Wikidata:Property proposal/Archive/22

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

number of spans

Descriptionnumber of spans a bridge has
Data typeNumber (not available yet)
Domainbridge (Q12280)
Allowed values1 to 1000000
ExampleEast Bloomsburg Bridge (Q15133614) --> 6
Proposed by--Jakob (talk)

Skyscraper Center ID

Descriptionbuilding ID as used on the CTBUH's
Data typeString
Domainplace (buildings)
ExampleTour First (Q1509747) => 993
Sourcesearch on
Proposed byDanrok (talk) 14:18, 1 February 2014 (UTC)
This appears to be a good reliable source for building facts and statistics. See About the CTBUH. Danrok (talk) 14:18, 1 February 2014 (UTC)
Danrok, do you have a dataset that you want to import?--Micru (talk) 21:57, 8 March 2014 (UTC)
I don't have data to import. But, it's possible that they might supply some. Danrok (talk) 22:13, 8 March 2014 (UTC)
I was asking because I didn't see any data to import in any wikipedia, but if they provide the data, then I support it.--Micru (talk) 15:42, 9 March 2014 (UTC)
  Support. Frequently used in Wikipedia (see search added above). ----- Jura 10:26, 10 May 2014 (UTC)
  Support --LydiaPintscher (talk) 22:37, 10 May 2014 (UTC)

China railway station telecode

   Done: no label (P1306) (Talk and documentation)
DescriptionA three-letter code for railway stations used in China railway system
Data typeString
Template parameter"电报码" in zh:template:Infobox China railway station
ExampleShanghai railway station (Q529728) => SHH
Proposed byLiangent (talk)

Migrate data in Property:P296. See Wikidata:Properties for deletion and above: this is another coding schema used in the same railway system. Liangent (talk) 08:12, 19 October 2013 (UTC)

  Support --A.Bernhard (talk) 09:10, 23 October 2013 (UTC)
  Oppose until consensus is reached at Wikidata:Properties_for_deletions#Bahnhofscode_.28P296.29 --Pasleim (talk) 22:00, 29 October 2013 (UTC)
  SupportIvan A. Krestinin (talk) 16:52, 11 November 2013 (UTC)
  Support. Needed for infobox ----- Jura 10:26, 10 May 2014 (UTC)


DescriptionOpenStreetMap tagging schema (a Key:key or Tag:key=value)
Data typestring => URL's to link to OSM-Wiki and OSM-Taginfo-invalid datatype (not in Module:i18n/datatype)
Domaine.g. lighthouse (Q39715) (classes of geographical objects in Wikidata)
Allowed valuese.g. "Tag:man_made=lighthouse" Tagging schema in OpenStreetMap
ExampleExample above would generate two links: and .
On a higher level are Keys like that I would connect with railway (Q22667)
Robot and gadget jobsNo.
Proposed byKolossos (talk)

I'm trying also on a low level of objects to link between OpenStreetMap and Wikipedia/Wikidata (see project WIWOSM), but with this proposed property I try to link between projects on the higher level of classes. These object classes are represented in Openstreetmap by the tagging schema. I would try for each entry in "Map_Features"-site (see source) to find an item in Wikidata. So I see a usage for this property only for some hundreds of items. If it's not matching exactly, I would not fill this property. At some points we will have discussion, because it is definitly not easy in each case. So we would learn where both projects matching good and where not. Perhaps we see that the structure in OpenStreetMap is for some point better than ours. Benefit can be in both directions Wikidata would get some map-icons and OpenStreetMap could profit from the translations in Wikidata. Later, tools can use this links so that we could compare e.g. all lighthouses in OpenStreetMap with all objects that are instances of class lighthouse (Q39715) and so on.

After approval I would need some help for the template that generates the URL's. I would also try to add this connection of both projects in taginfo so that we would have this link also in the backward direction. --Kolossos (talk) 19:19, 12 April 2014 (UTC)

  Support. Sounds to be useful. Where do you plan to place the links apart from one of them at the item page using MediaWiki:Gadget-Preview.js? Isn't there yet a connection at least from one to the other? Do you want to have both at the item talk page? Regards, --Marsupium (talk) 12:59, 13 April 2014 (UTC)
  Comment Surely the link to the OSM-Wiki is more important, because it's closer to the heart of OSM. This link will go to the English description page, and on each of these pages is a small link to Taginfo. Also Taginfo has links to the OSM-Wiki in different languages. So if you and others are thinking two links are to confusing, I can live with one link to the Wiki. Taginfo is more technical and provides statistics, links to Overpass-Turbo and JOSM. --Kolossos (talk) 16:15, 13 April 2014 (UTC)
  Support. After disussing some questions I understand, that this fits with the possibility easy to make maps out of Wikidata. I can use this also here in east germany, poland and czech via [1]. Conny (talk) 19:21, 14 April 2014 (UTC).
  Support I also discussed this with Kolossos and got convinced very fast. This will be a very powerful link to show geographical items on maps and much more. As said this property will be used on few high-level items, e.g. bridge (Q12280). Every item that's an instance of (P31) bridge (Q12280) will then be linked to the corresponding OSM tag. --Thiemo Mättig (WMDE) 16:49, 17 April 2014 (UTC)
  Weak support. I think this should be implemented, but in a cleaner way than what's proposed above, e.g. with property name "OpenStreetMap tag" and without including the "tag:" prefix in the values, as that would be redundant. Also, it would be better if the values were not free-form text (i.e. strings), but an item instead (not sure that's the proper term within Wikibase's data model); That way the data model would be cleaner since tags can only be added if they're already presented in the Wikidata system. We could also include a threshold for adding such tags to the list of OSM tags in the data model, something like a minimum of X usages as counted by Taginfo. This approach would help consolidate the tags (since adding a new tag isn't left to a user's sole discretion as happens in OSM), and would also make it easier to manipulate geodata without doing a lot of string parsing. --Waldir (talk) 15:22, 26 April 2014 (UTC)
  Comment The "tag:" is not redundant, because there exit also some rare cases of "key:" e.g. Key:admin_level. We should have the flexibility to support theses cases. A cleaner way would be in my eyes to split "key" and "value" in two separate fields of one property, but this seems not possible. The tagging in OSM is OSM-specific, so I see no link to Wikidata-items, e.g. not everyone would combine "prisons" in the category "amenity"[2]. Also for object classes with small number of items I see a small risk of miss-using this feature. It will be a lot of work to support only entries of the map feature page. --Kolossos (talk) 16:41, 29 April 2014 (UTC)

  Done as OpenStreetMap tag or key (P1282). Notification - Kolossos, Marsupium, Conny, Thiemo Mättig (WMDE), Waldir.--Micru (talk) 09:54, 3 May 2014 (UTC)

first written record

Descriptionthe year or exact date when the item was first mentioned in written records
Data typePoint in time
Template parameter"year_of_first_mention" in en:Template:Infobox German location
Domainplaces, persons from ancient or medieval times, institutions etc.
ExampleMunich (Q1726) => 1158
Sourceexternal reference
Proposed bySlomox (talk)

Many populated places were founded before writing became common in that part of the world. So we don't know when the places were founded, but at some time they suddenly appear in written records. Munich (Q1726) was first mentioned in the Augsburger Schied (Q760561) from 1158. The relevant information is the time, so the property should be of type "time" and the exact record (Augsburger Schied) can be optionally stored as a qualifier.

Besides the most obvious use case, populated places, this property could be used for many other items as well. For some people we don't know their birth and death dates, but we know that they appeared in records at some time and vanished from them later. The same applies for buildings, institutions or even whole people (like the Chauci (Q168549) being mentioned in 12 BC for the first time). Slomox (talk) 10:47, 5 February 2014 (UTC)

Maybe "time of first record"? Their existence isn't always written e.g. hieroglyphics. --Izno (talk) 16:22, 5 February 2014 (UTC)
Good idea. Not sure what should be the extent of this property though… Note sure it's a good idea to mix it with a floruit (Q36424).
Izno : hieroglyphs are a writing system (a strange one when you don't know it but still).
Cdlt, VIGNERON (talk) 12:10, 7 February 2014 (UTC)
  Support. Even if it's already done .. ----- Jura 10:26, 10 May 2014 (UTC)

PSGC identifier

   Done: no label (P1228) (Talk and documentation)
DescriptionPhilippine Standard Geographic Code: identifier for administrative divisions in the Philippines
Data typeString
Template parameternone
Domainadministrative territorial entity of the Philippines (Q1350310)
Allowed values9-digit number string
ExampleManila (Q1461) => 133900000 [3]
Format and edit filter validationnone
SourcePhilippine Standard Geographic Code home page
Robot and gadget jobsnone
Proposed byseav (talk)

This is similar to the other administrative identifier properties that are already existing for other countries. —seav (talk) 23:14, 20 March 2014 (UTC)

  Support -- Lavallen (talk) 06:16, 21 March 2014 (UTC)
  Support --Balû (talk) 07:02, 28 March 2014 (UTC)

Monument identifier Cologne

   Not done
DescriptionIdentifier for monuments in cologne (about 9000)
Data typeZeichenkette-invalid datatype (not in Module:i18n/datatype)
Template parameterKeine Infobox aber viele Listen (z.B. de:Liste der Baudenkmäler im Kölner Stadtteil Altstadt-Nord)
DomainDenkmäler in Köln/Monuments in Cologne
Allowed values+d
ExampleCologne Cathedral (Q4176) = "911"
Proposed by--Tobias1984 (talk) 17:17, 5 February 2014 (UTC)
  •   Support --Tobias1984 (talk) 17:17, 5 February 2014 (UTC)
  •   Oppose  Comment See #list of monuments 2. A property could be introduced for monument identifiers in North Rhine-Westphalia in general qualified by the specific "Denkmalliste" rather than for each municipality individually. If that should not be supported, one may use the general property catalog code (P528) instead.--Leit (talk) 19:03, 5 February 2014 (UTC) I do not oppose this property in general but if to be created it should be part of a general solution for cultural/architectural heritage monuments in North Rhine-Westphalia.--Leit (talk) 19:24, 5 February 2014 (UTC)
  •   Support Raymond (talk) 21:38, 5 February 2014 (UTC)
  • Slight   Oppose as by User:Leit's argument above: In the long run we would need literally thousands of those (BTW mutually exclusive, complementary) properties and I don't see how anyone could even select them all in order to make use of them in a larger scale than just picking some. (We'd have to classify them or invent some kind of virtual or meta-property just to maintain them). Of course any of them could be linked to a different database (Duisburg might be a better example than cologne since they have indivdidually selectable entries) and distinct databases (wikidata-)traditionally suggest distinct properties. -- Gymel (talk) 23:05, 5 February 2014 (UTC)
  • @Tobias1984, Leit, Raymond, Gymel:   Not done No consensus and discussion stagnated three months ago. --Jakob (talk) 12:06, 14 May 2014 (UTC)

full list of instances

   Not done
Descriptionlist of these items
Data typeItem
Template parametervarious, often in parentheses after an item
Domainanything with a long list of incidents or examples or incumbents
Allowed valueswikipedia list
Sourcesame source that is used for current incumbent
Proposed byArctic.gnome (talk)
  •   Support as nominator -- We need some kind of inverse of is a list of (P360). A page like pope (Q19546) really needs some way to link to list of popes (Q40483) given how intimately connected the two items are. --Arctic.gnome (talk) 20:07, 7 December 2013 (UTC)
    Sounds like redundant data.   Oppose. --Yair rand (talk) 00:28, 9 December 2013 (UTC)
    Redundant to what? If there is an existing property to link an item to the list of that item, I'll just use it instead. --Arctic.gnome (talk) 00:42, 9 December 2013 (UTC)
    Redundant to is a list of (P360), which would be the inverse, as you mentioned above. --Yair rand (talk) 03:12, 9 December 2013 (UTC)
    Why is that bad? Don't we already allow lots of inverse properties, like father & child, artist & discography, edition & edition of, encodes & encoded by, topic's main category & category's main topic, succeeded by & proceeded by? If we have a property that only goes in one direction, you wouldn't be able to find the pair of the second item unless you searched through "what links here". --Arctic.gnome (talk) 03:47, 9 December 2013 (UTC)
  • {s}} I like the first use the best. --Jakob (talk) 15:56, 26 February 2014 (UTC)
  •   Support just case #1.--Micru (talk) 12:20, 29 April 2014 (UTC)
  •   Oppose}Changed my mind after experimenting with lists. If it is a "full list" the scope is too reduced, and if it is just the inverse of "is a list of", then it should be named differently.--Micru (talk) 13:55, 6 May 2014 (UTC)
  •   Strong oppose „full list”? A query does this job better. --Succu (talk) 21:51, 5 May 2014 (UTC)
  •   Oppose per Succu. Emw (talk) 18:52, 10 May 2014 (UTC)
  •   Comment Hi, as Wikidata relies on open-world assumption (Q851949)      , we need a way to say that we got the whole list of instances in the database. Of course retrieving the instances list by a query is possible, but closing the list, as in saying there is only and will ever only have this precise list needs to be indicated. By the number of instances and the number of results in the query this is possible, or by a sequence of followed by (P156) terminated with followed by (P156) we got another possibility, but we must still keep this problem in mind. This property is another possible way. TomT0m (talk) 19:16, 10 May 2014 (UTC)

  Not done Substantial opposition. Use the other option (followed by:"no value")--Micru (talk) 10:08, 17 May 2014 (UTC)

claimed name

   Not done
Descriptionsee the example below.
Data typeMultilingual text (not available yet)
ExampleSenkaku Islands (Q29460) <claimed name> 钓鱼岛及其附属岛屿 (zh-hans)/Diaoyu Islands (en)/釣魚島およびその付属島嶼(ja) <claimed by> People's Republic of China (Q148), 釣魚台列嶼 (zh-hant/ja)/Diaoyutai Islands (en) <claimed by> Taiwan (Q865), Senkaku Islands (en)/尖閣諸島 (ja/zh) <claimed by> Japan (Q17)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:46, 25 March 2014 (UTC)

  Not done Datatype not available, re-open discussion when the mono-lingual data type can be used.--Micru (talk) 06:48, 30 April 2014 (UTC)
@Micru: "Datatype not available" is a strange reason to deny a property. A lot of properties has been approved before the datatype was available.
But I think it is a good idea to wait with the discussion of properties like this until we know how other name-properties will be used. -- Innocent bystander (talk) (The user previously known as Lavallen) 08:27, 1 May 2014 (UTC)

proved by

   Done: proved by (P1318) (Talk and documentation)
Descriptioncountry, organization, person which proved (to demonstrate that something is true or viable / to give proof for) sth. isually used as qualifier.
Data typeItem
ExamplePythagorean theorem (Q11518) => Pythagoras (Q10261)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:46, 25 March 2014 (UTC)

@GZWDer: I am testing a new property assessment method, according to it:
  • previous or requested use in MW projects: 1 (not that I know of, but wp might be interested)
  • fulfills a structural WD/WP/task force need: 5 (needed to represent scientific/math proofs)
  • data availability for importing or generated value: 2 (data available, manual input)
  • easy to import or easy to understand: 3 (easy to understand, no bot jobs)
  • plans to re-use data: 1 (not that I know of, but wp might be interested)
  • property ownership: 3 (implicated user, no information given about re-use plans)
Result: 15pts   weak support. Recommendation to improve rating: the property seems interesting, but the proposer should put effort in making it known or integrated into infoboxes or a task force.--Micru (talk) 07:02, 30 April 2014 (UTC)
  Oppose use stated in (P248). --Succu (talk) 21:45, 5 May 2014 (UTC)
Wrong. stated in (P248) will not be qualifiable, and won't point all the time to the original publication. For example it does not fit with the Pythagoran Theorem, as ... stated in ... what ? A book that said he proved the theorem ? This does not work.
  Support I already did this proposition in natural sciences, if I can remember well. TomT0m (talk) 19:03, 18 May 2014 (UTC)

topic consists of


Motivation. GZWDer (talk) 15:46, 25 March 2014 (UTC)

  Support Needed to organize items that are part of another item.--Micru (talk) 07:35, 30 April 2014 (UTC) Can be queried from the inverse property--Micru (talk) 10:17, 17 May 2014 (UTC)
How would this be used? Would human history (Q200325) "consist of" history of Africa (Q149813), history of Europe (Q7787), modern history (Q3281534), and pretty much every other item on history? Would all inverses of <subclass of> also fit this property? Or non-inverses, for that matter? What does it mean for a "topic" to "consist of" something other item/topic? This seems rather unusable. --Yair rand (talk) 14:44, 30 April 2014 (UTC)
I often used part of (P361) for sciences ou of feelings. For example for Gödel's incompleteness theorems (Q200787)      wrt. mathematical logic (Q1166618)     . Clearly, if a science or a field of science is a collection of knowledge, then those theorems are a part of it, so
⟨ Godel first incompleteness theroem ⟩ part of (P361)   ⟨ mathematical logic (Q1166618)      ⟩
. Another way to express this would be to create a class mathematical logics theorem with
⟨ Godel first incompleteness theroem ⟩ instance of (P31)   ⟨ mathematical logics theorem ⟩
. I don't think those two ways are enough, so I   Weak oppose this property. TomT0m (talk) 14:56, 30 April 2014 (UTC)
Yair rand, TomT0m: Reviewing the label, I realize that it is not the best one. What about "(secondary) approaches to this topic"? Think of the relatioship between water (Q283)=>properties of water (Q6478447).--Micru (talk) 16:14, 30 April 2014 (UTC)
Or maybe a shorter name for both properties could be "facet"/"facet of".--Micru (talk) 17:29, 30 April 2014 (UTC)

  Not done Inconclusive support. Use only the inverse property (facet of (P1269)), propose again if there are more uses.--Micru (talk) 10:17, 17 May 2014 (UTC)

facet of

   Done: facet of (P1269) (Talk and documentation)
Descriptionthe reverse property of these 3 requests above.
Data typeItem
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:46, 25 March 2014 (UTC)

  Support. We need a way to link sub pages to the main page. Filceolaire (talk) 13:52, 14 April 2014 (UTC)
  Support, but with a different name: "part of topic".--Micru (talk) 07:21, 30 April 2014 (UTC)
  • "part of topic" was not very clear, experimenting now with "facet of".--Micru (talk) 09:43, 1 May 2014 (UTC)

topic of WikiProject

   Not done
Descriptiontopic of WikiProject (aka Expeditions in Wikivoyage)
Data typeItem
ExampleWikiProject United States (Q10816993) => United States of America (Q30)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:54, 25 March 2014 (UTC)

  Not done--Micru (talk) 07:29, 30 April 2014 (UTC)


   Not done
DescriptionThe reverse property of that above.
Data typeItem
ExampleUnited States of America (Q30) => WikiProject United States (Q10816993)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:54, 25 March 2014 (UTC)

  •   Support --Jakob (talk) 12:00, 26 March 2014 (UTC)
  •   Oppose I don't like the mix between data and organizational stuff. Better use a model to put in the talk page. Snipre (talk) 11:18, 3 April 2014 (UTC)
  •   Oppose. --Yair rand (talk) 01:06, 12 May 2014 (UTC)

  Not done Substantial opposition.--Micru (talk) 10:12, 17 May 2014 (UTC)

non-free data available at

   Not done
DescriptionURL where we can look for a non-free data
Data typeURL
ExampleReggiolo (Q952) population (P1082) somevalue <non-free data available at> <related URL>
Robot and gadget jobsYes
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 07:33, 12 April 2014 (UTC)

  •   Support - A while back I thought about proposing something like this for linking to non-free images, but this idea is better. --Jakob (talk) 11:47, 12 April 2014 (UTC)
  Support. Since we can't copy the data, linking to it is the next best thing. Mushroom (talk) 18:23, 12 April 2014 (UTC)
  Weak oppose It looks like a poor solution. -- Innocent bystander (talk) (The user previously known as Lavallen) 09:05, 19 April 2014 (UTC)
  • Interesting. I   Support support the general idea, I   Weak oppose am not sure I like to use SomeValue for that. It looks like an interesting workaround. Undecided. --Denny (talk) 19:23, 21 April 2014 (UTC)
  Weak oppose We have the data or not. In other case just put directly the information in WP. It is just a patch to solve a licence problem and in that case better bypass WD and put all the data in WP which respects the minimal requirements for most not CC0 databases. But as no licence change is planned, why not. Snipre (talk) 14:03, 28 April 2014 (UTC)
  Support. Filceolaire (talk) 21:09, 1 May 2014 (UTC)
  Oppose I see no sense in this property. --Succu (talk) 21:52, 1 May 2014 (UTC)

  Not done No consensus.--Micru (talk) 10:20, 17 May 2014 (UTC)


   Done: floruit (P1317) (Talk and documentation)
DescriptionThe date on/during which the subject was active. "Floruit" (abbreivated as "fl.") is commonly used when, for example, an author's birth and death dates are unknown but we do know when they were publishing (see floruit (Q36424) for reference).
Data typePoint in time
Domainhuman (Q5) mainly, although there appears to be some case for using it for other things
Allowed valuesTime/date
ExampleWilliam Baldwin (Q8004982) => fl. 1547; Samuel Austin (Q7410810) => fl. 1629
SourceAuthority controls and other biographical/bibliographical sources contain this information.
Proposed byAdamBMorgan (talk)

One of four properties I'm proposing to handle approximate dates. Wikisource will need something like this for several authors, Commons probably needs it for some creators and Wikipedia for some biographies. It can be used for other things, in the same way as the standard term, for when those things have "flourished". - AdamBMorgan (talk) 21:30, 16 April 2014 (UTC)

I'm a bit uneasy about this. The two examples you give each refer to a single year from which I am guessing that these authors are known for a single publication in which case why not just use notable work (P800) to link to that work? Other cases will need a start and end date for Floruit. It probably won't do any harm though and it will be on a limited enough number of items that it can be changed later if need be. Filceolaire (talk) 21:34, 17 April 2014 (UTC)
The examples were just the first I found. There are more on, for example, WorldCat or the Dictionary of National Biography, I just needed some that already had data items. Many do have a range to show when they are known to be active. "Floruit" is pretty standard for this sort of this; it's the only to date some authors and creators. - AdamBMorgan (talk) 01:51, 19 April 2014 (UTC)
  Support Existing data, good for unknown authors.--Micru (talk) 07:31, 30 April 2014 (UTC)
@AdamBMorgan:: Template:Infobox person on enwp has the parameter "years active" which I saw in use on the infobox for David Duchovny. This would seem to be a better name for this property except that it should be two properties - "Years active - from' and 'Years active - end' with 'Floruit' as an alias for both of these. With those changes I would support this. Filceolaire (talk) 15:18, 2 May 2014 (UTC)
  Question Isn't it a duplicate of something like
⟨ Annakin Sky Walker ⟩ occupation (P106)   ⟨ Jedi ⟩
start time (P580)   ⟨ date of birth ⟩
end time (P582)   ⟨ Darth Vador apparition ⟩
 ? TomT0m (talk) 16:16, 2 May 2014 (UTC)
  Comment Some databases - like International Plant Names Index (Q922063) use floruit (Q36424) in the sense of „to appear” (example) because they do rely on a first paper. Often it's simple to provide more accurate informations (in this case see his Curriculum Vitae). --Succu (talk) 19:35, 5 May 2014 (UTC)

earliest date

   Done: earliest date (P1319) (Talk and documentation)
DescriptionQualifier marking the earliest possible date of the statement where the actual date is unknown (ie. the lower bound of a specified period). Similar to start time (P580) but that isn't quite right for this usage.
Data typePoint in time
Domainhuman (Q5), event (Q1656682)
Allowed valuesAs standard of time/date properties.
ExampleChristopher Columbus (Q7322) => date of birth (P569): Unknown value => Earliest date: 31 October 1450; Isaac Asimov (Q34981) => date of birth (P569): Unknown value => Earliest date: 4 October 1919
SourceStandard authority control references.
Proposed byAdamBMorgan (talk)

A qualifier to help handle unknown and uncertain dates. Where the official date is "between X and Y", this will be X. Personally, I intend to use it for authors: for example, set date of birth (P569) to unknown value then use this and another qualifier to set out the bounds of the period in which the author was born (rather than just leave it unknown or set to a generic decade or century). It could be used for other events with uncertain dates, however. - AdamBMorgan (talk) 21:41, 16 April 2014 (UTC)


   Not done
DescriptionApproximate date qualifier. Standard term for "about this date" which could be used where the actual value is unknown value.
Data typePoint in time
Domainhuman (Q5), event (Q1656682)
Allowed valuesAs standard of time/date properties.
ExampleMarcus Licinius Crassus (Q175121) => date of birth (P569): Unknown value => Circa: 115 BC; Urban II (Q30578) => date of birth (P569): Unknown value => Circa: 1042
SourceStandard authority control references.
Proposed byAdamBMorgan (talk) 22:09, 16 April 2014 (UTC)

Another qualifier, the last of four proposals to help handle uncertain dates. Circa is a common term (see circa (Q5727902) for reference) but one not currently supported by Wikidata to the best of my knowledge. With this qualifier, any date could be set to unknown value and the "circa" qualifier used for an clear approximation. This will be needed to fully support Wikisource and it will also be needed for Commons (creator biths/deaths and creation dates) and Wikipedia articles. - AdamBMorgan (talk) 22:09, 16 April 2014 (UTC)

  Oppose. There is already provision for dates to have an uncertainty which can be set to a month, a year, a decade, a century, a millenium etc. using the current UI. Filceolaire (talk) 21:21, 17 April 2014 (UTC)
True, but that's rather clumsy. The academic use of "circa" doesn't come with a defined degree of error but it's usually based on some research to narrow it down to something at least a little more specific than a decade (which would be the most likely of the existing options). We would be losing some precision if this information is not recorded. Selection by decade also has a problem at the borderlines: for example, "c. 1855" is straightforward enough but is "c. 1850" in the 1840s or the 1850s? The actual date could be in either decade; the user making that choice increases the potential error in the data held. (As a qualifier, it could also actually be paired with a date set to a decade, rather than the "unknown value" as described.) - AdamBMorgan (talk) 01:45, 19 April 2014 (UTC)
The data structure and API actually allow you to specify a date like "between 1846 and 1855" or "between Apirl 18 and 23 1956". All that is needed is a bit of user interface code to allow dates like this to be entered directly. It seems there is no feature request for this on Bugzilla yet[4], go ahead and add one! -- Duesentrieb (talk) 19:48, 30 April 2014 (UTC)
Duesentrieb: Wouldn't be that Bugzilla61909?--Micru (talk) 12:41, 1 May 2014 (UTC)
See Wikidata:Property proposal/Archive/17#circa.--GZWDer (talk) 12:22, 29 April 2014 (UTC)

  Not done WD can manage this on its own, it might need an agreement about how to translate the "c." into ranges and how to represent it.--Micru (talk) 12:41, 1 May 2014 (UTC)

element of

   Not done
Descriptionused for single items in contrast to subclass of (P279) and part of (P361) where instance of (P31) conflicts with classical usage
Data typeItem
Domainchemical element (Q11344) and others to be identified
Allowed valuestype of linked items (Q template or text), list or range of allowed values, string pattern...
Examplecopper (Q753) => group 11 (Q185870) and period 4 (Q239825) and d-block (Q214503)
Proposed bygangLeri aka לערי ריינהארט (talk)

Wikidata:Project chat#chemical_element relates to some unsolved mappings of chemical elements.
The property might be used in similar hierarchical systems. gangLeri לערי ריינהארט (talk) 13:45, 30 April 2014 (UTC)

לערי ריינהארט: Is this not very similar to the proposal "in list" above?--Micru (talk) 13:39, 30 April 2014 (UTC)
There was an edit conflict. Please dive me some time to read that proposal. לערי ריינהארט (talk) 13:45, 30 April 2014 (UTC)
It is not exactly the same. It is a general set relation. לערי ריינהארט (talk) 14:06, 30 April 2014 (UTC)
  •   Oppose similar to instance of (P31). This is set membership versus class membership. An atom is an instance of the atom concepts, we just have to define the classes of atoms and to class those classes themselves. TomT0m (talk) 13:51, 30 April 2014 (UTC)
  •   Oppose. group 11 (Q185870) and period 4 (Q239825) and d-block (Q214503) are all classes (or sets) each of which has Copper as instance of (P31) that class. Each of these classes are of course a subclass of 'element'. Filceolaire (talk) 21:55, 1 May 2014 (UTC)
  •   Oppose. Exactly redundant with instance of (P31), which has the semantics of 'element of' from set theory (more on that in my reply to gangLeri here). That said, subclass of (P279) would be the better property to use for the examples in the proposal, not P31. Emw (talk) 22:42, 4 May 2014 (UTC)
  •   Oppose. Already discussed in the chemistry propertiy proposals section. Snipre (talk) 13:45, 14 May 2014 (UTC)

statement disputed by

Descriptionqualifier to link to a source that is disputing this statement
Data typeItem
Allowed valuessources
ExampleWaltz No. 17 in E-flat major, Op. posth. (Q16747520) <composer> Frédéric Chopin (Q1268) // <statement disputed by> Chomiński catalog (Q16749680)
Format and edit filter validationno filter
Sourceit links to the source itself
Proposed byMicru (talk)

Motivation. References are always supposed to support the claim, but this qualifier is needed to represent counter-claims, when another source just says that the statement is dubious.--Micru (talk) 08:57, 7 May 2014 (UTC)

  • The idea sounds interesting but I am not sure I understand how you propose to implement it. I guess we should have both the source that initially made the claim and the source that disputed it in the "references" section, like this ? --Zolo (talk) 07:34, 8 May 2014 (UTC)
    Zolo, yes that would be it. The normal reference and the critical reference will be both in the references section. This qualifier will point out to the one that is being critical with the statement.--Micru (talk) 08:51, 8 May 2014 (UTC)
  •   Support. Great idea. See also said to be the same as (P460). Emw (talk) 17:07, 10 May 2014 (UTC)
  •   Support, this seems a good way to support disputed information.
  •   Oppose. Mixing sources with content is problematic. We need a better solution than just grafting the sources for or against a statement onto the statements themselves. I'd even prefer using this as a plain source property than having it as a qualifier. --Yair rand (talk) 01:05, 12 May 2014 (UTC)
    Yair rand, could you please give an example of what a better solution would look like? And by using it as a "source property" do you mean having a property called "relationship of source with statement:[dispute, doubt]" to be used in the "references" section?--Micru (talk) 11:10, 14 May 2014 (UTC)
  • @Micru, Zolo, TomT0m, Emw, Yair rand:   Done --Jakob (talk) 12:10, 14 May 2014 (UTC)
    Jakob Why have you created the property when there are still questions open?--Micru (talk) 12:59, 14 May 2014 (UTC)
    There was sufficient support to create it. Also, it seems that Yair rand is calling for an overhaul of the whole sourcing system, which is not the scope of this discussion. And using it as a source property as xe suggested could work. --Jakob (talk) 01:47, 15 May 2014 (UTC)
    Jakob, thanks for your explanation, probably it was a good decision. Let's use it and we'll see how well it goes.--Micru (talk) 08:28, 16 May 2014 (UTC)
@Micru:, I think that what Yair rand meant is that this property should be used in the reference section, which is not the same thing as the qualifiers. I agree, and I think you do to, as you supported my example this (it's just that you mixed up the words, sources are not qualifiers). --Zolo (talk) 10:09, 18 May 2014 (UTC)
@Zolo: Not sure if I understood correctly your example. Please check the disputed statement in Waltz No. 17 in E-flat major, Op. posth. (Q16747520) [5] and tell me if that is what you had in mind.--Micru (talk) 15:13, 18 May 2014 (UTC)
@Micru: no actually I had this in mind. I think it makes more sense: this is a statement about what the source says, not really a qualifier of the statement. It is sort of the inverse of p248 (p248: source says X, p1310: source says not X. When transcluded in Wikipedia, I would imagine it in a footnote like: "source: value of p238, for a contrarian view, see value of p1310. ". --Zolo (talk) 15:29, 18 May 2014 (UTC)
@Zolo: The only problem I see with this is that it is not really visible...and it is something you really want to convey clearly. It would be also nice if we could instead qualify the statement with a more generic property that would allow us to add different nuances. For instance calling this property "type of statement" with values "criticism", "uncertainty", etc. That way perhaps we would achieve the needed separation between information/source, plus conveying a different meaning in a visible way, but we would have to add the statement twice with the same value, one of them qualified with "type of statement:criticism". What do you think of it?--Micru (talk) 16:28, 18 May 2014 (UTC)
@Micru: on the face of it, it seems attractive to have "type of statement" as a qualifier in addition to "disputed by" in the source, but it makes the structure more complicated. If you want to say "composed by Chopin according to X but not to Y" you have to specify that "type of statement = criticism" only applies to source B and not to source A. And if you want to put "type of statement" as a qualifier rather than in the references, it seems that you need to have 2 "composed by Chopin" statements (1) "composed by Chopin source: Kobylańska Katalog" 2) "composed by Chopin type of statement "criticism", source Chomiński catalog"). That said, it may be ok to do it that way. Something similar seems to be required for other qualifiers (if you want to say: composed by Chopin determination method (P459) musical analysis, source A and composed by Chopin determination method (P459), handwriting analysis, source B, you may also need two different statements).--Zolo (talk) 20:28, 18 May 2014 (UTC)
Yes, that was what I meant. Using the property in the references section makes for a clearer division between sources of statements and the statements themselves. There are some clear problems with this, but I think they're less severe than the problems posed by putting source data in as qualifiers.
It also occurs to me that in certain cases, one could just reframe "X says the value of Y is not Z" as "X says the value of Y is (unspecified, maybe "unknown") but definitely not (qualifier) Z", by using a separate statement. --Yair rand (talk) 00:21, 19 May 2014 (UTC)
Zolo, Yair rand: If we change the name of this property to "type of statement" (to be used as Zolo described), then this property can also be used to capture uncertainty as Random knowledge donator requested on the project chat. Any concern about it?--Micru (talk) 11:14, 19 May 2014 (UTC)
Yes, if we're going to touch the roots of Wikidata, we need a clear plan, usecases and a RfC. These are complex questions. TomT0m (talk) 11:47, 19 May 2014 (UTC)
Thinking some more about it "type of statement" does not sound like such a good idea. If X shows that Y it is not by Chopin", then if X is the most authoritative source, we would have something like "composer: Chopin, type of statement: refutation, rank:preferred". But rank = preferred means that it will be the value retrieved by default. And if the client does not analyze all qualifiers (which would be a very stringent requirement), which means that it is likely to get "composer: Chopin" while it should really be "composer: not Chopin". So I'd say that we should stick to "disputed by" in the source (in this case: "composer: Chopin, source(stated in Z-disputed by Y), rank: deprecated. We also have the issue that 3 ranks may not be enough and that ranks conflate relevance and reliability, but if we are displeased with that we should try to have a built-in solution (split into relevance rank and reliability rank, and/or add levels). Trying to get around that using qualifiers appears to make the structure rather confuse. --Zolo (talk) 17:58, 19 May 2014 (UTC)

date of first historic record

   Not done
Descriptiondate of an entity's first historic record
Data typePoint in time
Domainhistoric noble families (family (Q8436)); historic persons (human (Q5)) that no birth date is known of but that appear in historic records; towns whose date of foundation is unknown, more or less every geographical location (geographic location (Q2221906)) and basically any kind of thing
Allowed valuesany date
ExampleAbensberg (Q319056) => 1065, Römhild (Q535758) => 800, pizza (Q177) => 997, violin (Q8355) => 1525 with some qualifier like "term" => "vyollon" as well as => 1535 with "term" => "violino"
Proposed byRandom knowledge donator (talk)

  Not done Redundant.--Micru (talk) 10:13, 17 May 2014 (UTC)

Participant alias Candidate, Competitor

   Not done
Descriptionperson or team that participated in a tournament, league, election or event.
Data typeItem
Domainevents such as sports competitions and elections
Allowed valuespersons, sportsteams
Example2013–14 La Liga (Q6086248)=>Real Madrid CF (Q8682); 2008 presidential election (Q4612418)=>Barack Obama (Q76)
Format and edit filter validation.
Sourceexternal reference, Wikipedia list article (either infobox or source)
Robot and gadget jobs.
Proposed byFilceolaire (talk)

property for listing the competitors in a competition. Filceolaire (talk) 15:23, 14 April 2014 (UTC)

  Oppose. Redundant to participant (P710). Mushroom (talk) 19:53, 14 April 2014 (UTC)
  Oppose. Proposal withdrawn per Mushroom. Filceolaire (talk) 21:59, 1 May 2014 (UTC)
  •   Not done Withdrawn by proposer. --Jakob (talk) 23:00, 1 May 2014 (UTC)

Central Bank

   Done: central bank (P1304) (Talk and documentation)
DescriptionCentral bank for a country
Data typeItem
Allowed valuesName of the Central Bank
Proposed byMcnabber091 (talk) 19:06, 14 April 2014 (UTC)
  Support, I do not see any other simple, reliable way to get that. --Zolo (talk) 07:52, 26 April 2014 (UTC)
  SupportWylve (talk) 06:33, 11 May 2014 (UTC)


Descriptionidentifier in the Dictionnaire du Jura
Data typeString
Template parameterfr:Modèle:DIJU
Domainperson, place
Allowed valuessee sample
ExamplePhilippe Robert: 3822

Makes it easier to reference it. ---- Jura 06:36, 12 April 2014 (UTC)

  Comment Template used only 23x, no article for Dictionnaire du Jura. --Kolja21 (talk) 20:33, 12 April 2014 (UTC)
The template is recent. There are about 6000 entries in this encyclopedia. --- Jura 00:35, 13 April 2014 (UTC)
  Support -- SERutherford (talk) 21:11, 26 April 2014 (UTC)
  Support --Micru (talk) 16:55, 30 April 2014 (UTC)

DNB editions

   Done: DNB editions (P1292) (Talk and documentation)
DescriptionID for the items in the catalogue of the German National Library (Q27302) (DNB). Single edition of a book, not to be confused with authority file GND ID (P227).
Data typeString
Template parameterde:Vorlage:DNB; DNB @ de:Vorlage:BibISBN
Domainversion, edition, or translation (Q3331189)
Allowed valuesnumbers
ExampleDer Herr Präsident (Q15220473) => 450155102
Proposed byKolja21 (talk)

Discussion about importing the template de:Vorlage:BibISBN, see de:Vorlage Diskussion:BibISBN#DNB-Eintrag. --Kolja21 (talk) 22:57, 26 April 2014 (UTC)

  Support --Micru (talk) 16:55, 30 April 2014 (UTC)
  Support -- SERutherford (talk) 22:16, 2 May 2014 (UTC)


DescriptionID on the Gran Enciclopèdia Catalana (Q2664168)
Data typeString
Template parameterused on Template:Cite GREC (Q8218905) mainly on ca-wp to source over 6000 articles.
Allowed values7 numbers, format:èdies/gran-enciclopèdia-catalana/EC-GEC-$1.xml
ExampleWikipedia (Q52) => 0283878
Format and edit filter validation(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
Robot and gadget jobsimport from ca-wp, or extract form viaf, import from template
Proposed byMicru (talk)

Motivation. To better manage the registers and to allow re-use.--Micru (talk) 17:18, 30 April 2014 (UTC)

  Support--Pere prlpz (talk) 20:56, 30 April 2014 (UTC)
  Support - SERutherford (talk) 22:17, 2 May 2014 (UTC)
  Support - Kolja21 (talk) 01:05, 3 May 2014 (UTC)
  Support --- Jura 13:16, 3 May 2014 (UTC)

Copyright license

Withdrawn as a duplicate of Property:P275.

   Not done
Descriptionthe copyright license under which a work is available
Data typeItem
Template parameter"license" in en:Template:Infobox journal
Domainarticle (Q191067), book (Q571), work (Q386724)
Allowed valuesfor a start, instances of Creative Commons licenses (Creative Commons license (Q284742)), plus perhaps CC0 (Q6938433) and public domain (Q19652).
ExampleEcological guild evolution and the discovery of the world's smallest vertebrate (Q15567682): Creative Commons Attribution (Q6905323)
Format and edit filter validationas per allowed values
Sourcetypically the URL at which the item is hosted; some dedicated services have the information for some subgroups of relevant items (e.g. CrossRef for scholarly articles)
Robot and gadget jobsFor scholarly articles licensed compatibly with CC BY-SA, WikiProject Open Access is planning to set up a bot to handle licensing information
Proposed byDaniel Mietchen (talk)

I think we need licensing information on Wikidata but I am not entirely sure how to approach this best, so I am posting here to get the discussion going. As for data type, it may be useful to be more specific (e.g. CC BY 4.0 - or perhaps even the language- and country-specific deeds - instead of Creative Commons Attribution (Q6905323)). As mentioned in the allowed values, it would make sense to include things like CC0 or PD here, but both are technically not licenses, so perhaps the property would have to be named differently. In principle, it would be desirable to extend this beyond CC licenses, so I have opted not to put CC into the property name, but I can see that it would also make sense (particularly now that no good repository of other licenses exists) to make this property more specific by naming it CC license or open license or some such. Daniel Mietchen (talk) 14:47, 9 May 2014 (UTC)

  Support (Conflict of interest declaration: I am working for pay for Wikiproject Open Access). This property will allow "signalling" of licenses - putting icons next to citations - across wikis, which is the aim of our project. The benefit of signalling, is know what citations you will be able to read, or be "paywalled" from, before you click. Plus, having the licenses visible will promote the awareness of open licenses to readers. Maximilianklein (talk) 15:58, 9 May 2014 (UTC)
  Question Daniel, Max: What is the difference between this proposal and the existing copyright license (P275)?--Micru (talk) 17:17, 9 May 2014 (UTC)
Thanks for spotting that, Micru - I simply wasn't aware of copyright license (P275). It seems that there is no difference that would be practically relevant now (although CC0/PD technically do not fall under licenses). So I am withdrawing this proposal. Not sure how to mark that in the template - guidance appreciated. --Daniel Mietchen (talk) 18:41, 9 May 2014 (UTC)
No problem! Already marked as   Not done--Micru (talk) 21:31, 9 May 2014 (UTC)

Blood Type

   Not done
Descriptionsee Wikidata:Property proposal/Archive/10#Blood Type (en) / 血液型 (ja) and Wikidata:Project chat#Blood type property. ONLY BE USED WITH A RELIABLE SOURCE.
Data typeItem
Allowed valuesA,B,AB,O
ExampleAkira Ishida (Q418736)=>O (source in jawiki)
SourceMust be with reliable sources in case of being against the private data policy of a lot of countries about medical data
Proposed byGZWDer (talk)

@Napoleon.tan, Snipre, Tobias1984, Presentime, Pere prlpz: please comment. GZWDer (talk) 16:00, 14 January 2014 (UTC)

  OpposeI see two reasons for not creating this property:
  • as already said, blood type is a medical data and medical data is particularly sensible. And even if they is a source I am not sure that everyone accepts the spreading of this kind of data. Just an example: blood type is depending on the parents blood type and by in some cases analyzing blood type within a family can show biological relations which are not corresponding to the legal relations (to be simple, in some cases this can show that a child is not the biological child of 2 parents). So if this can be public in some way this doesn't mean that everyone wants to see this kind of information largely published.
  • then the question is to know if this kind of information is easily accessible for a large number of persons. Blood type is not a characteristic of a person, this doesn't say anything about him. I know that in some countries blood type like Japan is important but if this concerns only the japanese Wikipedia perhaps this doesn't need to be on wikidata but can stay on the japanese infoboxes. If this information is only available for some thousands of persons and is interesting only few wp's, I think we can discuss about its usefulness at a worlwide scale. This is a personal point of view but a database with a lot of properties but not filled in a systematic way is not very interesting, I prefer few properties which are present in most items of the same class offering a real added value in term of query and analysis.
Can someone estimate the percentage of persons infoboxes in the jp:WP which have this property filled ? Snipre (talk) 18:05, 14 January 2014 (UTC)
  • Privacy: This property can be used based on reliable source. I would add only with reliable, publicly accessible, secondary sources. Then we won't make public any data which is not public already. Anyway, I think there will be a lot more people worried by us publishing their age, names of parents or sons, ancestry, criminal history, or other data about them, than people worried by us spreading their blood type. Furthermore, publishing it in wikidata or publishing it in Japanese Wikipedia makes no difference.
  • Utility: I must admit that being interested in celebrities'blood type sounds bizarre to me. Anyway, if some people from another culture or with different interests than me are interested in bloody type, I don't see any harm in helping them to get that information. In fact, since it is part of human knowledge published in reliable sources, giving access to it is in line with WMF goal.
  • "World-wide": If just one WMF project was going to use this information, you could be right that Wikidata won't be of great help in such job and information could be just stored locally in infoboxes. But if such information is going to be used by two projects (e.g. Japanese and Korean or Ainu wikipedias) or more, Wikidata can be a great help.
On the other hand, I don't expect most Wikipedias using this property. Therefore, having it for Japanese use won't make blood type appear in enwiki infoboxes.--Pere prlpz (talk) 20:17, 14 January 2014 (UTC)
I am not sure that only the limited information of these four types: "A,B,AB,O" would be enough to prove that somebody is or isn't related to each other. If you find a piece of clothes from Thirty Years' War (Q2487) in Germany and find A-type blood in it. Then it's suddenly more likely that the owner comes from Sweden, but it does not prove anything. (And this example comes from real life.) -- Lavallen (talk) 20:30, 14 January 2014 (UTC)
@Lavallen: Just have a look at this. This can't be used for all cases but they are some constrained cases. Snipre (talk) 21:00, 14 January 2014 (UTC)
@Lavallen: Blood types can be used to prove that a given person is not the offspring of a certain couple. This is the main reason to use this property only when blood type has been published by publicly accessible secondary reliable sources.--Pere prlpz (talk) 09:23, 15 January 2014 (UTC)
@Pere prlpz: Here we reach the problem: what is a reliable source for this kind of data ? Again, as it is a normally a medical data, the only person who can say the blood type is the person itself: there is no public document with this kind of data unlike for names, age or adresses. All other sources are problematic because you don't know how they get the data and if the person of interest gave his agreement. This is a problem for living persons, for dead persons you can assume that some documents are public. For me reliable sources are not enough, an agreement or a public declaration of the person of interest about this data is necessary in order to respect the private life of the persons. You can find a lot of data from people press (relationship, disease,...) but in some cases the persons laid some complaint against the newspapers: can we use this kind of information in that case even if the data is published ? Snipre (talk) 13:20, 15 January 2014 (UTC)
@Snipre: We have long worked definitions of what a reliable source is in all Wikipedias. We can treat reliable sources about blood type just in the same way as we treat reliable sources about any personal data. For instance, how do you know age, marital status or number of children of any notable people? You just check in books, newspapers or other reliable sources. If there is no information in any reliable source about a certain property for a given person, we just don't use that property for that person. For example, most of people with infoboxes in Wikipedias don't have information about spouses - it happens in Wikipedia in the same way as it happens in reliable sources. Very few people is expected to have information in "blood type" property, a lot less people than nearly any other personal data. This is not a problem. This is just what reliable sources do.
The only difference between blood type and other potentially private data (let's say, age) is the number of notable people with coverage by reliable sources and the number of readers and wikipedias interested in using it. Although, except for quantity, blood type is not different from age.--Pere prlpz (talk) 17:18, 15 January 2014 (UTC)
  Support I think this is reasonable. We can add it to people where it has relevance and can be sourced. It can also be used as a qualifier for medical topics. Certain diseases are tied to the blood type and we should be able to express that. --Tobias1984 (talk) 10:34, 15 January 2014 (UTC)
This information cannot be added if not the person him/herself has provided it. The person hirself is the only primary source except for medical journals, which we never have access to, and never should use. I'm afraid that the published blood-type therefor sometimes is fake, but I cannot see that as a problem. Wikidata will never be used to give medical treatment to these people. -- Lavallen (talk) 19:55, 15 January 2014 (UTC)
Look at this example: en:Blood type distribution by country. That is information we need to bring into the database. We are not going to add it to every person on Wikidata, but probably to a few people where it hase relevance. But we also need to store statistical information and the information how blood groups make certain diseases more likely. --Tobias1984 (talk) 11:41, 16 January 2014 (UTC)
  Support I think it is useful for Japanese voice actors or animation characters. --Konggaru (talk) 09:12, 28 January 2014 (UTC)
  Support per Konggaru. Amine projects would like to have it in infoboxes at least for characters. -- Vlsergey (talk) 11:22, 12 February 2014 (UTC)
  •   Support, useful data. If used in an item on a living person, use reliable sources (as always). --Jakob (talk) 21:54, 25 February 2014 (UTC)
  •   Oppose. First, this is medical data, and as such subject to additional legal protection in many jurisdictions. I anticipate needless complications arising from this if Wikidata collects such data. Second, it is not useful for the purposes of our projects, as the only real use of this data is in actual medical applications, for which no professional would rely on Wikidata. The apparent facetious/prurient/superstitious interest in this data as related to celebrities in Japan appears to me to be not within the scope of the purpose of our projects. Sandstein (talk) 20:47, 6 April 2014 (UTC)
  Support. I like to think no data is beyond the scope of our project, provided it is useful to someone and properly sourced. Blood type is big in Japan, the Japanese Wikipedia uses it, so I can't see why we shouldn't include it. As per the legal issues, by using this property we would be republishing data that is already public. If someone tried to defame someone on Wikidata using their blood type (I can't quite picture that, but let's imagine), the data would have no source so the claim would be removed. Mushroom (talk) 22:30, 6 April 2014 (UTC)
  Support because this information would be expected from some of our readers, and this is not a western-centric project. Snipre's concerns are valid, but they can be addressed by constraining this property to country of citizenship (P27) => Japan (Q17).--Underlying lk (talk) 22:05, 22 April 2014 (UTC)
  Oppose Sounds like Stasi 2.0. --Kolja21 (talk) 22:26, 7 May 2014 (UTC)

  Not done Conflictive property that might have legal repercusions since the servers are in US.--Micru (talk) 17:32, 9 May 2014 (UTC)

place of activity, place of work

   Done: no label (P1298) (Talk and documentation)
Descriptionthe place a person has worked (place of activity); not to be confused with employer (P108).
Data typeItem
Allowed valuesitem (geographical feature (Q618123))
ExampleRichard Grammel (Q2149608) => Gdańsk (Q1792) (Source: GND 118697099)
Sourceauthority files (and others)
Proposed byKolja21 (talk)

We have residence (P551) (Wohnsitz) but authority files often uses "place of work" (Wirkungsort) instead. See also: Wikidata:Requests for permissions/Bot/PLbot 4. Kolja21 (talk) 22:10, 7 April 2014 (UTC)

Perhaps a better translation would be "place of activity" and should not be confused with a formal connection as expressed by "affiliation" (I thought there was a property for this but I cannot find it at the moment). -- Gymel (talk) 10:41, 12 April 2014 (UTC)
Good idea. I've changed the label. AFAIK we don't have "affiliation". We only have employer (P108). --Kolja21 (talk) 15:08, 13 April 2014 (UTC)
  Support--Micru (talk) 13:01, 3 May 2014 (UTC)
  Support ----- Jura 11:58, 7 May 2014 (UTC)

Descriptionid in the Swiss Parliament's "Database concerning members of parliament since 1848" (more than 3400 persons)
Data typeString
Template parametersee fr:Modè,,,
Domainperson (members of the Swiss parliament)
Allowed valuesid (see sample)
ExampleJean Guinand: 99
Sourcesee fr:Modè and others

Allows easy referencing for parliamentarians. --- Jura 06:40, 12 April 2014 (UTC)

  Support. Mushroom (talk) 18:24, 12 April 2014 (UTC)
  Comment. Similar to US Congress Bio ID (P1157), MEP directory ID (P1186), Riksdagen person-ID (P1214). --- Jura 14:55, 4 May 2014 (UTC)
  Support -- Innocent bystander (talk) (The user previously known as Lavallen) 06:58, 12 May 2014 (UTC)


   Done: instrument (P1303) (Talk and documentation)
Descriptionthe instrument that a person plays (includes vocals)
Data typeItem
Template parameternone
Allowed valuesmusical instruments
ExampleRingo Starr (Q2632) => drum (Q11404)
Format and edit filter validation(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
Sourceexternal reference, Wikipedia list article (either infobox or source)
Robot and gadget jobsShould or are bots or gadgets doing any task with this? (Checking other properties for consistency, collecting data, etc.)
Proposed byFilceolaire (talk)

Just as occupation:sportsperson can have the property sport (P641):long jump (Q170737) so occupation musician can have this property.

It can also be used as a qualifier to performer (P175) on an album or a single item to specify what each performer did on that item just as character role (P453) can be used to qualify cast member (P161) on film items. Filceolaire (talk) 12:55, 14 April 2014 (UTC)

Place of origin (Switzerland)

Description"Lieu d'origine/Heimatort/Luogo d’origine" of a Swiss citizen. Not to be confused with place of birth, domicile or residence.
Data typeItem
DomainHumans of Swiss Nationality (any time), maybe some Austrians too.
Allowed valuesone or several municipalities of Switzerland (current or former)
ExampleGustave Ador: Geneva [6][7][8] and Vuiteboeuf [9]

. --- Jura 04:04, 15 April 2014 (UTC)

  Comment Standard element in biographies on Swiss people. --- Jura 18:24, 2 May 2014 (UTC)
See the references for the above sample. --- Jura 11:58, 7 May 2014 (UTC)
  Comment added labels in en/it/de. --- Jura 08:21, 10 May 2014 (UTC)
  Support A hereditary property for people. Used to be of higher legal importance than place of birth (Q1322263) and de:Bürgerort states that even currently Swiss Identity documents present this property instead of place of birth (Q1322263). According to de:Heimatrecht a similar concept applies to Austria between 1849 and 1939. -- Gymel (talk) 09:24, 10 May 2014 (UTC)
  Question However, to be strict, values must be of type civic community (Q1021096)? For most items these simultaneously are municipality of Switzerland (Q70208) as subclass of municipality (Q15284), but in other cases like Bern (Q70) and civic community of Bern (Q1014896) they are (since 17something or much later?) extremely distinct? -- Gymel (talk) 09:58, 10 May 2014 (UTC)
No, I think it should be territorial municipalities existing when the person lived. civic community (Q1021096) just group(s) everyone who has the same place of origin in some cantons. --- Jura 10:34, 10 May 2014 (UTC)
But considering Leo Merz (Q2800146), HLS: "Merz, Leo // geboren 13.7.1869 Thun, gestorben 21.9.1952 Bern, christkath., von Hägendorf und Bern". Isn't it that he died in Bern (Q70) but the proper item for his Heimatort Bern should rather be civic community of Bern (Q1014896)? -- Gymel (talk) 12:17, 10 May 2014 (UTC)
Values for Leo Merz (Q2800146) would be Hägendorf (Q67184) and Bern (Q70).
civic community of Bern (Q1014896) is a more like a corporation of citizens than a "place of origin". While he as alive, Leo Merz may be been entitled to participate in that corporation. Some of the civic community (Q1021096) limit benefits to citizens that are also resident in the municipality. --- Jura 12:51, 10 May 2014 (UTC)
But isn't it true that any person "heimatberechtigt" in Bern is member of the Burgergemeinde by birth and opting out is impossible except by changing the Herkunftsort to some other place (I'm aware of the fact that the incorporatedness of the Burgergemeinde Bern is somehow exceptional even for Switzerland but otherwise there is no difference)? -- Gymel (talk) 16:20, 10 May 2014 (UTC)
For people with "place of origin" = "Bern", this might be correct (or slightly more complicated), but in any case, the "place of origin" is "Bern" and not "Burgermeinde Bern". --- Jura 16:50, 10 May 2014 (UTC)

  Support. And for Leo Merz I would say "place of origin" = "Bern" but I'm not an expert on this. --Pasleim (talk) 17:33, 20 May 2014 (UTC)

Parent language family

   Not done

Makes property P133 (P133) obsolete. This property is like parent taxon (P171), but for languages. This property is needed as P133 is a broad category, while this proposed property gives a more precise family, where the family's item could use this property to become broder (such as English (Q1860) => Anglic (Q1346342) => Anglo-Frisian (Q5329170) => North Sea Germanic (Q30134) => West Germanic (Q26721) => Germanic (Q21200) => Indo-European (Q19860) ). Ebe123 (talk) 23:26, 25 February 2014 (UTC)

  •   Support Could be very useful. --Jakob (talk) 01:11, 3 March 2014 (UTC)
  •   Support --Micru (talk) 18:05, 3 March 2014 (UTC)
  •   Oppose Per TomT0m.--Micru (talk) 22:13, 4 March 2014 (UTC)
  Oppose, use subclass of (P279) with and items like . to know which level of the hierarchy we are in. I have already done it a few times probably, no need for a specific property. TomT0m (talk) 18:13, 3 March 2014 (UTC)

  Not done Per comments.--Micru (talk) 13:08, 1 May 2014 (UTC)

NLA party identifier

DescriptionNational Library of Australia party identifier
Data typeString
Domaingeographic location (Q2221906), organization (Q43229), human (Q5)
ExampleRichard Weyl (Q7329920) => 1095692; Isaac Aaron (Q15994254) => 1500864 & 1465747 (example of duplicate yet to be resolved); Andrew Barton Paterson (Q436699) =>; Commonwealth Scientific and Industrial Research Organisation (Q1117048) => 458251
Format and edit filter validation10 digit number
SourcePeople Australia has a Open Archives Initiative Protocol for Metadata Harvesting (Q2430433) interface to extract the information.
Proposed byJohn Vandenberg (talk)

The primary identifier allocated & used by Australia for people, organisations and places, including subjects with no relation to Australia. It is led by the National Library of Australia (Q623578), all other significant Australian libraries are members of the consortium, and they are integrating other datasets built by organisations, and is now a mandated part of research data at universities[10]. 'Trove' is the public website to access the data, 'People Australia' is a funded program to hold/build the dataset.

Regarding automation for population of this field, the NLA party ID exists on ~250 Wikipedia pages (see w:Special:LinkSearch/, however the People Australia records can be correlated with Wikipedia entries via identifiers that are common to both, such as Template:Cite Australian Dictionary of Biography (Q6703319). John Vandenberg (talk) 07:47, 6 April 2014 (UTC)

  Question John: How does this identifiers compares with Libraries Australia ID (P409)?--Micru (talk) 18:14, 30 April 2014 (UTC)
  • @billinghurst, Micru:, the simple explanation is that Libraries Australia ID (P409) is one of several identifier for authors/subjects of items held by libraries in Australia, and is managed using librarian practises without public use as a after-thought (it is like a national OCLC number), whereas the party id is a unique identifier for things known by any contributing member to the open consortium (which includes major datasets already built by large groups, but every Australian academic can also contribute individually) with Libraries Australia only being the primary dataset and the NLA acting as manager, and is managed using modern minting practises via modern protocols, with use by the general public as the primary objective. For NLA identifier 35411562, which is Andrew Barton Paterson (Q436699) the NLA party record is (as EAC 2004++ XML), where you can see 35411562 and many more identifiers listed, making this the primary entry point into Australian public datasets (esp. research, but other federal funded programs are also requiring grant recipients align themselves with NLA party IDs). John Vandenberg (talk) 20:18, 30 April 2014 (UTC)
    JV knows more about this stuff than me, and I will bow to his expertise in this area.  — billinghurst sDrewth 14:49, 6 May 2014 (UTC)

  Support, but we need a unique name like "Trove ID", "NLA Trove" etc. and we need to rename Libraries Australia ID (P409). --Kolja21 (talk) 00:23, 3 May 2014 (UTC)

I am not sure that renaming NLA is necessary, it aligns with VIAF usage, and our nomenclature. I would agree that this proposed property should not be that similar so it can be differentiated, and wonder whether it should be aligned with something like "PeopleAustralia-identifier".  — billinghurst sDrewth 14:49, 6 May 2014 (UTC)

'Trove' cant be in the name, as 'Trove' is a front-end website that uses many different IDs, and it does not mint these IDs. 'People Australia identifier' would be acceptable to start with, however this identifier is also for other types of entities such as CSIRO (see examples). This proposed identifier is called NLA party ID[11], or NLA Persistent Identifier. Property:P409 should probably be renamed, as it is actually the Australian National Bibliographic Database identifier (which was based on the 'Australian Name Authority File', which uses different IDs), and is served via the 'Kinetica Service'. It is common to call Libraries Australia ID (P409) the 'AuCNLKIN ID', where AuCNLKIN is the MARC Org code for 'Australian Commonwealth National Library Kinetica Service'. It is also common to call the old NLA identifier the 'Libraries Australia identifier'. Whatever we call them, I think we will need to add notes into the description to help the user pick the right property, and periodically run a bot that resolves the property values to detect if the wrong property has been used. John Vandenberg (talk) 03:02, 8 May 2014 (UTC)

  Done As NLA Trove people ID (P1315). Notification John Vandenberg, billinghurst, Kolja21. John, please add the aliases and documentation as needed.--Micru (talk) 10:34, 17 May 2014 (UTC)

EGAXA identifier

   Done: EGAXA ID (P1309) (Talk and documentation)
Descriptionidentifier issued by the Bibliotheca Alexandrina (Q501851)
Data typeString
Template parameterwill be used by ar:template:Authority control
Domainpersons (and organizations?)
Allowed valuesvtls (?) + numbers
ExampleNaguib Mahfouz (Q7176) = viaf:89803944 => vtls001414794 (or 001414794?)
SourceVIAF, library catalog
Proposed byKolja21 (talk)

Needed for Arabic Wikipedia (Q199700). (AFAIK till now the identifier is only searchable through VIAF. Any expert on Egypt knows more? Do we need the additive "vtls"?) --Kolja21 (talk) 14:32, 6 May 2014 (UTC)

  Support. "vtls" appears to be part of the number, like the "n" in LCCN numbers, rather than the name. - AdamBMorgan (talk) 21:18, 9 May 2014 (UTC)
  Support --LydiaPintscher (talk) 21:35, 10 May 2014 (UTC)
  Support ·addshore· talk to me! 11:10, 13 May 2014 (UTC)

in list

   Not done
Descriptionlist concerned by the statement / item
Data typeItem
Domainstatements like preceded by / followed by, items that are part of a list
Allowed valuesitems which are sequences
Format and edit filter validation(exemple : 7 chiffres peuvent être validés avec le filtre d'édition Special:AbuseFilter/17)
Robot and gadget jobsDevrait-il y avoir/Existe-t-il des robots ou des gadgets qui effectueront des tâches avec cela ? (Vérifiez les autres propriétés afin d'être cohérent, de collecter des données, etc.)
Proposed byTomT0m (talk)

@Zolo, Lavallen, FelixReimann, GZWDer, Jakec: It's that, or we use part of, I would prefer that. It apply to any kind of lists, the type of the list is a good information of the kind of sequence, so it seems generic enough to me ... TomT0m (talk) 22:27, 15 February 2014 (UTC)

So an item like "Astronomical naming conventions for minor planets" (I am not aware of any item like that today) would fit for example 4 Vesta (Q3030). That sounds good!
But I have mainly used P155/156 for the evolution of administrative division (see Ytterlännäs rural municipality (Q13604045)). Would communal reorganisation (Q1497011) fit for such items? - Then, I am not sure "part of" would fit for that purpose. We maybe need this new property, then? -- Lavallen (talk) 05:54, 16 February 2014 (UTC)
@Lavallen: I guess the most accurate list for that would link a geographical territory item, such as an item for an urban area with the local jusrisdiction in which he belongs, and that the law about the territorial entity is a qualifier or a source for the change of jurisdiction in this case ... I don't realise if it is a realistic solution in this case though.
For example if <urban area X> and <urban area Y> where two <village X/Y(resp)> that appeared to have been reorganized as a single unit <XY city>, we could have <urban area X jurisdictions> as the list item and
⟨ town XY ⟩ follows (P155)   ⟨ village X ⟩
in list Search ⟨ urban area X jurisdictions ⟩
⟨ town XY ⟩ follows (P155)   ⟨ village Y ⟩
in list Search ⟨ urban area Y jurisdictions ⟩
... TomT0m (talk) 16:21, 16 February 2014 (UTC)
These are actually municipalities (local political autonomic areas), not urban areas (who are mainly statistical units). (I do not know any good way to describe the succession of urban areas yet.) I do not completely understand what you mean with jurisdictions here. Municipalities of different kinds are of course jurisdictions in themselves, who are independent up to the central government, with one exception. "Municipalsamhällen" did not used to be completely independent to the municipalities they were connected. -- Lavallen (talk) 19:09, 16 February 2014 (UTC)
I think this amounts to the same thing as using this property as qualifier. I tend to like using qualifiers better as it is more concise (we do not have the rather useless repetition "instance of French cabinet" / "in list: French cabinets").
The solution you propose would probably easier to use in templates fr:Modèle:Insérer dynastie but I really don't like those templates :] And for other purposes, using "instance of French cabinet, succeeded by X" is more straightforward, as we avoid intermediate items like history of french governments (Q684321) that might sometimes be machine-confusing or human-confusing. --Zolo (talk) 19:10, 16 February 2014 (UTC)
@Zolo: This does not work if the item belongs to several sequences. For example, a serie episode belongs to a season and to the whole serial. There is several Pierre Laval government in the french Third Republic, which are identified by a number on (working in importing governments into Wikidata :) ) which does not follow immediately on the sequence of governments, there may be albums that are sequel of others in a discography but not chronologically on the same discography ... This may not be useful in all cases but I already seen several cases in which this can be used. TomT0m (talk) 19:20, 16 February 2014 (UTC)
@TomT0m:, that's true that element order may not always be obvious, but I do not understand how this property would improve things. If you want to say that Gouvernement Jacques Chirac (2) comes chronologically just after Laurent Fabius cabinet, but that as a "Jacques Chirac cabinet" it comes after Jacques Chirac 1, how would those two differ:
  1. instance of Cabinet of Francepreceded by Laurent Favius Cabinet / instance of Jacques Chirac (1) Cabinet preceded by Jacques Chirac 1
  2. preceded by Laurent Fabius Cabinet in sequence: Cabinet of France preceded by Jacques Chirac (1) Cabinet in sequence:Jacques Chirac Cabinet
-Zolo (talk) 22:23, 16 February 2014 (UTC)
@Zolo:OK you meant the solution to qualify the type claims. I don't really like these solutions as it seems a little weird for my computer scientisf self to do that, a type usually defines (or is defined by) the properties of its instances. Adding informations when we type something is OK for template type (that defines the concrete type actually) but it's not the case here. In particular, Jacques Chirac Cabinet is a subtype of Cabinet of France. Here in a sense we have two kinds of list, one who is attached to Cabinet of France and one who is attached to Jacques Chirac Cabinet, and we define them when we instanciate the classes, in a sense. Qualify the type claim is kind of an equivalent call a constructor with a parameter in object programming, a notion that absolutely does not exists here. OK, I'm beginning to sort out the node into my head :) here instance of (P31) is supposed to have a typing semantic, like in RDF/OWL. But using following / preceeded by we give it a sequence semantic (to the class), and classes have usually not a list semantic but a simple set of elements semantic. By expressing in separate statement it seems cleaner : we don't mix the two semantics. OWL Punning OWL Punning allows to treat an item as both a class and an individual, so there is no conceptual difficulties for him to be both a class and a list, but using the two semantics in the same claim seems not so clean, or even impossible in systems like OWL punning.
Of course some people in this project do not care of this kind of details or easily overcome by a little data processing when exporting datas (I understand that) but imho trying to be as clean as possible might not be useless (the cost is not very high, just a qualifier and one or two claims). I think Emw would back me up here (even if he would like to be even more aligned with OWL :) ) TomT0m (talk) 22:59, 16 February 2014 (UTC)
@TomT0m:. That seams to make sense. Still I am pretty sure I have seen cases where "preceded by" is not perfectly clear even when we know what sequence it is referring to. I'd like to find some of these before I make up my mind: not sure whether we should have special properties like "chronologically preceded by in sequence", or simply add statements to the sequnce item, like "instance of chronological sequence.--Zolo (talk) 07:26, 22 February 2014 (UTC)
So 'Star Wars A New Hope' would be in 2 lists, one with the sequence in which the films were issued, the other by internal chronology? That could work. See the discussion about Preceded by and followed by and the difference between a series of books/TV shows (previous not replaced) and a sequence of political office holders (each replaces the previous). Going to think some more before voting. Filceolaire (talk) 21:02, 24 February 2014 (UTC)
@Zolo, Filceolaire: Thinking about this a little more, I think it would be better if this property was used as a main property with precedeed/followed by as qualifier, for example
⟨ Toy Story 2 (Q187266)      ⟩ in list Search ⟨ Q2732470 ⟩
follows (P155)   ⟨ A Bug's Life (Q216153)      ⟩
followed by (P156)   ⟨ Monsters, Inc. (Q187726)      ⟩
and . Maybe with a rank numerical qualifier to put the position in the list ... TomT0m (talk) 23:11, 4 March 2014 (UTC)

  Wait This property seems redundant with queries, I would prefer to wait and see how queries are implemented before dedicating time to lists that maybe can be generated automatically.--Micru (talk) 12:38, 29 April 2014 (UTC)

No it's not redundant with queries. It's an added information, John might be both preceeded by Pierre as the head of its company and by Paul in its appartment ownership. You can't know which sequence is it just with follows (P155) alone. TomT0m (talk) 12:45, 29 April 2014 (UTC)
You might be able to query anything (appartment owner, head of company) and sort the results by "start/end date", or "date of creation", or "rank" or whatever. precedeed/followed by doesn't seem necessary in the examples you give.--Micru (talk) 12:57, 29 April 2014 (UTC)
Then you don't need follows (P155) and followed by (P156) anywhere. Yet they are very convenient to navigate beetween items. Redundancy is not so much a problem, just because we could do this another way does not mean it's bad. Plus this work if you don't know dates. TomT0m (talk) 15:06, 29 April 2014 (UTC)
Another feature is to query the first item member is a list item : you query the item which as no succeeded by or no value for the property in this list. TomT0m (talk) 15:12, 29 April 2014 (UTC)
@TomT0m: I have been using part of (P361), and for me it does the job. See example works by Chopin. You could add the "preceded/followed" as qualifiers of "part of" to have a better sorting. Works for you too?--Micru (talk) 14:04, 6 May 2014 (UTC)

  Not done Use part of (P361)->List with qualifiers "followed by" and "preceded by".--Micru (talk) 12:58, 23 May 2014 (UTC)

available data at

Descriptionurl that contains data relevant for this item
Data typeURL
Allowed valuesany
ExampleUS ->, with qualifier "license:CC0"
Robot and gadget jobsYes
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 07:33, 12 April 2014 (UTC)

  Support. Since we can't copy the data, linking to it is the next best thing. Mushroom (talk) 18:22, 12 April 2014 (UTC)
  Comment at least remove the non free in the name to be NPOV. There could be free datasets we will not import in Wikidata for a number of reasons, ne reason to restrict the range of this property. TomT0m (talk) 11:19, 2 May 2014 (UTC)
TomT0m: modified and   Support.--Micru (talk) 13:06, 23 May 2014 (UTC)

latest date

   Done: latest date (P1326) (Talk and documentation)
DescriptionQualifier marking the latest possible date of the statement where the actual date is unknown (ie. the upper bound of a specified period). Similar to end time (P582) but that isn't quite right for this usage.
Data typePoint in time
Domainhuman (Q5), event (Q1656682), intellectual work (Q15621286)
Allowed valuesAs standard of time/date properties.
ExampleChristopher Columbus (Q7322) => date of birth (P569): Unknown value => Latest date: 30 October 1451; Isaac Asimov (Q34981) => date of birth (P569): Unknown value => Latest date: 2 January 1920
SourceStandard authority control references.
Proposed byAdamBMorgan (talk) 21:56, 16 April 2014 (UTC)

Corresponds to the previous proposal, earliest time: Another qualifier to help handle unknown and uncertain dates. Where the official date is "between X and Y", this will be Y. For authors, set date of birth (P569) could be set to unknown value then "earliest date" and "latest date" to set out the bounds of the period in which the author was born (rather than just leave it unknown or set to a generic decade or century). As with "earliest date", it could be used for other events with somewhat vague dates. - AdamBMorgan (talk) 21:56, 16 April 2014 (UTC)

  •   Support. It could also be used for works, e.g a photograph that could not have been taken after a given date. Thryduulf (talk: local | en.wp | en.wikt) 12:11, 22 April 2014 (UTC)
  •   Oppose this and "earliest date". If I understand correctly, this will be built into the time datatype directly, so that indicating the earliest and latest times could be done without needing a separate statement. We just need to wait until the UI development is done. --Yair rand (talk) 02:03, 24 April 2014 (UTC)
  •   Support for now. It can me moved later on if there is an alternative system available. And besides since "earliest date" was created, it makes sense to either have both created or none.--Micru (talk) 13:09, 23 May 2014 (UTC)
  • @AdamBMorgan, Thryduulf, Yair rand, Micru:   Done --Jakob (talk) 16:34, 23 May 2014 (UTC)

Abbreviated form

   Not done
DescriptionThe general Abbreviation, initialism or acronym corresponding to the complete name (as opposed to P743 (P743) which is for stuff like New York City City → New York)
Data typeString
Template parameter"Abbreviation" in en:Template:Infobox journal or en:Template:Infobox constellation
Allowed valuesAlmost any
ExampleDoctor of Philosophy (Q752297) → Ph.D.
Sourceexternal reference, Wikipedia list article (either infobox or source)
Proposed byCirceus (talk)

Obvious case use. Circeus (talk) 15:10, 26 April 2014 (UTC)

  • This is the third time this has been proposed. See WD:Property_proposal/Archive/7#Code.2C_Acronym_or_Abbrevation and WD:Property_proposal/Archive/19#abbreviation.   Oppose, yet again. To quote Byrial: "This is not about the item, but about the names of the item. Even in the same language the item can have several synonymous names." --Yair rand (talk) 12:43, 30 April 2014 (UTC)
  •   'Support. @Yair rand: This is about the name of the item which is an important property of the item. Aliases can include all kinds of info including common mis-spellings. This property is for recording one type of names for the item. Alternatively we rename P513 (P513) to 'name and use it for all the different types of name with instance of (P31) as a qualifier for the type of each name (birth name, stage name, art name, alias, nom-de-guerre, nom-de-plume, etc.) Filceolaire (talk) 21:23, 1 May 2014 (UTC)
See the discussion at Property_for_deletions/Archive. Filceolaire (talk) 21:36, 1 May 2014 (UTC)
@Circeus: Mind to explain your „Obvious use case” a little bit further? As far as I know journals and constellations have well known (specialized) abbreviations. --Succu (talk) 21:47, 1 May 2014 (UTC)
It seems pretty obvious, and I don,t see why we'll want to have a bunch of super specific categories ("Authors of plant names work abbreviation", "International Astronomical Union constellation abbreviation"). This is the sort of things people will be adding because it is just obvious, and they will be shoving it in inappropriate properties, and alternative names just don't always cut it (besides, I don't see how this is NOT a useful property in general). Some acronyms (e.g. UNPROFOR) can probably use P743 (P743) (which I found used in Doctor of Philosophy (Q752297), hence the proposal), but I find that property to be annoyingly vague. Besides, an abbreviation can be an initialism in a language and an acronym in another, such as English UN vs. French ONU, only one of which seems to clearly fall under "short name" for me. Circeus (talk) 16:07, 2 May 2014 (UTC)
Thanks for your annotations. This it's simailar proposing a unstructured free text property, so I   Oppose. --Succu (talk) 21:01, 2 May 2014 (UTC)

  Not done Circeus, this is language-dependent information, consider to propose it again when the mono-lingual datatype is available.--Micru (talk) 13:14, 23 May 2014 (UTC)

Complete name

   Not done
DescriptionThe converse of the above "abbreviation of" because certain things (such as certain substances or proteins, people and organisations) are better noun by their abbreviated form. Could also be used for people to distinguish Björk (Q42455) or Madonna (Q1744) (using this) from Eminem (Q5608) (which uses P513 (P513))
Data typeString
Allowed valuesAlmost any
ExampleBjörk (Q42455) → Björk Guðmundsdóttir, UNICEF (Q740308) → United Nations International Children's Emergency Fund
Sourceexternal reference, Wikipedia list article (either infobox or source)
Proposed byCirceus (talk)

Converse of above. Circeus (talk) 15:10, 26 April 2014 (UTC)

  Not done--Micru (talk) 13:15, 23 May 2014 (UTC)


   Not done
Descriptionmandatory property for both piped titles at one of the WMF languages, value should be the highest possible hierarchical item; suggested also for items without instance of (P31), part of (P361), and subclass of (P279)
Data typeItem
Allowed valuesnot allowed: WMF types as: disambiguation pages, category pages, list pages, template pages
Examplesample items that would use that property, with proposed values; example: circle (Q17278) => mathematics (Q395)
Proposed bygangLeri aka לערי ריינהארט (talk)

The name can be understood in many languages; it is better then "term in". If no description is available one should add also a description as term in "foobar"; depending on the existence of the properties instance of (P31), part of (P361), and subclass of (P279) a more detailed description can be added.
I have seen a lot of part of (P361) statements which are used with the same scope. However mass addition of part of (P361) overwhelms the "tree" tool and makes it useless. It makes sense to replace these statements with the new property. לערי ריינהארט (talk) 13:27, 29 April 2014 (UTC)

One example of tree overview for earth science (Q8008) is Not sure if context will help here. לערי ריינהארט (talk) 02:56, 30 April 2014 (UTC)
context can help to increase "granularity" without interfering with the part of relations which should be used for major relations (as the parts of science, major research domains etc.).
context might be regarded as a lexicographic property do distinguish semantically different items. It is not part of the description, it is part of the item and it is machine readable. This may minimize the efforts in adding descriptions for thousand of items. 14:03, 30 April 2014 (UTC)
  •   Comment At first sight this seem exactly like a Mediawiki category, a thing that we exactly want to leave behind.
This is not the same as Mediawiki category. Search is one of the weakest points because only a small amount of items have a usefull description in English. context is a not related to a specific language. You may search for motion (physics) or motion (legal) or motion (democracy) or motion (geometry) if you are using the English interface but this is no help if you are using another language / another schrpt. לערי ריינהארט (talk) 23:17, 30 April 2014 (UTC)
  •   Weak oppose until a convincing usecase is found. TomT0m (talk) 14:13, 30 April 2014 (UTC)

example collection: stratigraphy (Q134783) is part of (P361) geology (Q1069) ; it is a major "term"
en:trap does not list trap with context volcanology/vulcanology. Please see en:Siberian Traps. A trap with that context is normally listed in the terminology/index. In Wikidata such an item might be difficult to find. לערי ריינהארט (talk) 15:12, 30 April 2014 (UTC)

tree via subclass of (P279) for physician who specialized (Q16031226) provides an useful overview while tree via part of (P361) for medicine (Q11190) is flooded by minor topics.
@Snipre I tried to use part of (P361) for the systematic about sciences and generally for field of work and used subclass of (P279) for occupations/professions. Regards gangLeri לערי ריינהארט (talk) 15:38, 30 April 2014 (UTC)

additional arguments: It happens that items are imported and will stay orphaned for a long time. It takes a lot of time to identify such unlinked items. @TomT0m: transuranium element (Q106844) was and stays orphaned while transactinide element (Q428778) is hardly to integrate in the main group / period chemical element structure based on the classical periodical table.
In order to identify all WD topics one needs to verify each language, each category and each page. Why would somebody do that? It is really a waste of time. I am still hunting orphaned pages from Norse mythiology ... לערי ריינהארט (talk) 22:50, 30 April 2014 (UTC)

I understand to connect the item to the graph, but I don't understand how it's good to connect with a meaningless property. It's just an opportunity to find a good way to model things or to create a relevant meaningful property. This would be a poor workaround who discourage the going further step and the submission to the problem to community. TomT0m (talk) 09:52, 1 May 2014 (UTC)


   Not done
DescriptionThe field (of study) related to an item. Intend to replace use of part of (P361) in graph (Q141488).
Data typeItem
Domainany term
Allowed valuesany subclass of academic discipline (Q11862829)
Examplegraph (Q141488)=>mathematics (Q395) and computer science (Q21198)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 10:09, 10 May 2014 (UTC)

  Weak oppose : I think we just remove the person reference in the field of work (P101) property to generalize it. Very similar of this property except for the domain. TomT0m (talk) 15:24, 10 May 2014 (UTC)
  Oppose. Per TomT0m. Emw (talk) 17:07, 10 May 2014 (UTC)

is a superclass of

   Not done
Descriptionsubclasses of the item; inverse of subclass of (P279)
Data typeItem
Domainanything with incidents
Allowed valuestype of subclasses
Examplesample items that would use that property, with proposed values; example:

geographic region (Q82794) => is a superclass of =>region of Italy (Q16110)
president (Q1255921) => is a superclass of =>province president (Q3911031)
mayor (Q30185) => is a superclass of =>mayor (Q670106)
mayor (Q30185) => is a superclass of =>Mayor of New York City (Q785304)
administrative territorial entity of Japan (Q1850442)=> is a superclass of =>municipality of Japan (Q1054813)

administrative territorial entity of Japan (Q1850442)=> is a superclass of =>special wards of Tokyo (Q308891)
Proposed byOursana (talk)

Motivation. Necessary to make the relations between the item classes clearOursana (talk) 23:11, 12 May 2014 (UTC)

  Oppose Reasonator can show pretty well the superclass tree as well of the subclasses of an item, so I think it's not necessary at that point. TomT0m (talk) 20:57, 14 May 2014 (UTC)
  Oppose The reverse relationship is in no need of explicit marking. Currently, has part or parts (P527) is usually the relevant property when this truly needs to be expressed (see e.g. Mascarene Islands (Q205695)). Circeus (talk) 15:54, 15 May 2014 (UTC)

  Not done--Micru (talk) 13:16, 23 May 2014 (UTC)

MusicBrainz instrument ID

Descriptioninstrument identifier per the MusicBrainz open music encyclopedia
Data typeString
Domainmusical instrument (Q34379)
Allowed values36 character Universally Unique Identifier
Exampleaccordion (Q79838) => bdf08ac2-b9c2-4391-85e5-9a7716bdd690
Format and edit filter validation[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}
Robot and gadget jobsUser:MineoBot can import them
Proposed byMineo (talk)

Mineo (talk) 11:08, 17 May 2014 (UTC)

  Support Most other MusicBrainz entity IDs are supported, and this seems like a good candidate to also support. I see no reason not to include this. (MB instruments already link to Wikidata.) --Freso (talk) 19:26, 21 May 2014 (UTC)
  Support --CatQuest, The Endavouring Cat (talk) 19:31, 21 May 2014 (UTC)
  Support--Micru (talk) 16:37, 23 May 2014 (UTC)

intrinsic property

   Not done
Descriptionproperty of a system or of a material itself or within (see: w:Intrinsic and extrinsic properties)
Data typeItem
Allowed valuesall
Exampleaddition (Q32043)=>associativity (Q177251), physical object (Q223557)=>density (Q29539)
Proposed byMicru (talk)
  • Needed to add more information about classes.--Micru (talk) 14:13, 23 May 2014 (UTC)
  •   Support curious where it may lead. --- Jura 05:16, 26 May 2014 (UTC)
  •   Strong support good idea. TomT0m (talk)
  Comment It could even be used to suggest properties for the class instances. Say density (Q29539) is linked to a property expressing this relationship beetween an object and a value for example. TomT0m (talk) 16:27, 26 May 2014 (UTC)
  •   Weak oppose. The 'physical object (Q223557)=>density (Q29539)' example suggests this property will be used to unwittingly construct an upper ontology that has lurking alignment problems with major upper ontologies outside Wikidata. A statement about an entity's intrinsic properties is not a suggestion; it formally entails that it has a (possibly unknown) value for the property.
For example, in Basic Formal Ontology 2.0, "photon" is a (distant) subclass of "material entity" and "object", which are analogous to Wikidata's "physical object" item. Photons have no mass, though, so we cannot say that photons are physical objects and also say 'physical object (Q223557)=>density (Q29539)' and also be aligned with BFO. BFO is the most widely used upper ontology in the natural sciences, so it would be advantageous to be aligned with that.
Beyond that note of caution, I would like to see some more background research on how formal ontologies in the wild (or even just in the literature) deal with intrinsicality. As the Stanford Encyclopedia of Philosophy notes in, there is wide disagreement on the precise definition of "intrinsic" and "extrinsic" beyond familiar examples from physics like "mass" and "density". And as far as I'm aware, no major ontology or ontology language supports this distinction.
In summary, I have two concerns. First, I'm concerned that this property and "extrinsic property" will exist be but under-researched and under-defined, and end up causing substantial community confusion like, say, member of (P463). Second, I am concerned that this property is far-reaching yet unconventional in the Semantic Web. Emw (talk) 01:51, 29 May 2014 (UTC)

  Not done Not ready yet. Needs more preparation.--Micru (talk) 11:29, 30 May 2014 (UTC)

extrinsic property

   Not done
Data typeItem
Allowed valuesall
Examplephysical object (Q223557)=>weight (Q25288)
Proposed byMicru (talk)
  • Needed to add more information about classes.--Micru (talk) 14:13, 23 May 2014 (UTC)
  Support goes with previous. --- Jura 05:16, 26 May 2014 (UTC)
  Comment not opposed per se., but it needs more thinking. I think we would need to add informations about the type of context needed. For example, weight (Q25288) depends on the gravity field the object is into. If we want to add it to an item, it would need to be qualified by on <Earth (Q2)> surface for example. TomT0m (talk) 16:23, 26 May 2014 (UTC)
talk:As a first step I have proposed "depends on", to define the "expected input variable" of the class. We can either use the same property as qualifier or think of a new one to "pass the value".--Micru (talk) 13:45, 27 May 2014 (UTC)

  Not done Not ready yet. Needs more preparation.--Micru (talk) 11:29, 30 May 2014 (UTC)

depends on

   Not done
Descriptioncondition(s) that affect the expression of this class
Data typeItem
Allowed valuesall
Exampleweight (Q25288) <depends on> gravitational field (Q558066)
Proposed by--Micru (talk) 13:36, 27 May 2014 (UTC)

  Not done Not ready yet. Needs more preparation.--Micru (talk) 11:29, 30 May 2014 (UTC)

number of seats

DescriptionThe number of seats that a house (camera) in a Parliament/Congress/similar has.
Data typeNumber (not available yet)
Template parameter"members" in en:Template:Infobox legislature
Domainlegislative house (Q10553309), upper house (Q637846), lower house (Q375928)
Allowed valuesNatural numbers
ExampleHouse of Lords (Q11007) => 779
Sourceen:Template:Infobox legislature
Robot and gadget jobsA bot could collect the data from articles like en:Seanad Éireann, en:National Assembly (South Korea), ...
Proposed byIt Is Me Here t / c

The point of this would be to have a central location for updating the sizes of cameras. This is particularly useful when these change, and the numbers are used at e.g. en:Template:Infobox political party/seats, itself typically embedded in en:Template:Infobox political party. It Is Me Here t / c 15:16, 30 May 2014 (UTC)

  •   Support --- Jura 09:40, 1 June 2014 (UTC)
  • Support SPQRobin (talk) 13:05, 8 June 2014 (UTC)
  Done SPQRobin (talk) 13:26, 8 June 2014 (UTC)

Freebase identifier (en)


Descriptionidentifier in the Helveticarchives database, the archive database of the Swiss National Library
Data typeString
Template parameterTemplate:Helveticarchives (Q16323070)
Domainpeople, organizations
Allowed valuessee sample
ExampleAlice Rivaz: 165123

Link to archives of materials of these persons. --- Jura 00:29, 13 April 2014 (UTC)

  Support Templates introduced in de:WP and fr:WP some days ago by an apparently institutional account. Thus reliable data with even some entitlement for completeness is to be expected. -- Gymel (talk) 10:05, 13 April 2014 (UTC)
  Support -- SERutherford (talk) 02:19, 18 April 2014 (UTC)
It used to be fine, now it does not work: VIAF as a code for Helveticarchives. The example given is not representative being a sub-section of the catalog, and the : "format as a regular expression", [1-9]\d* , with an external link ???? is just very confusing, to say the least. Please revert to something simple and straithgtforward. --DDupard (talk) 16:53, 14 April 2018 (UTC)

Shown with features

Descriptionsecondary features of something depicted. To be used as a qualifier of depicts (P180)
Data typeItem
ExampleOlympia (Q737062) , Victorine Meurent (Q271273) shown with features: nude, bracelet, supine position
Proposed byZolo (talk)

See Wikidata:WikiProject_Visual_arts/Questions#.22Depicts.22. That may sound a bit messy but subdviding it in many subproperties ("wearing", etc.) sounds worse to me. I think it should even incorporate color (P462), when used as a qualifier of depicts (P180), so that we do not need to deal with several properties. Any suggestion for a better label is welcome ;]. User:Zolo
Jane023 (talk) 08:50, 30 May 2013 (UTC)
User:Vincent Steenberg
Marsupium (talk) 13:46, 18 October 2013 (UTC)
GautierPoupeau (talk) 16:55, 9 January 2014 (UTC)
Multichill (talk) 19:13, 8 July 2014 (UTC)
Susannaanas (talk) 11:32, 12 August 2014 (UTC) I want to synchronize the handling of maps with this initiative
Mushroom (talk) 00:10, 24 August 2014 (UTC)
Jheald (talk) 17:09, 9 September 2014 (UTC)
Spinster (talk) 15:16, 12 September 2014 (UTC)
PKM (talk) 21:16, 8 October 2014 (UTC)
Vladimir Alexiev (talk) 17:12, 7 January 2015‎ (UTC)
--Oursana (talk) 13:16, 17 May 2022 (UTC) Sic19 (talk) 21:12, 19 February 2016 (UTC)
Wittylama (talk) 13:13, 22 February 2017 (UTC)
Armineaghayan (talk) 08:40, 10 March 2017 (UTC)
Musedata102 (talk) 20:27, 26 November 2019 (UTC) Hannolans (talk) 18:36, 16 April 2017 (UTC)
Zeroth (talk) 02:21, 4 June 2018 (UTC)
Infopetal (talk) 17:54, 9 August 2019 (UTC)
Karinanw (talk) 16:38, 24 March 2020‎ (UTC)
Ahc84 (talk) 17:38, 26 August 2020 (UTC)
Bitofdust (talk) 22:52, 26 March 2021 (UTC)
Mathieu Kappler
Zblace (talk) 07:22, 24 December 2021 (UTC)
Oursana (talk) 13:16, 17 May 2022 (UTC)
  Notified participants of WikiProject Visual arts Zolo (talk) 18:27, 14 January 2014 (UTC)

  Oppose for the moment at least. You did not fill the domain and range (allowed values) in the template, though you imply them:
  • The range sounds far to big with cloths, colors etc. I guess we should have a properties each for colors, cloths and so on. To me, subporperties sound better. Such a property would not have much meaning, used in qualifiers to describe paintings this might be enough, but I'd like to use more general solutions:
  • The domain is far to small. We can take the same property to describe a thing in reality and fiction/arts. We do not need a property "shown with features" in addition to "wearing", "colour" and such properties for real things and perhaps even an property "described with features" for literature!? It will probably be better to describe fictional and real things with the same properties.
And let us first of all see how other formal description frameworks for visual works handle this. Or do you yet know similar solutions? Regards, --Marsupium (talk) 13:47, 19 January 2014 (UTC)
@Marsupium:. I had not thought of literature, but I agree it may be useful there too. At to what exactly we should do, I cannot think of any practical alternative. If we want to go into subproperies, it might get really complicated.
The only website I know supporting something like this is Joconde, and it is just as unspecific as what I propose [12] . Other databases only use a few very general values for their "depict" properties, and this is what is suggested in CONA. But of course, they cannot be as specific as we can, because they do not have like us millions of items about notables people, places etc. ;). --Zolo (talk) 18:19, 15 February 2014 (UTC)
Thanks for pinging me! And thanks for your comment. I think that we have to work out another solution. Especially because of what you pointed out with your last phrase. But if you really need this property for any purpose I do not want to be an obstacle during my absence on WD ;). So I gladly recall my opposition on condition that we may later replace it by other properties and delete it if we agreed on an eventually better model for the description of objects in images. However, I'd probably prefer the use of properties like color (P462) (Cave! The colour of – say – an oil painting may be different from the colour of an object depicted by the corresponding position of the oil painting since the image may show the object in a special light!), "shown with other object" (superclass of wearing), busy with activity and the like.
Best regards, --Marsupium (talk) 14:53, 19 February 2014 (UTC)
P.S.: You might also examine the model of Iconclass (Q1502787): – though it is quite messy as far as I remember. I plan a more general examination on formal models for the description of images (CONA etc.) anyway, but I won't begin before spring and perhaps considerably later. --Marsupium (talk) 14:53, 19 February 2014 (UTC)
  Comment I am leaning towards 'wearing' instead with other properties that are more specific about what they refer to but can be used more generally for fictional characters and real world things. Filceolaire (talk) 04:55, 25 February 2014 (UTC)
  Comment As Filceolaire, I also prefer to have specific properties like: wearing, posture, cropping, etc.--Micru (talk) 17:25, 3 March 2014 (UTC)
@Micru, Filceolaire: I don't agree we need lots of properties for that. This seems to be the right level of genericity : we know it applies to some artistic representation, photography or artworks. There is a very big number of things in a very big number of situations. This thing/situation or properties are pairs of item value which can precisely define the subject, and the types of this item will gives us the nature of this feature. IMHO we got all the information we might want in a flexible way, we don't need a ton of specific properties for each cases, at least until we got a more precise idea of the problems to solve. TomT0m (talk)
  Support generic and usable solutions, large enough to be useful, even if we will have to sort out a little how it will be used. We need some usecases and showcase them. TomT0m (talk) 17:39, 3 March 2014 (UTC)
TomT0m: In that case I would suggest changing the label to something like "characteristics of depiction", the current one seems unclear.--Micru (talk) 17:45, 3 March 2014 (UTC)
If we do not know the things that shall be expressed, we do not need any properties at all, those things addressed here can be expressed with very few precise properties. I please everyone to take a look at the Iconclass classification: It presents such badly formalised descriptions. I'd be glad if we are able to do a better job here. --Marsupium (talk) 15:15, 20 April 2014 (UTC)
  Support Let's try it out.--Micru (talk) 14:32, 13 June 2014 (UTC)

Bibcode reference code

   Done: bibcode (P1300) (Talk and documentation)
Descriptionw:Bibcode identifier for the periodical
Data typeString
Domaincreative work
Allowed valuesvalid on periodical (Q1092563)
ExampleCelestial Mechanics and Dynamical Astronomy (Q2792544) => CeMDA
Sourcew:User:Bibcode Bot/Journal list
Robot and gadget jobsw:User:Bibcode Bot currently hosts this data at w:User:Bibcode Bot/Journal list.
Proposed byJohn Vandenberg (talk)

Bibcode is similar to CODEN (P1159), for a different discipline. It would be good for the bibcode bot to be able to use Wikidata as its data store, making it easier to add new values as journals change series or new journals arrive. This was proposed to be added to English Wikipedia at w:Template talk:Infobox journal#ADS bibcode. I suggested at Property_talk:P819#Use_on_journals that it might be appropriate to reuse that property for journals, but I think a separate property is more appropriate. John Vandenberg (talk) 14:59, 7 March 2014 (UTC)

  Comment sorry my missing answer here. What is the difference between ADS bibcode (P819) and this new property? --Paperoastro (talk) 20:53, 9 March 2014 (UTC)
Hi User:Paperoastro. This property will have a different set of values, it will be allowed on periodicals instead of articles, and it has a different format - CeMDA instead of YYYYCeMDA..... When people see a ADS bibcode (P819), they will expect it is an article which can be accessed using that bibcode. This proposal for a bibcode reference code will only uniquely identifies a periodical. John Vandenberg (talk) 23:44, 9 March 2014 (UTC)
I understand your proposal, but when I query here, the result is always in the format YYYYCeMDA... See for example this link. Where is the difference? --Paperoastro (talk) 22:19, 10 March 2014 (UTC)
The format of articles in the journal is YYYYCeMDA... The code for the journal is CeMDA. It is useful to have the journal code as a property if you wish to covert a textual citation into an article bibcode. e.g. if I have "Morais, M. H. M.; Namouni, F.; Retrograde resonance in the planar three-body problem, Celestial Mechanics and Dynamical Astronomy, Volume 117, Issue 4, pp.405-421", and I want to create a bibcode, then I need some way to look up "Celestial Mechanics and Dynamical Astronomy" and obtain CeMDA, so that I can create the string 2013CeMDA.117..405M. If we have this property, or someone suggests another property to store this string, I can lookup "Celestial Mechanics and Dynamical Astronomy" in wikidata, and obtain 'CeMDA' via a property. John Vandenberg (talk) 23:45, 10 March 2014 (UTC)
Ok, if I understand correctly, you would use publication date (P577), volume (P478), page(s) (P304), author (P50) and this new property to build the ADS bibcode. In this case, I
  Support this new property, but probably it is better to change the name, because this code (I suppose) exist only in ADS. --Paperoastro (talk) 14:18, 11 March 2014 (UTC) P.S.: a technical problem: exist a manner to know if the list of author in P50 is in the right order?
Offtopic, the 'C' in is not mandatory; also works. Regarding checking the order of authorship, I dont think we have any tool doing that yet, and the problem is much broader than bibcode. We could check our data against bibliographic databases like Crossref (Q5188229). John Vandenberg (talk) 05:38, 6 April 2014 (UTC)

is depicted, depicted in, represented by, inverse of depicts (P180)

   Done: depicted by (P1299) (Talk and documentation)
Descriptioninverse of depicts (P180) to add to the motif
Data typeItem
Template parameterHier den Infobox-Parameter einfügen, falls existent. Beispiel: "population" in en:template:infobox settlement
Domainany motifs: religious, historical event (Q1656682) events, landscapes, (fictional) persons depicted in a book, film, TV show, myth or any artwork etc.
Allowed valuescreative works, books of the bible
Exampleexample: Battle of Anghiari (Q1779008) => is depicted by The Battle of Anghiari (Q2045726), while inverse The Battle of Anghiari (Q2045726) uses depicts (P180) Battle of Anghiari (Q1779008); Daniel (Q171724) => Book of Daniel (Q80115)
Format and edit filter validation(Beispiel: eine siebenstellige Zahl kann mit dem Missbrauchsfilter 17 überprüft werden)
SourceExterne Referenzen, Listen in der Wikipedia (entweder Infobox oder Quelle)
Robot and gadget jobsFühren Bots oder Helferlein irgendwelche Tätigkeiten mit dieser Eigenschaft aus oder sollten sie solche Tätigkeiten ausführen? (etwa in dem sie andere Eigenschaften auf Konsistenz überprüfen, Daten sammeln etc.)
Proposed byOursana (talk)

Motivation. Oursana (talk) 10:27, 27 March 2014 (UTC)
Merged with same proposal of Fileolaire :Oursana (talk) 11:36, 3 April 2014 (UTC)
This is the inverse property of depicts (P180) for use on the items about each character in a creative work. It can have chapter (P792) and section, verse, paragraph, or clause (P958) as qualifiers for the chapter and verse in which the character is mentioned. Having this inverse property means we don't have use depicts (P180) to list every character in a work on the item for the work. Filceolaire (talk) 16:22, 30 March 2014 (UTC)

  Support, sounds useful. However, I do not get what is meant by ”Suggested values: (deprecated)“?? BTW: By translating the German description above one might open the possibility to join the discussion for others. --Marsupium (talk) 11:42, 18 April 2014 (UTC)
deprecated is simply kept from the template without further meaning, the description is also done in English--Oursana (talk) 19:32, 20 April 2014 (UTC)

SMDB identifier

DescriptionA little similar to IMDb ID (P345). Svensk mediedatabas ("SMDB") ("Swedish media database") contain entries of Swedish media. The database covers media stored at Kungliga biblioteket ("the National Library of Sweden"). It stores most Swedish media since 1979, but also some older stuff.
Data typeNumber (not available yet)
Domainfilm (Q11424) (film), single (Q134556) (single), album (Q482994) (album), specific broadcasts (TV-shows and radio shows) and "interactive multimedia".
Allowed values0-9
ExampleRallybrudar (Q10648678) => 002465142 (to link to, The Sign (Q638334) => 001504588, Look Sharp! (Q945464) => 001476627.
Sourcew:Svensk mediedatabas, w:sv:Svensk mediedatabas, w:sv:Mall:SMDb, about SMDB at
Robot and gadget jobsA bot could possibly import data for the some 100 pages on Swedish Wikipedia that uses a template that link to SMSB. There are also some 4000 pages on Swedish Wikipeida that mention the link (external links or as a source in a reference), but it might be a good idea to verify them one by one before importing to verify that the link points to an entry in the database that correspond to the title of the Wikipedia article. (Do we have an interface for that? It might come in handy in situations like these.)
Proposed byBensin (talk)

Motivation: A lot of Swedish media is covered by this database and I think it would be great to be able to link to entries in this database both to identify instances of creative works but also to use as a source for claims of who created a work (director, songwriter), who appeared in a film. --Bensin (talk) 21:06, 8 April 2014 (UTC)

Service discontinue at

   Not done
Descriptionsee example
Data typePoint in time
ExampleWindows XP (Q11248)=> April 8, 2014
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 13:44, 9 April 2014 (UTC)

  Not done Redundant with service retirement (P730).--Micru (talk) 12:57, 1 May 2014 (UTC)

public source code repository

Data typeURL
Template parameter?
Allowed valuesany url that is a public repository
ExampleMatplotlib (Q2985668) = ""
Format and edit filter validationunique value, ...
Robot and gadget jobsCould be parsed from Wikipedia articles.
Proposed byTobias1984 (talk)

Motivation: Make queries and links possible to where source code is kept. -Tobias1984 (talk) 15:57, 13 May 2014 (UTC)

Employer Identification Number

DescriptionEmployer Identification Number (Q2397748): The EIN is a unique nine-digit number assigned by the United States Internal Revenue Service (IRS) to business entities and non-profit organizations for the purposes of identification.
Data typeNumber (not available yet)
Template parameter"tax_id" in en:Template:Infobox non-profit
Domaincorporation (Q167037), nonprofit organization (Q163740) and/or company (Q783794)
Allowed valuesEINs are written in a nine digit series, for example XX-XXXXXXX.
ExampleGoogle (Q95) => 77-0493581 (see also Google's annual report (Q699735) here:
SourceInternal Revenue Service (Q973587) issues EINs. They are listed in annual reports submitted to the U.S. Securities and Exchange Commission (Q953944)
Robot and gadget jobsBots could populate a great deal of EINs for non-profits/corporations/companies.
Proposed bySERutherford (talk)

EINs would be an authoritative point of reference for specific U.S. non-profit/company/corporation profiles. SERutherford (talk) 01:05, 7 April 2014 (UTC)

  Question: Would this field have stability issues? Don't a lot of organizations change their EINs whenever they reorganize? This didn't used to be be common, but it has been since the 1980s. It seems that this number would change out from under an item regularly enough that it could become outdated. Also, there may be arguments about which corporation a wiki page is really about; wiki company article subjects often tend to actually follow a business operation, not a legal corporation, when the subject comes up. Perhaps neither of these are reasons to reject the property, though; but if it's approved, there should probably be documentation to the effect that EIN shouldn't be applied to items that relate to more than 1 historic corporation (or it is obvious and uncontroversial what single series of corporations were the single "main" organization in series). – The preceding unsigned comment was added by Closeapple (talk • contribs) at 2014-04-26T19:57:37 (UTC).
  Comment: The issue of stability is a good point, but wouldn't be much different than the common practice of legal entities' changing their names over time as they reorganize (e.g., ChevronTexaco to Chevron Corporation (Q319642)) and is manageable. Also, a point regarding their stability, EINs are permanent once issued and cannot be reused once canceled (see: -- SERutherford (talk) 21:46, 26 April 2014 (UTC)
I would ask that we better differentiate this to make it country specific or specific to the body, eg. US-EIN, or IRS-EIN. There is a whole world out there and disambiguating is necessary. Similarly, the other descriptive components are are more specific.  — billinghurst sDrewth 14:54, 6 May 2014 (UTC)
Maybe it should be FEIN if we use it; some state tax agencies in the U.S. have a so-called state "EIN" but I think it's always the FEIN unless one hasn't been issued for some obscure reason. (That reason would have to be very obscure, since any entity (including a person) having any employees, and any corporate body filing any federal tax forms, more or less requires an FEIN, even if the organization is filing as tax-exempt.) But what else would EIN or FEIN mean as a property, that would make it ambiguous? We don't normally prefix other properties with their countries, do we? Isn't that what the property description is for? --Closeapple (talk) 17:47, 6 May 2014 (UTC)
This not a item, it is a property and proposed as "Employer Identification Number", not EIN. We don't have "authority control"s, we have specific, and nowhere else do you see any duplication of a property name, so please be specific. Otherwise we end with a string of US-dominated strings, so disambiguate a property now.  — billinghurst sDrewth 14:29, 9 May 2014 (UTC)
  •   Support, enthusiastically. mapping these to our items may be difficult at times, and will require many-to-one and one-to-many mappings, with date range qualifiers, but that is very useful information to have handy in order to use in subsequent searches through government databases. John Vandenberg (talk) 21:33, 26 April 2014 (UTC)
  •   SupportYerpo Eh? 06:55, 7 May 2014 (UTC)
  • @SERutherford, Closeapple, billinghurst, John Vandenberg, Yerpo:   Done as a string. --Jakob (talk) 11:55, 7 May 2014 (UTC)

Berlin borough key

   Not done
Descriptionunique two-number key for Berlin boroughs
Data typeString
Template parameterNAME-OBEREBENE
Domainborough of Berlin (Q821435) (borough of Berlin)
Allowed values"01" to "12"
ExampleTempelhof-Schöneberg (Q158106) => 07
Proposed bySrittau (talk)

Srittau (talk) 01:37, 27 April 2014 (UTC)

  Question Would this (and the following) be special cases of Wikidata:Property_proposal/Place#German_Part_of_Municipality_Key? -- Gymel (talk) 06:49, 27 April 2014 (UTC)
Unfortunately not. It seems that Berlin has its own system for its districts and suburbs. --Srittau (talk) 23:45, 27 April 2014 (UTC)
Two digits for districts, two more for suburbs (AGS for anything "Berlin" is 11000000?). I have the impression that there is a uniform scheme all municipalities in Germany have agreed upon ("Gemeindeteilschlüssel", [13]). To construct appropriate subdivisions is the responsibility of the respective municipalities, therefore there is no general meaning apart from "level-1 subdivision" and "level-2 subdivision". However for purposes of destatis the municipalities do not have freedom to enumerate them in arbitrary form. -- Gymel (talk) 00:20, 28 April 2014 (UTC)
  Oppose. Allowed values = "01" to "12". --Kolja21 (talk) 01:23, 3 May 2014 (UTC)
  Comment: It seems that this is limited to exactly 12 Wikipedia items, already known, and the code is not duplicated or helpful for grouping the items in any way. Sometimes a unique code for only a dozen items might still be useful: There are probably some codes on Wikidata for admin-1 level divisions of specific countries with very few items at the admin-1 level. (I don't know which, though.) But I think there needs to be some evidence that this code might be the link to some useful information somewhere. Does some significant data use this number, rather than using the "AGS" or "Gemeindeteilschlüssel" mentioned above? --Closeapple (talk) 18:09, 6 May 2014 (UTC)
  •   Not done no support for creation. --Jakob (talk) 22:47, 6 June 2014 (UTC)

Berlin locality key

   Not done
Descriptionunique four-number key for Berlin localities
Data typeString
Template parameterORTSTEILNUMMER
DomainOrtsteil (Q253019) (Ortsteil)
Allowed values\d\d\d\d
ExampleLichtenrade (Q821101) => 0706
Proposed bySrittau (talk)

Srittau (talk) 01:37, 27 April 2014 (UTC)

  Oppose Sorry, properties are for general use, not for single cities. --Kolja21 (talk) 13:10, 27 April 2014 (UTC)
So there is already a property to describe this number? If not, your opposition is not very constructive. --Srittau (talk) 23:46, 27 April 2014 (UTC)
  Stöder These kinds of properties are almost always limited to one single nation. A German bundesland has the size of a smaller nation, so I see no problem. -- Innocent bystander (talk) (The user previously known as Lavallen) 09:20, 30 April 2014 (UTC)
According to de:Liste_der_Bezirke_und_Ortsteile_Berlins there exist 96 Ortsteile within the 12 districts (four digits, the first two of them reflect the district number of the proposal above), and they are not to be confused with the 95 statistical subdistricts with three-digit-numbers (I probably have confused them already in my comment to the property above). It should be explained for what purpose these Ortsteilnummern might be of use, IMHO it usually suffices to have items for all entities and not to give an arbitray, albeit official numbering. -- Gymel (talk) 11:38, 30 April 2014 (UTC)

  Not done Srittau, Kolja21, Innocent bystander, Gymel: Not enough support.--Micru (talk) 06:47, 11 June 2014 (UTC)

OpenCorporates identifier

Descriptionstring that uniquely identifies a legally distinct corporate entity.
Data typeString
Domaincompany (Q783794), organization (Q43229), corporation (Q167037), juridical person (Q155076)
Allowed valuesJurisdiction prefix (ISO3166-2 format) plus identifier based on official registry entry; format ^[a-z_]+\/\w+$
ExampleGoldman Sachs (Q193326) => us_de/2923466 (see or
Robot and gadget jobsBots could populate identifiers using the OpenCorporates name resolution service. Bots could also be written to create company entries where they don't already exist.
Proposed byGangle

An open, authoritative identifier with good worldwide coverage for all corporate entities is needed to cross-reference statements about corporations. This is complementary to Legal Entity Identifier (P1278), because Legal Entity Identifiers are still in the process of standardization and won't have decent worldwide coverage for many years. OpenCorporates identifiers are based on jurisdiction and the company number issued by the corporate registry, and are completely open. An identifier also provides a cross reference to the actively managed corpus of open corporate metadata on Gangle (talk) 10:55, 13 May 2014 (UTC)

  Support ·addshore· talk to me! 11:11, 13 May 2014 (UTC)
  Support --LydiaPintscher (talk) 15:18, 13 May 2014 (UTC)
  Support John Vandenberg (talk) 11:04, 14 May 2014 (UTC)

Italian Chamber of Deputies ID

DescriptionID code from the Italian Chamber of Deputies for incumbent and former deputies
Data typeString
Domainpolitician (Q82955)
Allowed valuesp[number] (ex. "p15080")
SourceChamber of Deputies of Italy (Q841424)
Robot and gadget jobsbot uploading
Proposed bySannita - not just another sysop

This is the first step, in order to import the datasets from Chamber of Deputies of Italy (Q841424) (which unfortunately are still CC-BY). It is anyway a relevant source of infos about Italian politicians, and a potential mine for new data. Sannita - not just another sysop 11:29, 17 May 2014 (UTC)

  Support, seems ok. Just interesting: is there some website where to see all of those ID's? --Stryn (talk) 11:48, 17 May 2014 (UTC)
Datasets are available here (link in Italian). Here's an example. --Sannita - not just another sysop 12:30, 17 May 2014 (UTC)


   Done: EPSG CRS (P1338) (Talk and documentation)
DescriptionEuropean Petroleum Survey Group Geodesy ID for geodetic datum
Data typeString
Template parameter?
DomainGeodetic datum items
Allowed valuesstring of numbers
ExampleSWEREF99 (Q10659701) = "4619" ( (usually a Wikipedia article covers a group of geodetic data, that have some differences in very specific details. We could either create items for these, or just assign more than one code to the item (see:
Proposed by--Tobias1984 (talk) 18:50, 19 February 2014 (UTC)

  Question Tobias1984, how will you import the data? Parsing it from the site, or from wikipedias? How is it planned to use this data?--Micru (talk) 12:41, 29 April 2014 (UTC)

@Micru: - I thought adding this by hand should work well (There aren't that many articles about projections). The data would just serve as general information querying. For example if I am looking for the type of map projections used in Sweden I would get the information from the example. A long term goal would be to categorize the historic and official maps on Commons with this property. And even farther in the future we might have a database of every map ever made by an official institute, with the correct metadata attached to it. Tobias1984 (talk) 14:36, 29 April 2014 (UTC)
@Tobias1984: I am testing a new property assessment method, according to it:
  • previous or requested use in MW projects: 0 (or at least I don't know about it)
  • fulfills a structural WD/WP/task force need: 0 (I haven't seen this in the context of a task force/infobox)
  • data availability for importing or generated value: 5 (the data seems valuable)
  • easy to import or easy to understand: 0 (manual)
  • plans to re-use data: 5 (enables map categorization)
  • property ownership: 3 (implicated user, but no group backing)
Result: 13 points,   weak oppose for now. Recommendation to improve rating: get support from a task force/wikiproject so it it is not only a one-man job. Discuss with any Wikipedia possible data re-use.--Micru (talk) 06:35, 30 April 2014 (UTC)
  •   Support. Tobias does an immense amount of good work in our geological data and I trust his judgement here. This property is really more about authority control in the natural sciences, which is not controversial. I don't think there should be a high bar for the creation of such properties. Emw (talk) 20:22, 31 May 2014 (UTC)
  •   Support --- Jura 09:39, 1 June 2014 (UTC)
  • @Tobias1984, Micru, Emw, Jura1:   Done, there is sufficient support for the creation of this property. --Jakob (talk) 11:44, 1 June 2014 (UTC)

territory claimed by

Descriptionvery useful item to POV use.
Data typeItem
Allowed valuescountry, organization, person
ExampleSenkaku Islands (Q29460) located in the administrative territorial entity (P131) Toucheng (Q706299) <claimed by> People's Republic of China (Q148) & Taiwan (Q865), Tonoshiro (Q11579187) <claimed by> Japan (Q17)
Sourceexternal reference, Wikipedia list article (either infobox or source)
Proposed byGZWDer (talk)

Motivation. GZWDer (talk) 15:46, 25 March 2014 (UTC)

@GZWDer: I am testing a new property assessment method, according to it:
  • previous or requested use in MW projects: 5 (infobox field)
  • fulfills a structural WD/WP/task force need: 5 (needed to represent territorial claims)
  • data availability for importing or generated value: 5 (the data seems available)
  • easy to import or easy to understand: 5 (bot jobs)
  • plans to re-use data: 5 (infobox)
  • property ownership: 3 (implicated user, no information given about bot plans)
Result: 28 points   Support.--Micru (talk) 06:44, 30 April 2014 (UTC)
  Oppose use stated in (P248). --Succu (talk) 21:44, 5 May 2014 (UTC)
Succu I think you misunderstood this property, I have changed the label. Check again to see if now it is more clear.--Micru (talk) 12:56, 23 May 2014 (UTC)
@Jakec: Territory claims are a special point of view. Seldom well founded. A possibillity to model these point of view claims part of (P361) (or something similar) together with stated in (P248). So I see no need for this allreday misused property. --Succu (talk) 20:19, 10 June 2014 (UTC)
  •   Support. Emw (talk) 20:24, 31 May 2014 (UTC)
  • @GZWDer, Micru, Succu, Emw:   Done Considering that the sole opposition was a misunderstanding, there is consensus for creation. --Jakob (talk) 22:44, 31 May 2014 (UTC)
  •   Oppose. This is indicating the source of a statement, so it should be in the sources. Use stated in (P248), per Succu. @Jakec: How did you come to the conclusion that Succu's opposition was a misunderstanding? --Yair rand (talk) 20:02, 10 June 2014 (UTC)
  • But the current use is certainly worth considering. --Jakob (talk) 21:27, 10 June 2014 (UTC)
Your „current use“ should be mapped with rank=prefered and a source stated in (P248). If... --Succu (talk) 21:45, 10 June 2014 (UTC)
But two countries may have equally strong claims to a territory. --Jakob (talk) 10:22, 13 June 2014 (UTC)
So set them both as preferred. --Yair rand (talk) 08:51, 19 June 2014 (UTC)

language of diffusion


As if figuring out the difference between language of work or name (P407) and original language of film or TV show (P364) wasn't hard enough, neither of these is usable for the language of media outlets (as opposed to specific works), hence this proposal, which should reduce the amount of constraints violations of these two properties. Circeus (talk) 00:28, 31 March 2014 (UTC)

  Oppose It seems redundant with language of work or name (P407) and the templates linked doesn't seem to make any distinction. I would recommend expanding the scope of language of work or name (P407) if necessary.--Micru (talk) 13:01, 23 May 2014 (UTC)

Wikisource category

   Not done
Descriptionequivalent to Commons category (P373)
Data typeput datatype here (item, string, media, coordinate, monolingual text, multilingual text, time, URL, number)-invalid datatype (not in Module:i18n/datatype)
Template parameterput Wikipedia infobox parameters here, if existing; ex: "population" in en:template:infobox settlement
Domaintypes of items that may bear this property; preferably use Q templates, as specialized as possible, or text. e.g. astronomical object (Q6999) (astronomical object), disease (Q12136) (disease), aircraft (Q11436) (aircraft), and more generally geographic location (Q2221906) (place), organization (Q43229) (organization), event (Q1656682) (event), work (Q386724) (creative work), etc. Special values (having specialized validation schemes): Persons, Taxons
Allowed valuestype of linked items (Q template or text), list or range of allowed values, string pattern...
Format and edit filter validation(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
Sourceexternal reference, Wikipedia list article (either infobox or source)
Robot and gadget jobsShould or are bots or gadgets doing any task with this? (Checking other properties for consistency, collecting data, etc.)
Proposed byYann (talk)

I am surprised that it doesn't exist yet. Yann (talk) 13:41, 31 March 2014 (UTC)

  Oppose Not really needed. You can just put it as a sitelink, but of course it is always better to link it from the item representing the category.--Micru (talk) 08:46, 7 May 2014 (UTC)
  • @Yann, Micru:   Not done - no support. --Jakob (talk) 17:37, 9 June 2014 (UTC)
    • @Jakob, Micru: Sorry, but I don't understand what is the issue here. This is needed and I don't see why this property could not be created. Also I was not informed of the opposition by Micru. Thanks, Yann (talk) 19:03, 14 June 2014 (UTC)
      • Yann, sorry that I didn't notify you, I thought you would check. Why do you need a property to link with the category, when it is already in the sitelinks here Category:École pratique des hautes études (Q13355080)? Perhaps I missed something?--Micru (talk) 10:49, 15 June 2014 (UTC)
        • @Micru: It is necessary in a case like Q668 (India). We have link with "Commons category": India, and we could have similarly "Wikisource category": India. We don't have a page on Wikisource called India, so we can't link Wikisource that way. It will be necessary to create interprojects links (Wikipedia article India to Wikisource category India). Regards, Yann (talk) 21:12, 17 June 2014 (UTC)
          • @Yann, tbh the "Commons category" property is redundant. The proper way to do it would be to use "topic's main category" to get the category item, and then from that item extract the sitelink to the category in wikisource. We already had the same discussion for Commons, and as far as remember that was the general consensus, to link categories with categories, pages with pages, and when some cross linking is necessary then use category's main topic (P301)/topic's main category (P910), to allow access to the other sitelinks. That won't work now (because of the missing aleatory access), but it seems counterproductive to create properties to link with each category of sister projects for something that is better done with a Lua script.--Micru (talk) 08:08, 19 June 2014 (UTC)
            • @Micru: Well, right now, the Commons category is useful. It allows interwiki links on Commons. Seeing that the deployment takes time, I only care about what works now, not what might work in an undefined future. Yann (talk) 18:33, 22 June 2014 (UTC)

don't merge with

   Not done
Descriptionthis item shouldn't be merged with that other item
Data typeItem
Template parametern/a
Domainitems appearing @Magnus Manske:'s
Allowed valuesitems
Example 1MISSING
Example 2MISSING
Example 3MISSING
Robot and gadget jobscould be added by wikidata-game

Pure maintenance property to avoid that such items get merged. --- Jura 05:25, 26 May 2014 (UTC)

  Question Since this is meta-information, what about using {{Item documentation}} instead?--Micru (talk) 08:35, 26 May 2014 (UTC)
  Oppose This is no real property but meta information. {{Item documentation}} or item talk pages should be used instead.Casper Tinan (talk) 18:24, 26 May 2014 (UTC)
  Support with renaming different from. This will look less like meta informationTomT0m (talk) 18:38, 26 May 2014 (UTC)
  Comment I don't think it's a bad idea to store informations like this in statements. Soon we will be able to store informations about properties, we will be able to class properties. On the other hand storing in templates is far less efficient to retrieve when we want these informations and is far too hacky. there exists properties in the semantic web to state two ressources denotes different individual, see differentFrom on OWL specification. TomT0m (talk) 18:38, 26 May 2014 (UTC)
  •   Support "different from". Isn't wikidata here to replace template based data in the first place? --- Jura 18:46, 26 May 2014 (UTC)
  •   Weak oppose. This property seems redundant with owl:disjointWith, i.e. Wikidata:Property_proposal/Archive/20#disjoint_with, or generally unnecessary. Also, the 'Example' field in the property proposal template is probably the most important field after 'Description'. Please don't leave it blank. Emw (talk) 01:40, 28 May 2014 (UTC)
: "disjoint with" means no instances in commmon. The property proposed here just means "not the same thing as". I think it would be useful, as I have reverted a few erroneous mergers like "Pokemon A" / "Pokemon A and its descendants". We could also imagine smarter pre-merger sanity checks based on part of (P361), subclass of (P279) etc, but that would be significantly more complicated. --Zolo (talk) 07:21, 28 May 2014 (UTC)
@Emw: As Zolo said. OWL defines to properties for difference, one for classes, and one for individuals. It seems important for disambiguation as we actually have a lot of items that can be both classified as individuals and instantiated (A car model can be a member of a car model class (think of a car model infobox and has instances (the car infobox)). TomT0m (talk) 11:33, 28 May 2014 (UTC)
  •   Question @Jura1: What would be the constraints on this property. There must be some; otherwise items would get completely bloated - there are millions of items that shouldn't be merged with a given item. --Jakob (talk) 01:55, 29 May 2014 (UTC)
    Not sure. I'd add it primarily through the merge tools @Magnus Manske:. This uses identical labels. Problematic there are mainly people with similar dates of births/nationalities/field of activity or similar topics that are not linked together with other properties. --- Jura 09:37, 1 June 2014 (UTC)
  •   Oppose, per Casper Tinan. --Yair rand (talk) 01:17, 9 June 2014 (UTC)
  • A property to support careless people playing a game which provides insufficient information to resolve a suggested task? Of course   Oppose. --Succu (talk) 20:30, 10 June 2014 (UTC)

  Not done No consensus reached.--Micru (talk) 19:58, 12 June 2014 (UTC)

Description in sources

Descriptionthe enumeration of sources (with URLs or item links as qualifcators) where item is described. Dictinary, encyclopedia, etc.
Data typeItem
Template parameternone
Domainany item
Allowed valuesany value, depicting valuable source of information
Format and edit filter validationeither link to item with wikisource link or URL should present (may be both)
Sourcetemplates such as ru:Template:ВТ-ЭСБЕ, ru:Template:ВТ-РБС, links to britanica articles from en-wiki, etc.
Robot and gadget jobsi can create the robot to fill those labels from ru-wikisource for Brockhaus and Efron Encyclopedic Dictionary (Q602358), Jewish Encyclopedia of Brockhaus and Efron (Q4173137) and Russian Biographical Dictionary (Q1960551) encyclopedias
Proposed byVlsergey (talk)

I believe it's better to use wikidata to organize all links to valuable enciclopedias (even if they were not used as sources in particular article) in some kind of navigation template like Template:Authority control (Q3907614). Colmmunity of each Wikipedia can decide, what sources they would like to see in this template, but common storage required to sync all data between different languages and betwwen different projects (i.e. wikipedia and wikisource) Vlsergey (talk) 12:37, 26 May 2014 (UTC)

  Support But with another name -> "source for the whole item".--Micru (talk) 11:35, 30 May 2014 (UTC)
Not Sure. There are cases where one source supports the whole item but there are (I believe) more cases where one sources supports 90% of the statement and another source supports the last 10% and I believe there would be a temptation to misuse. On the other hand there are items of marginal notabliity where we know there is only ever going to be one source. On the third hand For downloads and searches wikidata (as I understand it) looks more like a collection of statements than a collection of items. I wouldn't want statements to get detached from their sources. Filceolaire (talk) 13:11, 3 June 2014 (UTC)
This property is not for sources that support some statements in Wikipedia article. This is links to other encyclopedia articles about same item. They may be used as sources, but it does not means they were or they will be. -- Vlsergey (talk) 14:21, 3 June 2014 (UTC)