobecool: The item is indefinitely semi-protected, indeed. I have added the link you mention. Esteban16 (talk) 11:01, 27 August 2019 (UTC)

Somebody could merge "Kaprekar's constant" into "Kaprekar algorithm"


I am newbie in Wikidata.

And the form to merge entries shows me a conflict and I don't know how fix.

The entries are:

"Kaprekar's constant" Q854503

"Kaprekar algorithm" Q18413622

Thanks. --Hari Seldon (talk) 10:53, 27 August 2019 (UTC)

The algorithm and the constant are two different concepts and thus merging them shouldn't happen. In this case you will get an error that comes from the fact that frwiki has articles on both topics. ChristianKl❫ 11:13, 27 August 2019 (UTC)
@Hari Seldon: When looking at the error message it seems to me that it quite cleanly explains that frwiki has an article for both items. Is there something in it that feels unclear to you? Maybe we need a more clear error message? ChristianKl❫ 11:45, 27 August 2019 (UTC)

Removed or uncertain archaeological sites

For the next Wiki Loves Monuments competition, we have imported thousands of archaeological sites in Finland. We would like to be able to mark those of them that have been listed but marked as removed or uncertain to be able to use that for filtering them out from the lists. Which method to choose? Remove them entirely from Wikidata, or describe them with a property. Which one? – Susanna Ånäs (Susannaanas) (talk)

If the archaeological sites are real, they shouldn't be deleted from Wikidata. Can you tell us more about the criteria based on which you want to mark them? ChristianKl❫ 12:31, 27 August 2019 (UTC)


It depends on the situation:
Ayack (talk) 12:37, 27 August 2019 (UTC)
Answering both: The two types of sites that should be marked are
  1. The site is possibly an archaeological site rather than a pile of rocks
  2. The site is an archaeological site, but it has either been demolished or removed and relocated
Susanna Ånäs (Susannaanas) (talk) 12:56, 27 August 2019 (UTC)

Vision and strategy papers for Wikidata and Wikibase are published

Hi everyone :)

More than 7 years ago we started the development of Wikidata. We started out with the main goal of building a central data store for Wikipedia. Very quickly we expanded to also covering the other Wikimedia projects. Then came the interest from people outside Wikimedia to use Wikidata’s data to build their own apps, websites, visualisations and more. And last year we witnessed a significant interest in Wikibase, the software behind Wikidata. This means our work has expanded very significantly from what we started out with. We noticed that more and more people need clarity and certainty about where Wikidata and Wikibase are going. Because of this WMDE and WMF started discussing this more deeply and wrote down a lot of the things we have been talking about in the Wikidata and Wikibase communities over the last years.

The result of these conversations include one product vision paper and three product strategy papers that we are publishing today:

  • Vision: This paper gives a high-level overview of where we’d like to see Wikidata and Wikibase evolve and grow, based on the work of the past 7 years and many conversations with community members, movement partners, and other stakeholders.
  • Wikidata for Wikimedia projects: This paper dives deeper into how we see Wikidata developing in the context of the Wikimedia projects and how it can better support the projects.
  • Wikidata as a platform: This paper explores the value of Wikidata as a source of data for others outside Wikimedia to build upon.
  • Wikibase Ecosystem: This paper goes deeper into the idea of the Wikibase Ecosystem, a network of Wikibase instances which share data and other capacities.

I would love for you to read these papers now and give your feedback. We are collecting feedback until the end of October and will then integrate it as appropriate into a second iteration of the papers.

The papers are published at m:Wikidata/Strategy/2019. Please leave your feedback and questions on the associated talk page.

I am looking forward to reading your input.

Cheers --Lydia Pintscher (WMDE) (talk) 16:52, 27 August 2019 (UTC)

Alternative ship name: qualifier?

Hi, after editing the item about the vessel Dida, I got warning signs about the property 'Alternative name' that I had chosen to represent this information about the various names of the vessel:

  • Amstel (1998)
  • Fort Mingan (1995)
  • Nordbec (1970)
  • Oslo (1969)
  • Dida (1962)

The information blob says that 'Alternative name' should only be used as a qualifier, but I can't find the property to which this property belongs. Anyway such a qualifier would need a subqualifier for the year.

I believe it would be best to create a property (re)name or christening with qualifiers for date, place, the organisation (shipping line, navy) that ordered the (re)christening, and possibly the ceremonial christener and more. Merchant vessels are sold frequently as illustrated above and subsequent shipping lines generally apply new names.

Maybe there is already a solution in place I am not aware of? Thanks! — bertux 12:50, 27 August 2019 (UTC)

You should use official name (P1448) with start time (P580) and end time (P582) instead. Ayack (talk) 13:05, 27 August 2019 (UTC)
Actually, since there are various properties used (short name (P1813), native label (P1705), etc.) you should ask your question to Wikidata:WikiProject Ships. Ayack (talk) 13:14, 27 August 2019 (UTC)
See Wikidata talk:WikiProject Ships#Ship rechristening. Thanks! — bertux 19:17, 27 August 2019 (UTC)

Merge needed

If this is posted here wrong...., please replace it to were it belongs!

I like to see the item: No label defined (Q65767052) (is Dutch article “Zeslandentoernooi 2019 (vrouwen)” and created by User:BotMultichillT) merged with: 2019 Women's Six Nations Championship (Q58284509) (en:/fr:/it:) If someone of the wikidata-group can do so...; thank you very much. Greetings, Pucky (talk) 17:12, 31 August 2019 (UTC)

  Done --Liuxinyu970226 (talk) 23:54, 31 August 2019 (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. Matěj Suchánek (talk) 06:40, 2 September 2019 (UTC)

Use P131 as a qualifier for a settlement ?

It seems that there's a lot of misuses of P131 as a qualifier.
I think there's 2 typical misuses :

  • When P131 point to a division who is containing the qualified statement, as I can see, by example, in P19 :
    SELECT ?lnLabel ?laLabel ?item
      ?item p:P19 ?n .
      ?n ps:P19 ?ln .
      ?n pq:P131 ?la .
      SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
    Try it!
  • When P131 point to a subdivision of the qualified statement, as I can see, by example, in P17 :
    SELECT ?lnLabel ?laLabel ?item
      ?item p:P17 ?n .
      ?n ps:P17 ?ln .
      ?n pq:P131 ?la .
      SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
    Try it!

I think that in the first case, the qualifier can be delete. But before doing this on ~23,000 items, the French chat propose to me to expose the situation here to have more feedback.
So, what do you think about that Simon Villeneuve (talk) 20:53, 27 August 2019 (UTC)

It can be useful when the parent division is not obvious at the time used in a statement. Some Wikipedias probably want (for infoboxes) but can't derive a full chain of hierarchy at specific moment of time so they depend upon such information. --Infovarius (talk) 21:52, 27 August 2019 (UTC)
I'm curious (as I believe it is an overused property even as a main statement): are there any legitimate use of located in the administrative territorial entity (P131) as a qualifier to begin with? Circeus (talk) 23:15, 27 August 2019 (UTC)
Take Maade (Q1882240) - while using located in the administrative territorial entity (P131) as a qualifier for country (P17) is IMHO nonsense, it is totally valid for the mouth of the watercourse (P403) statement - and I also added it a few times for rivers in past that way. Ahoerstemeier (talk) 07:47, 28 August 2019 (UTC)
I think it can be useful for situations like with Middlesbrough (Q2673020) where located in the administrative territorial entity (P131) is North Yorkshire (Q23086) but on the next level it has alternatives. In most other cases, this is absolutely not necessary. Сидик из ПТУ (talk) 10:43, 28 August 2019 (UTC)
@Сидик из ПТУ: I don't understand : P131 isn't used as a qualifier on Q2673020. Simon Villeneuve (talk) 13:24, 28 August 2019 (UTC)
Yes, but I think this would be a great solution to build the correct geo-chains in similar cases. Using P131 here like a switch is better than applies to part (P518) on the next level because we can find more difficult cases where such refinements will be needed for each farm or rock, not just for five districts. For example, Q46706877 is situated on territories of three districts (read Ukrainian article). Сидик из ПТУ (talk) 13:51, 28 August 2019 (UTC)

Automatically display list of properties for an item

(Inspired by similar title above) Any big item with a lot of properties would be more browsable (and editable) if there were some "TOC" of properties used in the item. Can anyone write a gadget for it or may be even to propose this for Wikibase? --Infovarius (talk) 21:47, 27 August 2019 (UTC)

  Support --SilentSpike (talk) 09:10, 28 August 2019 (UTC)
  Comment I recall Wikidata items used to have TOC. But it showed just headings, like Claims or Wikipedia and was removed later. --Matěj Suchánek (talk) 16:08, 28 August 2019 (UTC)

Audio engineer

There seem to be differning uses of audio engineer (Q128124) (in English: "audio engineer", alises include "sound engineer") for both people who engineer speakers and the like, and people who assist record producers. Is there a better alternative for one of these, or do we need an additional item? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:15, 28 August 2019 (UTC)

@Pigsonthewing: I would love to see "recording engineer" separated out, at least. In general this area could do with some fleshing out by someone who knows the field. On a side note, the term "engineer" is used very loosely by English speakers while (at least in) Scandinavia it's a protected title that comes with a certain education. So what you call engineer, would often be a mere "technician" to us ;-) Moebeus (talk)

Words with a masculine and a feminine form

Beeyan Rachmat04 Fadirra

  Notified participants of WikiProject Language

In some languages such as Italian, French and German, nearly all proper names, occupations, nationalities and adjectives (maestro - maestra; lehrer - lehrerin; docteur - docteure; bianco - bianca) have a masculine and a feminine form. Currently, Wikidata has a lexeme for each form, but there's no link between one form and the other. What's the best solution:

  • Create "feminine form", "masculine form" and "neutral form" properties to link one form to the other;
  • Create only one lexeme for such words and point out the masculine and feminine forms in the Forms section.

Lore.mazza81 (talk) 00:09, 26 August 2019 (UTC)

You yourself used the term "form" to describe the differences, so it seems clear to me they should be considered different "forms" of the same lexeme. There may of course be other opinions on this. If somebody has a good reference, I'm sure this has been discussed in the lexicographical community! ArthurPSmith (talk) 17:33, 26 August 2019 (UTC)
@Lore.mazza81: Swedish has two, three or four "forms" depending on how you define the model. Add to that "gender-free form", which is none of these four (or was is five?). I would therefor prefer a property with the label "form" where you could add whatever you like of any form there could be in the wide world. IP (Hej!) 19:51, 26 August 2019 (UTC)
@ArthurPSmith: I prefer this option, too. It seems to me the best one also because it makes easier to use the "translation" property in the case of words that have no gender in a language, but many of them in another.
For example, the english word "white" can be translated as "blanco" and "blanca" in spanish, as "bianco" and "bianca" in Italian, as "blanc" and "blanche" in French, as "weiß" and "weiß" in German...
@Sextvå.tvånoll.ettsjunoll.sjufyra: What's the difference between what you propose and the "Forms" section?--Lore.mazza81 (talk) 01:07, 27 August 2019 (UTC)
I'm not sure if the samples from different lexical categories of random languages are really equivalent. Translations would generally be between specific senses of lexemes. Some languages have countless forms for most words, so you can't easily define them in the same way as in other languages. It should be possible to find an equivalence between languages independently of the chosen approach. --- Jura 02:40, 27 August 2019 (UTC)
@Lore.mazza81: My main point is that linking to the feminine form by a specific property does not make much sense. Then you have to create a new property for all forms there are in the universe. As I said, Swedish has more forms than only "feminine", "masculine" and "neutral". When I was in school, we learned four forms, the forth was a "second neutral" form. My sister learned two forms: neutral and "non-neutral". In practise we today use a "gender-free" form that isn't neutral. That expands your model of three new properties to six only by expanding into one single language. IP (Hej!) 08:41, 27 August 2019 (UTC)
@Sextvå.tvånoll.ettsjunoll.sjufyra: I don't like the property approach either.
@Jura1: In my opinion the spanish words "blanco" and "blanca" are different forms of the same lexeme, whose sense (or at lease one of its senses) is equivalent to (one of) the sense of the English lexeme "white". On the other hand, if we consider "blanco" and "blanca" two different lexemes (the current approach), none of their senses can be considered equivalent to any sense of the "white" lexeme.--Lore.mazza81 (talk) 02:31, 29 August 2019 (UTC)
I'd think Spanish would make just one entity for this adjective. We probably didn't quite get to the point to seek a way to flag senses of adjectives that apply to just some of the forms. --- Jura 08:14, 29 August 2019 (UTC)

Merger request (kind of)

There are two (2) overlapping items, Q1045852 and Q10809927, the former includes any articles in various language and the latter only the Vietnamese-language article w:vi:Quan (tiền), well "Quan (tiền)" is about the currency unit "String of cash coins", going down the list of articles from the item "Q1045852", the German, Lithuanian, and Korean articles are exclusively about the "貫" weight measurement unit used for cultured pearls, while the English, Mandarin, and Japanese are all primarily about the currency unit, but also fully explain what the unit of mass is. Obviously the Vietnamese article should be in the same list as the English, Mandarin, and Japanese articles as well as the Wikimedia Commons (Commonswiki) article while the German, Lithuanian, and Korean language articles are about a concept which is mentioned in all other articles except for the Vietnamese one.

While the simplest solution would be merging these two items, the scope of the articles overlap in some cases, but not in others, what would be the best solution here?

Maybe the English, Mandarin, and Japanese Wikipedia's should have separate articles for the weight unit, but this seems to be a bit more complicated than that. -- Donald Trung/徵國單  (討論 🀄) (方孔錢 💴) 19:29, 28 August 2019 (UTC)

In theory the Japanese language article could become a disambiguation and be split into the articles "w:jp:貫文" and "w:jp:貫目", the same could also happen on the English language article and the Mandarin one (which contains the least information on the Japanese unit used in the cultured pearl-industrial complex). -- Donald Trung/徵國單  (討論 🀄) (方孔錢 💴) 19:33, 28 August 2019 (UTC)

Interesting topic. We can not force Wikipedias to split or merge articles, it depends only their own policy and the size of the content in that language. To me it sounds like a valid case for merging, provided we specify the right instance/subclass statements. Cheers! Syced (talk) 03:18, 29 August 2019 (UTC)

Death properties

how does one do a "circa" death on here? in particular a "died after x-time/date" Lx 121 (talk) 08:18, 29 August 2019 (UTC)

sourcing circumstances (P1480) with circa (Q5727902) gives you circa. For died after x-time you can use century precision for the century in which someone died and then use earliest date (P1319) which the earliest date. ChristianKl❫ 09:22, 29 August 2019 (UTC)

Is there some particular help page where this is explained? - Jmabel (talk) 15:04, 29 August 2019 (UTC)

Help:Dates#Qualifiers. --Matěj Suchánek (talk) 15:24, 29 August 2019 (UTC)

Storing in Wikidata the preferred language variant used on Wikipedia?

Hi all,

I was asked a question by User:Boshomi that I’m not too sure about, so relaying here :)

On the German-language Wikipedia, editors decide when an article should use Austrian German (Q306626) spelling, and add it to a tracking category (Q32078103) − that’s about 17K articles.

(Worth noting that this use-case is not specific to German-language Wikipedia − English-language Wikipedia also has Category:Wikipedia articles by national variety of English (Q8110126))

User:Boshomi was wondering whether it would be possible to somehow store that information on Wikidata. It would then be possible to use that information on the article (for example for formatting the dates automatically in Citation templates, as I understand it).

As I understand it, whether an article 'should' use Austrian-German spelling is not something that can easily be inferred from the already existing properties (such as country (P17) or country of citizenship (P27)) − it’s an editorial decision made by the editing community based on several factors. It would then need to be stored as its own thing.

On the face of it, my first thought was that we store here statements about the entities, not about the Wikipedia articles about it (except for the Badges). That said, we do have several Wikidata property about Wikimedia entities (Q51118821). So I am relaying the question here in order to get a more definite answer :)

Thanks, Jean-Fred (talk) 16:29, 28 August 2019 (UTC)

To clarify the use case: that information is stored Wikipedia-side using tracking categories, but this is not accessible from other templates used on the page − or at least it’s not clear how to do so: maybe there is some Lua function to do exactly that :) Jean-Fred (talk) 16:37, 28 August 2019 (UTC)
I looked for a lua function and for magic words, but I could find nothing. The easy way is invoke a property from wikidata. –Boshomi (talk) 18:03, 28 August 2019 (UTC)
It seems to me like an editorial decision made at dewiki based on dewiki policy that only concerns dewiki (other Wikis might make different decisions about how a name is supposed to be spelled). Such information seems to me like it's ideally stored in a dewiki tracking category and not on Wikidata. ChristianKl❫ 09:27, 29 August 2019 (UTC)
@ChristianKl: Yes, it is an property of a dewiki article. The tracking category exits Q32078103. The use case for wikidata is template programming:
 if invoke dewikiLang=de-AT then Jänner else Januar end
 if invoke dewikiLang=de-CH then replace(sometextwith_ß , ß, ss) end
The result is a text output of templates that fits to the article, and there is no need for editing tousends of articles to put same extra parameter into each templates transclution.–Boshomi (talk) 18:03, 29 August 2019 (UTC)
  • May be these wikis should create appropriate "badges" which could be assigned to such articles? --Infovarius (talk) 20:14, 29 August 2019 (UTC)

Modeling the ratings and reviews of critics

Is anyone aware of a precedent for general modeling of quality ratings and reviews by critics?

Expert organizations and commentators routinely rate and review media, places, products, and organizations. Here are some examples:

Every big budget, book, music collection, video game, and movie gets critic ratings from lots of publications. In academia all sorts of individuals, organizations, and journals rate and review everything that is the subject of human discourse. Citizens, corporate entities, and government rate and review each other continually.

What discussion already exists about bringing this into Wikidata? I am especially interested in brief ratings, either numbers on a scale, a grade, single words, or a generalization like pass/fail. Let's put complex ratings and reviews, like essays, aside for now.

Possible paths forward
  1. Have Wikidata properties for every sort of rating
    • Seems unlikely as there are endless rating systems and the majority are small datasets of <100 ratings
  2. Have some general property which we adapt with qualifiers
    • Perhaps we match this with identifiers, as many ratings are from data collections?
  3. Something else?

Thanks. More than anything, I wish for anyone to speak up if they remember or can point to any previous or ongoing conversation. Blue Rasberry (talk) 14:14, 29 August 2019 (UTC)

Hi Bluerasberry. I can give an overview of how things are done for video game (Q7889).
The current status is captured at Wikidata:WikiProject_Video_games/Properties#Reception_properties:
Hope that helps, Jean-Fred (talk) 14:56, 29 August 2019 (UTC)


SELECT ?review ?game ?gameLabel WHERE {
  SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
  ?review wdt:P6977 ?game.
  ?game wdt:P31/wdt:P279* wd:Q7889.

Try it!

  Resolved This is the right answer and what I wanted. I gave a more complete thanks at WikiProject Video Games.

Thanks a lot. Blue Rasberry (talk) 20:04, 29 August 2019 (UTC)

Need physician physicist

Can we merge Q27459116 and deuteron (Q503527) ? Snipre (talk) 08:14, 28 August 2019 (UTC)

Physican here but a physicist would be more appropriate in this case ;-) Kpjas (talk) 08:35, 28 August 2019 (UTC)
@Snipre: I'm a physicist - though the properties on Q27459116 look more like a chemistry point of view. But both items are about the same fundamental object, so it seems fine to me to merge them. ArthurPSmith (talk) 18:11, 28 August 2019 (UTC)
I thought deuterium is an atom, and a deuteron is a nucleus of deuterium - meaning they have for example different mass?--Ymblanter (talk) 19:03, 28 August 2019 (UTC)
deuterium is an atom or isotope, yes. But the 1+ ion is missing its electron, so it's only the nucleus. ArthurPSmith (talk) 20:30, 28 August 2019 (UTC)
I thought deuterium is a (chemical) substance while deutron is a type of nucleus. --Infovarius (talk) 20:16, 29 August 2019 (UTC)
  Done Snipre (talk) 07:48, 30 August 2019 (UTC)

Missing items from WLM lists on itwikipedia

Hi, I am working with User:Giaccai on WLM lists for Tuscany and since one week we have been having problems with bizzare Listeriabot removal on itwikipedia. Some of these items were previously indexed but are not "seen" now by the system. For example this removal is where Q63982735 suddenly disappeared, few days ago it was only those items (that are still regularly removed if inserted). So it's hetting worse, apparently. Today we cannot insert almost anything from the civil parish of Vernio, at the bottom of the list example. We have similar situation in other local list but for some reason this is the worst one and it gives you a clear idea just in one place. I have however no clue why this happens so far, mostly because I have few time to analyze it. I hoped it would just naturally disappear...

The Italian community gave no resolutive answer by email so... any advice? It's not about local null edits or something like that. I also massively edits the items on various details just to be sure it could unblock something, it's not related to warning signs such as for example missing commons categories... One troubleshooting test I did not perform is to see if the problem is also on other wikipedias, I could do this as well but we are busy with hundreds of items to insert during these last days, i hope somebody can give me some advice before September, 1st. Otherwise the only temporary solution we have is to insert them manually every day to see them removed...--Alexmar983 (talk) 15:42, 30 August 2019 (UTC)

  • I think Q63982735 disappeared as the item was merged the day before with another item that was already in the list. --- Jura 15:51, 30 August 2019 (UTC)
Q66828154 is excluded from the query as P2186 has an end date. [24] --- Jura 16:02, 30 August 2019 (UTC)
but if the end date is this year, it should be excluded next year. i can try to put an end date and strat date together, it this is how is encoded.--Alexmar983 (talk) 16:04, 30 August 2019 (UTC) No let's try point in time.--Alexmar983 (talk) 16:07, 30 August 2019 (UTC)
ok now it works but I am not sure it's formally encoded correctly. For example if my WLM code is valid from year 2014 to 2019 (up to the end of the town council, as it could happen... I cannot encode it as "start date=2014, end date=2018", plus "point date=2019". it should be able to understand that end date implies the removal from the following year...--Alexmar983 (talk) 16:12, 30 August 2019 (UTC)
Maybe [25] or [26] works excludes the ones you are interested. The second query takes in account date precision. --- Jura 16:44, 30 August 2019 (UTC)
Don't worry, they are done manually to fix all minor details including external ID and commons category, so we know them quite well. we spot if a code is missing in the list. Thank you for the queries BTW, they might useful in other regions if they also refine this qualifier as we start to do (but still, this upper limit should not be encoded this way, when we say the last year we clearly include it).--Alexmar983 (talk) 17:16, 30 August 2019 (UTC)

Jura I still have doubts... I wish I had more times but WLM sucks a lot of energy with the social media campaign and reply to civil servants, I just try to do my best with some clear instruction and understand later. So if you have time to help me consider Calcinaia, I put "point in time 2019", as it should be and now it's gone. But I am quite it was not like that last year, I put "point in time" 2018 to at least one case, and they were listed. I might be tired but I feel something is different. You changed the code here right? They did modify something, maybe a comparison might show some clue. I am sorry i am just too full of stuff to process urgently to understand it right now. Thank you a lot.--Alexmar983 (talk) 17:58, 30 August 2019 (UTC)

  ?item wdt:P131* wd:Q16250 . 
  ?item wdt:P2186 ?wlmid .
  OPTIONAL { ?item p:P2186 / pqv:P580 [ wikibase:timeValue ?start ; wikibase:timePrecision ?startprec ]  }
  OPTIONAL { ?item p:P2186 / pqv:P582 [ wikibase:timeValue ?end ; wikibase:timePrecision ?endprec ]  }
  OPTIONAL { ?item p:P2186 / pqv:P585 [ wikibase:timeValue ?date ; wikibase:timePrecision ?dateprec ]  }

Try it!

Maybe the above helps. Seems I confused start time with point in time, but given the data, it didn't matter. If you know what dates you want to exclude (with or without a given precision), I can try to adjust the query. --- Jura 18:06, 30 August 2019 (UTC)

Automatically display lists of properties for Wikidata property items

For Wikidata property items like Q18610173 (Wikidata property to link to Commons), it would be helpful to display a list of the corresponding properties like icon, image, Commons gallery, etc. on the Q18610173 page. The closest we have right now is the "What links here" link, but that's not limited to just the properties that are an instance of the property item, and it's not as easily accessible as including the list directly on the page. Ariutta (talk) 17:26, 26 August 2019 (UTC)

The new gadget relateditems makes this possible. --Matěj Suchánek (talk) 08:39, 27 August 2019 (UTC)
That looks great. Thank you! Ariutta (talk) 21:45, 30 August 2019 (UTC)

Years of service for military person

I'm doing an integration of a local wiki "Infobox military person" with Wikidata, and there is a parameter "Years of service" which is a range of years, like "1879-1918". I was not able to find a property for that. Does anybody have an idea how can I define that in Wikidata? --StanProg (talk) 15:49, 28 August 2019 (UTC)

Use qualifiers start time (P580) and end time (P582) on whatever is the most relevant statement. Jheald (talk) 17:20, 28 August 2019 (UTC)
I've just opened Wikidata:Property proposal/Significant number. Perhaps it could help, when start and/or end times aren't known precisely? Nomen ad hoc (talk) 17:41, 28 August 2019 (UTC).
That could be occupation (P106) military officer (Q189290). Thank you both. :) --StanProg (talk) 19:26, 28 August 2019 (UTC)

this one might work work period (start) (P2031) Andber08 (talk) 21:56, 30 August 2019 (UTC)

How to find a relevant property

Sometimes I need to find a matching property to express a relationship. For example, today, I wanted to connect the alligator Saturn (Q30600555) to Moscow Zoo (Q613676), where he is held, and I needed to figure out the appropriate property for that.

One way that might be helpful is the following SPARQL query:

It tells you which properties connect animals with zoos how often - and in the Query Helper UI it should be easy to change either types to figure out good candidates for the property you are looking for.

I thought this might be helpful to share.

Also, I recommend reading the article on Saturn (Q30600555) --Denny (talk) 16:53, 30 August 2019 (UTC)

@Denny: oh that will be very useful. Thanks! - PKM (talk) 19:08, 30 August 2019 (UTC)
Very useful query! Perhaps it can be added to the current set of SPARQL examples. ElanHR (talk) 20:14, 30 August 2019 (UTC)

Rock paintings into paintings with genre rock painting?

I would like to discuss the justification for a mass edit of cave painting (Q214334) into painting (Q3305213) with genre (P136) cave painting (Q214334). Q11863834 one example. I wonder if it is a recommendation made in a project or just individual reasoning. In any case, I would like to oppose to this change, as it removes any hierarchies that cave painting (Q214334) could have that relate to archaeology, and promotes a visual arts perspective over others. @Nurni: – Susanna Ånäs (Susannaanas) (talk) 07:28, 29 August 2019 (UTC)

Provisionally AGREE with the above objections. seems arbitrary & limiting; could as easily be rationalised into a common group with rock-carvings into "cave art", with the property of being a painting :p & the archaeological considerations certainly apply! Lx 121 (talk) 08:30, 29 August 2019 (UTC)
Tried pinging the project Visual arts, but seems I cannot use the ping correctly. Sorry for the inconvenience, removed the blurb, but the respective members were probably pinged. – Susanna Ånäs (Susannaanas) (talk) 10:35, 29 August 2019 (UTC)
Agree - cave paintings are not, in the strictest sense, paintings as we refer to them in the d:Wikidata:Sum of all paintings project. Jane023 (talk) 12:57, 29 August 2019 (UTC)
As per this discussion I will revert those changes that I encounter in the set of items I am working with from the Finnish Heritage Agency (WLM for Finland). Thanks for your comments! – Susanna Ånäs (Susannaanas) (talk) 05:46, 31 August 2019 (UTC)
Consequently, the cave painting (Q214334) item needs revision, as its instance of (P31) is genre of painting (Q16743958). There is also a shift in the concept between the Finnish and English translations. The Finnish Wikipedia article and the source I added refer to paintings on rock surface while the English description and article refer to paintings on cave walls. I hope we can set up an alignment project for archaeological concepts! – Susanna Ånäs (Susannaanas) (talk) 06:17, 31 August 2019 (UTC)
@Susannaanas: I don't think I would call it a genre of painting. See Wikidata:WikiProject sum of all paintings/Top genres for often used genres. I made cave painting (Q214334) a sublcass of mural (Q219423) (which is a subclass of painting (Q3305213)). Multichill (talk) 14:56, 31 August 2019 (UTC)
Thanks, I think that should link it to painting. Perfect! – Susanna Ånäs (Susannaanas) (talk)

This user looks like spamming

Zyksnowy (talkcontribslogs)

-- 18:21, 29 August 2019 (UTC)

IMHO you should better consider opening an administrator noticeboard's request, if necessary. Nomen ad hoc (talk) 18:36, 29 August 2019 (UTC).
Some edits are clearly bad (spam or vandalism), some are good, the user has been around for a long time, and there are talk page discussions. Does anybody know the user?--Ymblanter (talk) 18:49, 29 August 2019 (UTC)
See also Wikidata:Requests_for_deletions/Archive/2018/08/06#User:Zyksnowy/sb for a previous discussion about this user.--GZWDer (talk) 20:47, 29 August 2019 (UTC)
Are the Chinese descriptions any good? I don't really like strings that get saved before the final version. --- Jura 03:15, 30 August 2019 (UTC)
It seems to me that most Chinese descriptions (final versions) added by the user are acceptable, but the overall quality is not that good, and some are probably machine-translated. --Stevenliuyi (talk) 04:44, 30 August 2019 (UTC)
I asked the user to stop saving random strings to zh aliases/descriptions. --- Jura 16:18, 31 August 2019 (UTC)

subclasses of physical quantity (Q107715)

I'm interested in an ontology of terms used in physical science. With it I found that the graph of subclasses of physical quantity (Q107715) contains a huge amount of objects which should not be subclasses of physical quantity, for example carnival (Q4618) should never be a physical quantity. Also the time measuring unit minute (Q7727) is not a physical quantity (it has a time interval but it should not be a subclass of "time interval"). Also events have time properties but they are never physical quantities. For more details about my concern see that discussion. ArchibaldWagner (talk) 19:52, 30 August 2019 (UTC) ArchibaldWagner (talk) 07:22, 31 August 2019 (UTC)

w:Template:Cite Q for de and es?

What needs to happen so w:Template:Cite Q includes links to version in Deutsch and Español?

It's currently available in 12 other languages but not Deutsch nor Español.

I've been wanting to translate the Wikipedia article on w:Julia Cagé from English into Spanish, but I want to use Wikicite for that. I translated that article from French into English, but I don't edit in Spanish enough to be sufficiently familiar with the citation template to want to even bother with that. I'd rather use Wikicite.

Thanks, DavidMCEddy (talk) 04:53, 31 August 2019 (UTC)

  • Not sure I understand, but the template you linked is on en-wiki. You'd need to create an equivalent template on the relevant wiki. If you aren't at the relevant level in Spanish (or German) or even just with the citation mechanisms in those Wikipedias, you'd need to get someone with the relevant skills to do that, which I would think you are more likely to find on the relevant wiki than here. - Jmabel (talk) 17:49, 31 August 2019 (UTC)

Lac (resin), shellac, and lac dye

lac (Q15112450) and shellac (Q429659) seem to be confused. I think this is because we really need three items: (1) "lac" (the insect resin) (= AAT 300012946), which is used for (2) "lac dye" (=AAT 300013605) and (3) "shellac" (varnish) (= AAT 300014918). The two current items look like (2) and (3) with some articles about (1) linked, but I am not certain. Can anyone help? - PKM (talk) 22:24, 29 August 2019 (UTC)

And here's the third item lac dye (Q31836648). I should be able to sort these now. - PKM (talk) 05:59, 31 August 2019 (UTC)
All sorted now and referenced: lac (Q15112450), shellac (Q429659), lac dye (Q31836648) and Indian lake (Q66921800). The sitelinks were actually fine judging by the content of the articles. - PKM (talk) 22:02, 31 August 2019 (UTC)

Harlem Yacht Club confusion

Harlem Yacht Club (Q20642319) is a yacht club on City Island (disclosure: I'm a member). There is (was?) also a rap group named Harlem Yacht Club, which as far as I know has no wikidata entry. What's confusing is that WorldCat has an entry which somehow merged the two (as in, "which part of unique identifier did you not understand?"). What I'm trying to figure out is if there's something messed up in wikidata that's causing this, or if it's something WorldCat managed to mess up on their own. I'm guessing the later, but figured I'd start here. RoySmith (talk) 23:39, 29 August 2019 (UTC)

I don't see that Wikidata has ever had anything about the rap group - certainly not in the history of that item. LoC and other linked sites also seem to be about the club, not the rap group. Seems to be purely WorldCat confusion? ArthurPSmith (talk) 00:16, 30 August 2019 (UTC)
Thanks. That's what I figured, but confirmation is appreciated. RoySmith (talk) 00:20, 30 August 2019 (UTC)

Bulk Loading of Entity URI

I am writing from the National Library Board Singapore. We have previously created a property National Library Board Singapore ID (P3988) to insert our name authority URI to existing Wikidata items. An example is the Wikidata item Catherine Lim (Q5052793) which is inserted with the National Library Board Singapore ID We would like to bulk import our name authority URIs that match with the same Wikidata items. How do we proceed? (Nlbkos (talk) 02:57, 30 August 2019 (UTC))

Not clear what advice you need. Maybe quickstatements to enter data. Maybe a federated SPARQL query - presuming you can federate your RDF datastore and wikidata - to match candidates (e.g. based on VIAF, other common identifiers, or names and dates of birth)? You certainly don't need permission to proceed; the major questions are how & on what will you match items with your records. --Tagishsimon (talk) 03:10, 30 August 2019 (UTC)
Did you already match them with Wikidata items? If yes, you could use QuickStatements to upload the identifiers.
If you first need to match them, it depends on the other data or tool you have available/is available. Apparently P3988 is also in VIAF, so it might be possible to find the applicable items with that. --- Jura 03:12, 30 August 2019 (UTC)
@Nlbkos: can you describe what you want to do? With this query you can get the list of what is already matched. Do you already have matches or are you looking for a tool to make these matches? Multichill (talk) 13:13, 31 August 2019 (UTC)
@Nlbkos: As there are currently only 42 items in Wikidata with a value for P3988, I assume you wish to add further values to Wikidata. If you have already made the matches in your system, you can export them as a TSV (or speadsheet) file, then add the IDs to Wikidata using QuickStatements; if you have not, and need help with the matching, and to then add the matched values to Wikidata, you can use Mix'n'Match. Please let me know if you need help with either. it would also be good if you could display Wikidata IDs (and links) on your pages, like gzmyZ3wowyA. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:47, 31 August 2019 (UTC)

How to connect a "conference proceedings series" to its "conference series"?

JSM proceedings (Q66458485) are the Proceedings of the Joint Statistical Meetings (Q6269403).

Shouldn't the two be linked? If yes, what's the preferred method? (Currently they are not.)

Thanks, DavidMCEddy (talk) 19:15, 31 August 2019 (UTC)

Secondarily, JSM proceedings (Q66458485) identifies its "publisher" as American Statistical Association (Q467040), but complains that "This statement has some potential issues." How can that complaint be removed? (The American Statistical Association is a professional society that publishes several professional journals as well as many other publications.) Thanks, DavidMCEddy (talk) 19:21, 31 August 2019 (UTC)

Closed Airport Property

Q1930822 uses P582 to indicate the end time of the airport. Q1436575 uses P576, but P576 says to use P730 for buildings. One of these should probably be changed to match the others since it represents the same underlying concept. Which is accurate? Charliefish (talk) 20:18, 28 August 2019 (UTC)

I think we should start a vote for this so -- 06:11, 29 August 2019 (UTC)
What about date of official closure (P3999)? Sjoerd de Bruin (talk) 09:45, 29 August 2019 (UTC)
Good idea, listed too. --Liuxinyu970226 (talk) 15:59, 30 August 2019 (UTC)
Let's start with understanding what do you mean by "end time of the airport"? Closing hour (if not working 24h/day)? End of functioning as airport (and turning into something else)? Demolishing the building? --Infovarius (talk) 20:05, 29 August 2019 (UTC)
@Infovarius: But now it looks like some users are very like using one property just because his friends are using that, so the vote is indeed needed here. @Swpb, Şêr, Aestrum, Arlo Barnes, Geertivp:@Toghrul Rahimli, Horcrux, Papuass, Sannita, Izno:@Trade, Sturban, Jérémy-Günther-Heinz Jähnick, ديفيد عادل وهبة خليل 2, Robert.Allen:@Naudefj, Manu1400, Schofför, Murma174: maybe these users should help us with a decision? --Liuxinyu970226 (talk) 04:10, 2 September 2019 (UTC)
Thanks, but I'm not sure I've edited an airport entity. Of course, more specific properties are often better (although generic ones allow large-scale analysis, they can be different to interpret without dedicated qualifiers to clarify). Arlo Barnes (talk) 02:24, 3 September 2019 (UTC)

end time (P582)

  •   Support Good property. -- 06:11, 29 August 2019 (UTC)
  •   Oppose, @ P582 should never be a main statement, who are using this property as so? --Liuxinyu970226 (talk) 04:11, 2 September 2019 (UTC)
  •   Support it is the best property and the main. Another solution will be to use significant event (P793) and have a statement "closing of the airport". Jérémy-Günther-Heinz Jähnick (talk) 06:12, 2 September 2019 (UTC)
    @Jérémy-Günther-Heinz Jähnick: Why? At least why do you think P582 can be used as main statement? --Liuxinyu970226 (talk) 21:51, 2 September 2019 (UTC)
    significant event (P793) permits to have a very precise statement, here for example the end of use of an airport that is different from a destruction or even the start of construction of towers in the site of an airport. The main problem we can have is the lack of translations for these items (and because you are a chinese speaker, you have a big idea of this problem). In my opinion it is the biggest problem in Wikidata, by the past I propose we give a list of the 5000 more used items and more crucial to local Wikipédias, in the idea to obtain a minimum of translations, considering a minimum of items translated permits to offer a basis to future translators, but this proposal was systematicaly ignored. end time (P582) is a "wide" property, so we naturally use it when we must indicate an end. Not everybody know all specific properties, for example two days ago I discover date of official closure (P3999). Jérémy-Günther-Heinz Jähnick (talk) 08:51, 4 September 2019 (UTC)
  •   Oppose - This is too generic. Geertivp (talk) 23:05, 5 September 2019 (UTC)

dissolved, abolished or demolished (P576)

This property helps mapping all abandonned/demolished airports
PREFIX wdno: <>
PREFIX rdf: <>
SELECT DISTINCT ?airport ?airportLabel ?pays ?paysLabel (SAMPLE(?coord) AS ?coord) ?ICAO ?IATA (SAMPLE(?Local_Code) AS ?Local_Code) WHERE {
  ?airport (wdt:P31/wdt:P279*) wd:Q62447.
   ?airport wdt:P576 _:b2.                                 # remove items with P576 (dissolved, abolished or demolished) as a main property
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q695850. }               # exclude military airports
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q1311670. }              # exclude railways stations
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q7373622. }              # exclude Royal Air Force
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q502074. }               # exclude héliports
  MINUS { ?airport wdt:P31/wdt:P279* wd:Q2265915.}                 # exclude vol à voile/gliders
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q44665966. }             # exclude  airports being build
  MINUS { ?airport (wdt:P31/wdt:P279*) wd:Q782667. }               # exclude motorways where plane can land
  MINUS { ?airport a wdno:P17.}                                    # exclude airports that aren't physically based in a country
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en,fr,[AUTO_LANGUAGE]". }
  OPTIONAL { ?airport wdt:P625  ?coord. }
  OPTIONAL { ?airport wdt:P239  ?ICAO. }
  OPTIONAL { ?airport wdt:P238  ?IATA. }
  optional { ?airport wdt:P17   ?pays.}
  OPTIONAL { ?airport wdt:P240  ?Local_Code. }
  OPTIONAL { ?airport wdt:P5699 ?Local_Code. }
  OPTIONAL { ?airport wdt:P5746 ?Local_Code. }
  OPTIONAL { ?airport wdt:P5851 ?Local_Code. }
  OPTIONAL { ?airport wdt:P6120 ?Local_Code. }
GROUP BY ?airport ?airportLabel ?ICAO ?IATA ?pays ?paysLabel
ORDER BY ?paysLabel ?coord}
Try it!
Bouzinac (talk) 13:15, 2 September 2019 (UTC)
  •   Support If we are talking about an airport actually closing down and ceasing to function as an airport, then this is the obvious property given how it's used with other organisations. MartinPoulter (talk) 10:35, 5 September 2019 (UTC)
  •   Oppose - This is rather used for organisations, or institutions; not for physical infrastructure. Geertivp (talk) 23:05, 5 September 2019 (UTC)

service retirement (P730)

  •   Oppose Bad property. -- 06:11, 29 August 2019 (UTC)
  •   Support - Could be used as well... Geertivp (talk) 23:05, 5 September 2019 (UTC)

date of official closure (P3999)

  •   Support - This is the exact property, because the airport buildings are (possibly) physically still there, possibly having another function, but the airport itself is not active any more... Geertivp (talk) 23:05, 5 September 2019 (UTC)

significant event (P793)

  •   Oppose - This is not telling what has happened; it would need another qualifier. Geertivp (talk) 23:05, 5 September 2019 (UTC)