Wikidata:Property proposal/Archive/25

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

has immediate cause

Descriptiongranular factor that directly resulted in this outcome or effect. Used in conjunction with 'cause' (i.e. underlying cause), when the 'immediate cause' directly followed any values in 'cause'. Multiple values allowed.
Data typeItem
Template parameterNone
DomainNone. Often but not always events or processes. See examples and discussion.
Allowed valuesAny. Often but not always events or processes. See examples and discussion.
ExampleCretaceous–Paleogene extinction event (Q55811) has immediate cause starvation (Q853930), impact winter (Q3032301)
Sourceexternal references
Proposed byEmw (talk)

In talk about causes, things are often broken down into "underlying cause", "immediate cause" and "contributing factor". This proposal is part of a series to add those properties to Wikidata and help us structure data about causes. The goal is let Wikidata model causation in a way that balances expressiveness and simplicity. It emerges from discussion at Property_talk:P828#A_better_way_to_model_causation (permalink). See there for extended examples and discussion. If these properties are created, then fuller guidelines and documentation will be put together based on the linked discussion and any comments below. Emw (talk) 03:50, 11 September 2014 (UTC)[reply]

Danneks, Genewiki123, Infovarius, Micru, TomT0m, Filceolaire, Tobias1984, WS, Matthiassamwald, Danrok, John Vandenberg, others, there seems to be general agreement in Property_talk:P828#A_better_way_to_model_causation for a three-tiered approach to modelling causation. Cause is a complex subject but I think we can grow guidelines over time as we gain experience with structuring data about causation in different domains. Please comment and vote here on this property, and also on the 'has contributing factor' property proposal below. Assuming consensus is shown here on this three-tier framework, I'll plan to label of P828 from 'medical causes' to 'has cause' and update its description accordingly. Thanks, Emw (talk) 04:15, 11 September 2014 (UTC)[reply]

has contributing factor

Descriptioncondition, event or other thing that significantly contributed to, but did not directly cause, this outcome or effect. Multiple values allowed.
Data typeItem
Template parameterNone
DomainNone. See examples and discussion.
Allowed valuesAny. See examples and discussion.
ExampleAmerican Civil War (Q8676) has contributing factor caning of Charles Sumner (Q5032419), Dred Scott v. Sandford (Q690462), Bleeding Kansas (Q331377)
Sourceexternal references
Proposed byEmw (talk)

See proposal for 'has immediate cause' above. Emw (talk) 04:05, 11 September 2014 (UTC)[reply]

sourcing circumstances

Descriptionsourcing circumstances
Data typeItem
Template parameterprimary dates and places in infoboxes
Domainqualifiers can be used for references and also for claims, if all specified references for claim would have the same qualifier
Allowed valuesvery limited set of values, precisely documented on property talk page, including:
    • "approximate value" -- the source specified value and explicitly stated that value is approximate, but specify no precision information. For example, "born around 1709" or "died near Berlin".
    • "possibly approximate value" -- the source specified not-ideally-precise value (like year instead of date of birth) and did not explicitly stated that value is approximate, but other sources explicitly did.
    • "unspecified calendar" -- the source did not explicitly specify used calendar for dates between 1582-10-15 and 1918-01-26 (or, for example, for dates in context when Japanese calendar could be used), calendar model is assumed basing on source information and/or comparison with other sources
    • "misprint" -- the information in source differs very little from other sources and suspect to have a typo
    • "miscalculation" -- source provides both calculation sources, methodology (or link to it) and calculation result, but result is different from other sources and manual calculation basing on provided description
  • "misprint" example
  • "approximate value" example
  • Format and edit filter validation
    Sourcesources themselves, sometimes infoboxes for import ("born about 1234 year")
    Robot and gadget jobssome conditions shall be checked using bot:
  • some values implies existence of other claims with sources and "deprecated" rank of the claim ("misprint", "miscalculation")
  • some values implies existence of other references to the same claim ("possibly approximate value", "unspecified calendar" (if attached to reference))
  • "unspecified calendar" could be used only for time
  • Proposed byVlsergey (talk)

    Need to model complex sourcing cases with boolean "flags" like specifies above for cases like "born around 1709". Vlsergey (talk) 12:28, 12 September 2014 (UTC)[reply]

    •  Support --Micru (talk) 13:17, 12 September 2014 (UTC)[reply]
    •  SupportAyack (talk) 13:46, 12 September 2014 (UTC)[reply]
    • Pictogram voting comment.svg Comment - Could you provide examples of various usages? This is essential to properly document its use. LaddΩ chat ;) 15:32, 13 September 2014 (UTC)[reply]
     Support OK now I get it. LaddΩ chat ;) 20:51, 15 September 2014 (UTC)[reply]

    Sports Reference ID

    DescriptionID on the website Sports Reference.
    Data typeString
    Template parameter"1" or "ID" in en:Template:Sports-reference and no:Mal:Sports-reference
    Allowed valuesNot sure how to put it in regex, but the format is for example ja/anne-jahren-1 for no:Anne Jahren
    ExampleAnne Jahren (Q256757) => ja/anne-jahren-1
    Format and edit filter validationAn edit filter could check validity
    SourceMainly en:Template:Sports-reference (and in other languages)
    Robot and gadget jobsMy bot can import values from various Wikipedias
    Proposed byJon Harald Søby (talk)

    This is a very useful database for sports articles, and being able to pull their IDs from Wikidata would be immensely helpful for the projects. Jon Harald Søby (talk) 17:05, 12 August 2014 (UTC)[reply]

     Support --- Løken (talk) 19:57, 16 August 2014 (UTC)[reply]
     Support linking to a sportsreference profile is common in EL sections on enwiki. --AmaryllisGardener talk 20:55, 16 August 2014 (UTC)[reply]

    ID on

       Done: ID (P1453) (Talk and documentation)
    DescriptionID on encyclopedia, based on Catholic Encyclopedia (Q4217446), in Russian
    Data typeString
    Template parameter
    DomainChristianity and related
    Allowed values\d+
    ExampleThomas Aquinas (Q9438) → 485
    Format and edit filter validation
    Sourcesite itself
    Robot and gadget jobsGadget URL: Thomas Aquinas (Q9438) → 485 →
    Proposed byVlsergey (talk)

    Another reliable encyclopedia link. Vlsergey (talk) 18:05, 14 August 2014 (UTC)[reply]

    WALS lect code

       Done: WALS lect code (P1466) (Talk and documentation)
    DescriptionLanguage code from the World Atlas of Language Structures (
    Data typeString
    Domainlanguoid (Q17376908) or a subclass (in practice languages, dialects, and "varieties")
    Allowed valuesstrings consisting of three lowercase letters
    ExampleMartu Wangka (Q3295916) -> mwa, Bokar (Q4938727) -> boq
    Format and edit filter validationregex /^[a-z]{3}$/ should work
    Source, has conversion between multitree (linguist list), ISO 639, glottocodes, and WALS codes
    Robot and gadget jobsIt could be done.
    Proposed byπr2 (t • c)

    This would be a useful addition, as some languages with WALS codes don't have ISO 639-3 (yet). πr2 (t • c) 02:51, 22 August 2014 (UTC)[reply]

    WALS genus code

    DescriptionLanguage "genus" code from the World Atlas of Language Structures (
    Data typeString
    Domainlanguoid (Q17376908) or a subclass (in practice language families and dialect continua)
    Allowed valuesstrings consisting of one or more lowercase letters
    ExampleHarakmbut languages (Q1584402) -> harakmbet, Pama–Nyungan (Q33942) -> pamanyungan
    Format and edit filter validationregex /^[a-z]+$/ should work
    Source, has conversion between multitree (linguist list), ISO 639, glottocodes, and WALS codes
    Robot and gadget jobsIt could be done.
    Proposed byπr2 (t • c)

    This would be a useful addition, as many language families with WALS genus codes don't have ISO 639-3 (yet). πr2 (t • c) 02:51, 22 August 2014 (UTC)[reply]

    WALS family code

    DescriptionLanguage (super-)family code from the World Atlas of Language Structures (
    Data typeString
    Domainlanguoid (Q17376908) or a subclass (in practice only language families)
    Allowed valuesstrings consisting of one or more letters that are valid WALS identifiers
    ExampleGuaicuruan languages (Q1974973) -> guaicuruan
    Format and edit filter validationregex /^[a-z]+$/ should work
    Source, has conversion between multitree (linguist list), ISO 639, glottocodes, and WALS codes
    Robot and gadget jobsIt could be done.
    Proposed byπr2 (t • c)

    This would be a useful addition, as some language families with WALS codes don't have ISO 639-3 (yet). Many of these have glottocodes however. πr2 (t • c) 02:51, 22 August 2014 (UTC)[reply]

    FIFA player code

       Done: FIFA player ID (P1469) (Talk and documentation)
    DescriptionPlayer code in the official FIFA database
    Data typeString
    Template parameterTemplate:FIFA player (Q6526850)
    Domainhuman (Q5)
    Allowed values\d+
    ExamplePretty much any footballer you can think of. Neymar (Q142794) => 314197
    SourceTemplate:FIFA player (Q6526850)
    Robot and gadget jobsMy bot can import values from various Wikipedias
    Proposed byJon Harald Søby (talk)

    This one would be great, because it is very versatile. Not only can you use it to link to the general stats and records for a player, but also to their Confederations Cup profile and World Cup profile and so on. Jon Harald Søby (talk) 16:42, 22 August 2014 (UTC)[reply]

    Patientplus ID

       Done: Patientplus ID (P1461) (Talk and documentation)
    DescriptionID at
    Data typeString
    Template parameterit'll be used directly
    Domaindisease (Q12136)
    Allowed valuesletters, numbers, and some extra characters like '()-
    ExampleAsperger syndrome (Q161790) => Asperger's-Syndrome
    Format and edit filter validation[A-Za-z\-'\(\)]
    Source[1] and on
    Robot and gadget jobsDexbot will do
    Proposed byAmir (talk)

    I was asked by one of medicine content writer in English Wikipedia Amir (talk) 15:14, 24 August 2014 (UTC)[reply]

     Support - Hi Amir @Ladsgroup:. Don't forget to ping the appropriate project if you request a property :) That will speed up voting. Tobias1984 (talk) 15:46, 27 August 2014 (UTC)[reply]
    Doc James
    Daniel Mietchen
    Andrew Su
    Sabah Ul-Hasan
    Projekt ANA
    Pavel Dušek
    Was a bee
    Chris Mungall
    Dr. Abhijeet Safai
    Sami Mlouhi
    Netha Hussain
    Abhijeet Safai
    Shani Evenstein
    ZI Jony
    The Anome
    Pictogram voting comment.svg Notified participants of WikiProject Medicine -Tobias1984 (talk) 15:46, 27 August 2014 (UTC)[reply]
     Support After seeing the outcome of a similiar project at en:Wikipedia_talk:WikiProject_Medicine#Wikidata_project_to_associate_drugs_with_interactions I feel like I would want this kind of property to use for other, similar government websites to get a comparable outcome. Blue Rasberry (talk) 16:47, 27 August 2014 (UTC)[reply]
    Thank you, I'll do the next time Amir (talk) 20:51, 27 August 2014 (UTC)[reply]


    DescriptionID for authors in Post-Reformation Digital Library (Q7233488)
    Data typeString
    Template parameterNot an infobox, but en:Template:PRDL uses this parameter
    Allowed valuesnumbers 1 to 6205
    ExampleJohn Calvin (Q37577) => 1 (see
    Format and edit filter validation1 to 4 digit number
    SourceI have a spreadsheet cross-referenced by Worldcat ID.
    Robot and gadget jobsI'm not sure how this works, but if it is possible for me to prepare the spreadsheet and hand it over to a bot that would be nice.
    Proposed byJfhutson (talk)

    A very useful website. Jfhutson (talk) 14:25, 1 September 2014 (UTC)[reply]

    Nupill Literatura Digital - Author

       Done: BLPL author ID (P1473) (Talk and documentation)
    DescriptionAuthor ID on Nupill Literatura Digital
    Data typeString
    Domainhuman (Q5)
    Allowed valuesnumbers
    ExampleMário de Andrade (Q203623) => 7416 (
    Proposed byLugusto (talk)

    Nupill Literatura Digital is an impressive database regarding authors and books in portuguese language and about portuguese speaking countries/cultures. Certainly a must have for Wikidata. Lugusto (talk) 23:24, 9 September 2014 (UTC)[reply]

     Support --Micru (talk) 13:15, 12 September 2014 (UTC)[reply]

    Nupill Literatura Digital - Document

    DescriptionWork ID on Nupill Literatura Digital
    Data typeString
    Domainwork (Q386724)
    Allowed valuesnumbers
    ExampleThe Posthumous Memoirs of Bras Cubas (Q1464493) => 136488 (
    Proposed byLugusto (talk)

    Nupill Literatura Digital is an impressive database regarding authors and books in portuguese language and about portuguese speaking countries/cultures. Certainly a must have for Wikidata. Lugusto (talk) 23:24, 9 September 2014 (UTC)[reply]

    @555, Micru: Created in addition to BLPL author ID (P1473). --Kolja21 (talk) 03:52, 17 September 2014 (UTC)[reply] identifier

       Done: ID (P1481) (Talk and documentation)
    DescriptionIdentifier of a historical place of the roman empire at
    Data typeString
    Domainarcheological and historical places, e.g. Fectio (Q2311719)
    Allowed valuespositive integer
    ExampleFectio (Q2311719) => 49
    Format and edit filter validation?
    Proposed byJakobVoss (talk)
    Discussion provides a mapping to Wikidata items. These links should be brought back into Wikidata to be reused by other parties. I contacted them to provide a dump of mappings to be imported in Wikidata. -- JakobVoss (talk) 13:52, 11 September 2014 (UTC) I've added a BEACON file to make life easier. It is at . Let me know if things need to change. Coming months I hope to link much more places.--René Voorburg (talk) 13:58, 12 September 2014 (UTC) -- Always uptodate is . The dataset now contain about 20.000 objects.René Voorburg (talk) 06:19, 16 September 2014 (UTC)[reply]

     Support --Kolja21 (talk) 23:34, 11 September 2014 (UTC)[reply]

    Stack Exchange tag

    DescriptionLink to tag for the concept on a Stack Exchange network website like
    Data typeURL
    DomainVery general, since Stack Exchange sites cover a wide variety of domains, although each site is usually rather well scoped.
    Allowed valuesURLs on any of Stack Exchange sites (full list) of the form "http://site/tags/tag"
    ExamplePHP (Q59) => ""; present perfect (Q1240211) => ""
    Sourceexternal reference
    Robot and gadget jobsBots could look for new tags, look for tag aliases, etc.
    Proposed byBskaggs (talk)

    I've curated a list of a few thousand Stack Overflow tags and their corresponding Wikipedia articles. I asked on the project chat about how to incorporate it, and Emw suggested that this seems like the right place to store the information. I just built a list for, but I'm sure this can be done for any of the other sites.

    I need a little clarification about creating such a property - is it acceptable to have multiple values of this property for the same entity? Because, for example, Apache HTTP Server (Q11354) should link to at least "apache" on, "apache-2.2" on, and "apache-2.4" on Also, notice how some sites have separate tags for distinct versions of some software; would this be fine? Finally some Stack Exchange correspond to things that only have a sub-section of a Wikipedia article, and I'm not sure of the right way to refer to them on Wikidata.

    Thanks for your input! Bskaggs (talk) 01:33, 12 September 2014 (UTC)[reply]

    Ruud Koot
    Daniel Mietchen
    Tinker Bell
    Jasc PL
    Tris T7
    Peb Aryan
    FWVH (passionné d'informatique et d'électronique)
    Sylvain Leroux
    Pictogram voting comment.svg Notified participants of WikiProject Informatics @Emw:

    • "is it acceptable to have multiple values of this property for the same entity?"
    Yes, I think that would be acceptable, especially in the initial implementation. Ideally, in my opinion, we would separate each version of a piece of software into its own Wikidata item even if it does not have a Wikipedia article, but that's probably too ambitious to start with. So if Apache HTTP Server (Q11354) links "apache" on, "apache-2.2" on, and "apache-2.4" on, that's fine. We can refine things later.
    • "Also, notice how some sites have separate tags for distinct versions of some software; would this be fine?"
    • "Finally some Stack Exchange correspond to things that only have a sub-section of a Wikipedia article, and I'm not sure of the right way to refer to them on Wikidata."
    Please create an item for those things. You can do so programmatically. Wikidata:Bots has a lot of bots that do this kind of stuff and Wikidata:Bot_requests has notes on past work.

    Ruud Koot
    Daniel Mietchen
    Tinker Bell
    Jasc PL
    Tris T7
    Peb Aryan
    FWVH (passionné d'informatique et d'électronique)
    Sylvain Leroux
    Pictogram voting comment.svg Notified participants of WikiProject Informatics (your previous might not have gone through.) Emw (talk) 02:22, 12 September 2014 (UTC)[reply]

     Support --Micru (talk) 13:15, 12 September 2014 (UTC)[reply]

    * Oppose Unless I'm greatly mistaken, StackOverflow is a forum of sorts, so why link to it? --Jakob (talk) 11:32, 16 September 2014 (UTC)[reply]

    •  Support Tag creation on stackoverflow is done by experienced users, users can not tag their questions with less common tags and stackoverflow itself is under "cc by-sa 3.0" so this is supporting another free knowledge database, I don't see any reason to oppose and like the idea. Also in future they can integrate their service with Wikidata for retrieving descriptions or we can compare our descriptions with theirs. –ebraminiotalk 13:45, 16 September 2014 (UTC)[reply]
    • @Bskaggs, Emw, Micru, Ebraminio: ✓ Done --Jakob (talk) 11:31, 19 September 2014 (UTC)[reply]

    Iran statistics id

    DescriptionIran's statistics id which is used at for Iran's settlement places
    Data typeString
    DomainIran's settlement places (Province, County, City, village)
    Allowed valuesNumber from 2-19 digit
    ExampleIran stat > 0302 for Ahar (Q911727)
    Source at excel files
    Robot and gadget jobsIt will useful for bots to update people census of Iran's settlements by excel files which are published by every 5 years
    Proposed byYamaha5 (talk)

    The suggested code will be useful to find correct places at statistical tables of Iran's statistics organization. Iran has many settlements which have the same name and disambiguate of them needs this code. also in other Iran's data centers like this they uses this code (كدآماري). Yamaha5 (talk) 13:15, 16 September 2014 (UTC)[reply]

    ✓ Doneebraminiotalk 00:17, 23 September 2014 (UTC)[reply]

    Wikipedian in Residence at

       Not done
    DescriptionA venue or organisation where the person has worked as a w:Wikipedian in Residence
    Data typeitem (perhaps also string?)-invalid datatype (not in Module:i18n/datatype)
    Allowed valuesPeople
    ExampleSarah Stierch (Q5060734) => Smithsonian Institution Archives (Q2154134)
    SourceList at
    Proposed byAndy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits

    If the community prefers, this could be reversed, as a property of the institution. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:19, 16 June 2014 (UTC)[reply]

    • Andy, we already have work location (P937) for such cases.--Micru (talk) 14:28, 17 June 2014 (UTC)[reply]
      • And how does that tell us that someone was a WiR (as opposed to a curator, or a janitor)? Also, many WiR posts are voluntary, not "work". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:14, 17 June 2014 (UTC)[reply]
        • Using occupation (P106) as qualifier you can describe the position. We don't have any property for salary yet (currency units still will take a while), but when we do you could enter "salary:no value" for volunteers.--Micru (talk) 19:40, 17 June 2014 (UTC)[reply]
    •  Oppose Redundant to the aforementioned property.--Jasper Deng (talk) 18:48, 15 August 2014 (UTC)[reply]

     Not done Redundant to existing property.--Micru (talk) 11:28, 17 August 2014 (UTC)[reply]


       Done: plea (P1437) (Talk and documentation)
    Descriptionqualifier for "convicted of". Indicates whether the person pleaded guilty, not guilty, etc.
    Data typeItem
    Domainhuman (Q5)
    Allowed valueslegal pleas
    ExampleMark Hofmann (Q4354485) => guilt (Q705178)
    SourceWikipedia articles, court documents, etc.
    Proposed byDanrok (talk)

    Needed for use with convicted of (P1399) property. Probably important that we have this for living persons especially those who pleaded innocent (and may well be). Danrok (talk) 12:18, 29 June 2014 (UTC)[reply]

    bulk proposal: rugby union players IDs

       Not done
    DescriptionBarbarians, All Blacks, British Lions, French Barbarians, Pro12, ErcRugby, ItsRugby
    Data typeString
    Template parameterall parameters of it:Template:Link statistiche rugbisti a 15 (ESPNscrum player ID (P858) and ID (P861) already exist)
    Domain occupation (P106) rugby union player (Q14089670)
    Allowed valuesstrings
    ExampleAldo Birchall (Q3609130) <ErcRugby> "1660"
    Format and edit filter validationproperty-by-property
    Sourceitwiki, external websites
    Robot and gadget jobsbots can import such data
    Proposed byRicordisamoa 15:20, 21 July 2014 (UTC)[reply]
    Are those club-specific player IDs which would only apply to a small group of rugby players? Not sure what the policy is for such rather special properties. --Bthfan (talk) 15:23, 21 July 2014 (UTC)[reply]
     Oppose In my opinion, most of these properties are not needed. For Barbarians, All Blacks, French Barbarians, you can use the following syntax :
    Richie McCaw (Q726148) member of sports team (P54) New Zealand national rugby union team (Q55801)
    ranking (P1352) 1014.
    What's more, neither the official british & irish lions' website nor the ERCrugby one do use proper identification numbers. I don't see the need to store numbers in wikidata then. I think it would be better to have one proposal for each type of ID. That way, we would be more able to sort out which property is needed and which is not. --Casper Tinan (talk) 20:12, 21 July 2014 (UTC)[reply]

     Not done Redundant to existing property.--Micru (talk) 11:29, 17 August 2014 (UTC)[reply]


       Done: FIDE player ID (P1440) (Talk and documentation)
    DescriptionID for chess players used by the World Chess Federation (FIDE)
    Data typeString
    Template parameter"FideID" in en:Template:Infobox chess biography and in de:Vorlage:Infobox Schachspieler (Template:Infobox chess biography (Q5619879))
    Allowed valuesThe IDs are strings of digits in range 0-9
    ExampleMagnus Carlsen (Q106807) → "1503014"; Bu Xiangzhi (Q470788) → "8601445"
    SourceFIDE (Lists with all players can be downloaded from; Wikipedia infoboxes (parameter FideID in enwiki and dewiki); Wikipedia Template (Template:FIDE (Q6309645), en:template:Fide, de:Vorlage:FIDE) as "id" or first unnamed parameter.
    Robot and gadget jobsIf chess players have this property, it will be easy for bots to automatically update Elo rating (P1087)
    Proposed byDipsacus fullonum (talk)

    Elo rating (P1087) changes every month, and it is a lot of work to keep the Wikipedias updated. This proposed FideID property for chess players, will make it easy to match chess players on FIDE's rating lists with their Wikidata items, so bots automatically can keep Wikidata (and indirectly Wikipedias using Wikidata) updated with the chess players' actual rating. Dipsacus fullonum (talk) 09:11, 30 July 2014 (UTC)[reply]

     Support, with timestamp. Conny (talk) 07:07, 6 August 2014 (UTC).[reply]
    Thanks, but with what timestamp? The FideID for a chess player will hopefully never change. Dipsacus fullonum (talk) 08:42, 6 August 2014 (UTC)[reply]
    @Conny: Are you possibly confusing this with FIDE rating? --Jakob (talk) 11:32, 6 August 2014 (UTC)[reply]
    Argh, yes. Thank you :) . Conny (talk) 11:37, 6 August 2014 (UTC).[reply]

    grave picture

       Done: image of grave (P1442) (Talk and documentation)
    Descriptionlink to the picture of the grave or tombstone of the deceased person
    Data typeimage-invalid datatype (not in Module:i18n/datatype)
    Template parameterindiquez ici les attributs d'infobox de Wikipédia correspondants, s'il en existe, par exemple : « population » dans fr:Modèle:Infobox Subdivision administrative
    Allowed valuesimages
    Exampleexemple d'élément qui utiliserait cette propriété, avec une valeur proposée; par exemple : Universe (Q1) => Earth (Q2)
    Format and edit filter validation(exemple : 7 chiffres peuvent être validés avec le filtre d'édition Special:AbuseFilter/17)
    Sourceimage de Wikipédia ou de commons /picture from wikipedia, or the existing Commons cat.
    Robot and gadget jobsDevrait-il y avoir ou existe-t-il des bots ou des gadgets qui effectueront des tâches avec cette propriété? Par exemple: vérifier les autres propriétés afin d'être cohérent, collecter des données, automatiser un lien externe, etc.
    Proposed byHsarrazin (talk)

    Motivation Hsarrazin (talk) 18:41, 6 August 2014 (UTC) Considering the very important number of images of people's grave, I would suggest to add a type of image (for people), to indicate that the picture is representing the grave or tombstone, not the actual person (or portrait).[reply]

    •  Support -- Vlsergey (talk) 08:58, 14 July 2014 (UTC)[reply]
    •  Support, please move the discussion to WD:PP. — Ivan A. Krestinin (talk) 21:31, 5 August 2014 (UTC)[reply]
    •  Support As producer of 5000+ pictures of graves, I can only agree Pyb (talk) 18:52, 6 August 2014 (UTC)[reply]
    •  Support. Emw (talk) 12:37, 7 August 2014 (UTC)[reply]
    • Pictogram voting comment.svg Comment There were already discussions about deleting all "type of image" properties and replace them with image (P18) and an adequate qualfier. --Pasleim (talk) 10:57, 9 August 2014 (UTC)[reply]
    • now, that would be ok to me too… - my proposal was, considering the existing system :)--Hsarrazin (talk) 18:08, 9 August 2014 (UTC)[reply]
    • I don't think that would work, since a grave picture really isn't an image of a person in any way. --Jakob (talk) 21:29, 9 August 2014 (UTC)[reply]
    • well, some grave have a portrait on it, and when there is no other pic for the person, it is better than nothing ;) - and much better than a plaque on a house the person lived in… at least, there is only 1 grave - and generally, the person is in it, but there may be a lot of places where the person has stayed :D --Hsarrazin (talk) 20:53, 11 August 2014 (UTC)[reply]

    @Hsarrazin, Vlsergey, Ivan A. Krestinin, Pyb, Emw, Pasleim: ✓ Done--Micru (talk) 11:25, 17 August 2014 (UTC)[reply]

    Commons Creator Template

    DescriptionFor many artists, Commons has a "Creator" template, for example Commons:Creator:Augustus Henry Fox.
    Data typeString
    Template parametern/a
    DomainPersons, groups of people
    Allowed valuesCreator-space page on Commons
    Examplesample items that would use that property, with proposed values; example: Augustus Henry Fox (Q16212572) => Commons:Creator:Augustus Henry Fox
    Format and edit filter validationCreator-space page on Commons
    SourceCreator-space page on Commons
    Robot and gadget jobsYes; note that Creator templates (such as the example above) may include a Wikidata ID.
    Proposed byAndy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits

    Thank you. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:13, 1 September 2014 (UTC)[reply]

     Support, a majority of creator templates include a wikidata ID, since it was the only way to create a link to wikidata and retrieve wikilinks through a script, at the time it was created. I worked a lot on those creators, to include wikidata and/or other informations, and there are a lot of information there, that should be recovered, before the template can be converted to use wikidata items ;) --Hsarrazin (talk) 12:04, 3 September 2014 (UTC)[reply]
    strong  Support: same with me --Oursana (talk) 08:17, 16 September 2014 (UTC)[reply]
     SupportAyack (talk) 11:21, 16 September 2014 (UTC)[reply]

    legal form

       Done: legal form (P1454) (Talk and documentation)
    Descriptionlegal form on an organization within the legal framework of its seat country (Inc., LLC, etc.)
    Data typeItem
    Template parameterHier den Infobox-Parameter einfügen, falls es einen gibt. Beispiel: "Rechtsform" (legal form) in de:Vorlage:Infobox Verband
    Domainorganization (Q43229), juridical person (Q155076)
    Allowed valuesGmbH (Q460178), foundation (Q157031), public establishment of an administrative nature (Q3244038), GmbH (Q460178) and many more
    ExampleQ169552 => Q897103
    Sourcede:Vorlage:Infobox Verband, etc.
    Robot and gadget jobsBot could extract data from articles with templates de:Vorlage:Infobox Verband, de:Vorlage:Infobox Unternehmen etc.
    Proposed byMB-one (talk)

    This property is of interest for any kind of organization. MB-one (talk) 11:11, 23 July 2014 (UTC)[reply]

    people missing

    Descriptionnumber of people reported missing as a result of this event
    Data typeNumber (not available yet)
    Domainearthquake (Q7944) and other disasters, wars, etc.
    Allowed valueswhole numbers
    Example2013 Lushan earthquake (Q11547847) => 23
    Sourceexternal reference, articles, etc.
    Proposed byDanrok (talk)

    Typical statistic for disasters which is normally reported by sources. Danrok (talk) 12:34, 11 July 2014 (UTC)[reply]

    journey origin

       Done: start point (P1427) (Talk and documentation)
    Descriptionthe starting place of this journey, flight, voyage, trek, etc. (excl. spaceflight launch)
    Data typeItem
    Template parameter"origin" in en:template:infobox aircraft occurrence
    Allowed valuesplaces
    Example1971 RAF Hercules crash (Q3997249) => Pisa International Airport (Q709169)
    Robot and gadget jobsbots can collect from infoboxes for some of these
    Proposed byDanrok (talk) 16:27, 7 February 2014 (UTC)[reply]

    journey destination

    Descriptionthe intended destination for this journey, flight, sailing, exploration, etc.
    Data typeItem
    Template parameter"destination" in en:template:infobox Airliner accident
    Domaingeographic location (Q2221906)
    Allowed valuesany geographic location, e.g. airport, city, island, etc.
    Examplesample items that would use that property, with proposed values; example: Iran Aseman Airlines Flight 6895 (Q741126) => Tehran Imam Khomeini International Airport (Q734729)
    Format and edit filter validation(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
    Robot and gadget jobsBots can fill in from infoboxes.
    Proposed byDanrok (talk)

    Similar property to start point (P1427) Danrok (talk) 11:19, 20 July 2014 (UTC)[reply]

     Support -Tobias1984 (talk) 13:04, 20 July 2014 (UTC)[reply]

    @Danrok, Tobias1984, Zellfaze: ✓ Done--Micru (talk) 11:47, 17 August 2014 (UTC)[reply]

    score method

       Done: score method (P1443) (Talk and documentation)
    Descriptionqualifier of "points/goal scored by (P1363)"
    Data typeItem
    Allowed valuese.g. for <points/goal scored by> use this property to indicate if it was a goal, a try, a conversion, an own goal, a behind etc.
    Example<goal scored by> Geoff Hurst (Q212738) => own goal (Q18533)
    Sourceexternal reference, Wikipedia list article (either infobox or source)
    Proposed byFilceolaire (talk)

    to record details of sports matches. Filceolaire (talk) 20:11, 10 June 2014 (UTC)[reply]

    •  Support. Probably it is better to use a less generic name... --Paperoastro (talk) 08:25, 12 June 2014 (UTC)[reply]
    score method? And, I wonder if instance of (P31) could be used for this, as a qualifier? instance of (P31) own goal (Q18533). Danrok (talk) 12:49, 30 June 2014 (UTC)[reply]

    @Filceolaire, Paperoastro, Danrok, Emw: Done as score method (P1443).--Micru (talk) 11:43, 17 August 2014 (UTC)[reply]

    assisted by

       Not done
    Descriptionperson who assisted. Use as qualifier to <goal scored by>
    Data typeItem
    Template parameterput Wikipedia infobox parameters here, if existing; ex: "population" in en:template:infobox settlement
    Domainsports matches
    Allowed valuesteam members.
    Example1966 FIFA World Cup Final (Q1065912) <goal scored by> Geoff Hurst (Q212738) =>Tofiq Bahramov (Q448820)
    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)

    to record details of sports matches. Filceolaire (talk) 20:11, 10 June 2014 (UTC)[reply]

    •  Support --Paperoastro (talk) 08:30, 12 June 2014 (UTC)[reply]
    • Pictogram voting comment.svg Comment Isn't this property get a bit too football-specific (if it's for football only, I think this should be mentioned in the domain)? Though to be honest I'm not sure how detailed Wikidata should/wants to collect such data. Not that someone on one day gets the idea to do a mass-import of basketball games and record all assists for every made basket (this would be a lot :-). --Bthfan (talk) 23:04, 13 June 2014 (UTC)[reply]
    @Bthfan:Assists are also recorded for basket-ball players. Yet, I agree with you : wikidata is not supposed to become a sports database.
    @Filceolaire: Do you know of any wikipedia infobox, list or table that would use this property ?Casper Tinan (talk) 21:32, 15 June 2014 (UTC)[reply]
    Casper Tinan Bthfan This isn't just for football. It's for every team sport which records such data. If there is a fan out there collecting these statistics and wants to donate them to someone who will make them freely available and provide the infrastructure for other fans to build on his work then I think that yes Wikidata could become a sports database, one which can be accessed by anyone for any analysis they want to do, without having to get a license to anyone - or it at least has the potential to do so some day. For now I'm just thinking about the world cup. Filceolaire (talk) 09:14, 16 June 2014 (UTC)[reply]
    •  Oppose We don't need such information. Referee, scorers and minute is enough. -- Blackcat (talk) 13:29, 18 June 2014 (UTC)[reply]
    •  Oppose Too much specific and too difficult to take note of. -- Sannita - not just another sysop 09:54, 20 June 2014 (UTC)[reply]
    • Pictogram voting comment.svg Comment I strongly disagree with the oppose reasons. "Too detailed" or "too specific" aren't good reasons to oppose a property proposal. What if we deleted Wikipedia articles for being too obscure? This is basically the same thing. If some items can use a property and they can't use other properties for the same purpose, I see absolutely no reason why a property should not be created. --Jakob (talk) 20:54, 27 June 2014 (UTC)[reply]

     Not done Per comments use instead contributor to the creative work or subject (P767).--Micru (talk) 11:37, 17 August 2014 (UTC)[reply]


       Done: title (P1476) (Talk and documentation)
    Descriptionshould replace the current P357 (P357) (type: string), with a more appropriate datetype
    Data typeMonolingual text
    Domaincreative works
    Allowed valuesany
    Proposed byZolo (talk)

    Motivation Zolo (talk) 07:39, 7 September 2014 (UTC) <add 2014-09-11>see also Wikidata:Properties for deletion for replacement of p357, P387 (P387), P392 (P392), P438 (P438) </>[reply]

    •  Oppose we can have single language property per item, so that shall be used as default language for all language-related properties. I don't want bunch of properties like "lang+title", "lang+subtitle", "lang+edition title", etc. -- Vlsergey (talk) 16:42, 7 September 2014 (UTC)[reply]
    • Vlsergey: I am not really sure of what you have in mind. We already have "title", "subtitle" etc. properties. Depending on the type of item there should be either one value (the title in the original language, or several, like one title per language). Currently, we specify the language using qualifiers. But now, we can use the "monolingual-string" datatype. The only difference is that language will now be required (which seems to make sense), and stored in the mainsnak's datavalue rather than in the qualifiers (which is more convenient). --Zolo (talk) 07:18, 8 September 2014 (UTC)[reply]
    • @Zolo: for every FRBR edition item we have language of work or name (P407) property that specifies language for all language-dependent properties. I don't like the idea to specify same language value several times in the same entity -- it is big overcomplication and redundancy (for FRBR work we also have original language of film or TV show (P364)). Currently the language can be specified once per entity, and language-dependent values with different language (for example, French name for English book, by the way, it would be original research to define that) can be specified using qualifiers in very rare cases. That is enough. -- Vlsergey (talk) 20:46, 8 September 2014 (UTC)[reply]
    @Vlsergey:, it is true that for items about a particular edition of a text, the language should be inferrable from the language of work or name (P407) of the item. However, in many cases (items about a book with translation, about a move, about an artwork). We need more than one title. In many cases, the label could do the job, but that is less precise and less powerful. For example, if the French title of a movie is different in Canada and in France, wich may happen, it is easier to use a statement (that can have qualifiers) rather than the label. Or if an artwork was given a title by an art historian, we should know in which language it is (for formatting reasons and others). It could be done analyzing the item given in the reference, but that may prove rather tricky. As a rule, I think it makes sense that when a string is in a particular language, this is indicated directly in the datavalue. This is really easier to work with than qualifiers. If the language is really easy to infer, adding it will not be much additional work. And if the language is not easy to infer, this is proof that adding it is useful.
    @Zolo: each translation shall be separate item i.e. FRBR Edition with it's own language of work or name (P407) value, it's own title, editors, publisher, dates, ISBN codes, etc. -- Vlsergey (talk) 07:41, 10 September 2014 (UTC)[reply]
    @Vlsergey: I know, but we also need items for the work, independently of the translation, and actually, this is the majority of our current items (en:War and Peace is not about an individual translation, nor is War and Peace (Q161531) only about the original edition). And for movies, we currently do not use separate items for each subtitled version. For visual artworks, it would just be meaningless (there is just one work, just different titles).--Zolo (talk) 08:28, 10 September 2014 (UTC)[reply]
    • Another argument against this property is titles in artificial languages. For example, what language shall we put for "Woblanadne Ctulhu Ftagn" title? It's obligatory, i.e. we won't be able just omit language, and not, it is not an English (it's only English transcription). We can't specify artificial language neither, because language list is fixed by server properties. -- Vlsergey (talk) 07:39, 10 September 2014 (UTC)[reply]
     Support It'is locigal the have the French mention after the Bonheur d'occasion of The Tin Flute (Q2910080). --Fralambert (talk) 22:15, 7 September 2014 (UTC)[reply]
    We have official name (P1448) for official name and label for local name. What is the use? can you show an example? --ValterVB (talk) 17:56, 8 September 2014 (UTC)[reply]
    @ValterVB: This property should replace P357 (P357), so it would have the same use as it (book titles, movie titles etc.) I would not necessarily be opposed with merging "title" and "official name" but that it is not we have at the moment. --Zolo (talk) 07:25, 10 September 2014 (UTC)[reply]
     Support the change in datatype Filceolaire (talk) 14:16, 16 September 2014 (UTC)[reply]
     Support. --Yair rand (talk) 22:41, 17 September 2014 (UTC)[reply]

    ==> Note This property replaced string property Property:P357, as discussed at Requests for deletions/Archive/2015/Properties/1#P357.

    birth name (monolingual text)

    Status:    Done


    ==> String-type P513 deletion discussion can be found at Wikidata:Requests_for_deletions/Archive/2014/Properties/1#P513.

    script directionality

    Descriptiondirectionality of writing system
    Data typeItem
    Domainwriting systems (alphabets, abugidas, etc.)
    Allowed valuesitems with subclass of (P279)=>script directionality (Q17321709)
    ExampleOsmanya alphabet (Q1377866): LTR
    Source[2], [3], Omniglot
    Robot and gadget jobsRobots can gather info and fill the property with it.
    Proposed byπr2 (t • c)

    Some scripts can be written in multiple directions, so different values could be specified. Also, it may be ambiguous whether TTB and "bottom up" goes from left-to-right or right-to-left, which could be specified as separate items. πr2 (t • c) 19:25, 24 June 2014 (UTC)[reply]

    It's also ambiguous whether LTR and RTL go downwards or upwards in block direction. --Yair rand (talk) 20:13, 24 June 2014 (UTC)[reply]
    @Yair rand: That's true, but I'm not aware of any LTR/RTL that goes upwards. Maybe we need two properties (or a qualifier) for this, unless you can think of a non-confusing way to encode it. The majority of scripts (especially ones still used today) are probably either LTR downwards or RTL downwards. πr2 (t • c) 21:19, 24 June 2014 (UTC)[reply]
    @Yair rand, PiRSquared17: I have added some of the necessary items (see above), more can be created as needed. If there are no objections I'll create this property in the next days.--Micru (talk) 19:26, 7 July 2014 (UTC)[reply]

    Orbits completed

    DescriptionOrbits completed in space missions
    Data typeQuantityValue (dimensionless)-invalid datatype (not in Module:i18n/datatype)
    Template parameteren:Template:Infobox spaceflight
    Domainspace missions
    Allowed valuesintegers
    Proposed byAdert (talk)

    It's very important for template spaceflight Adert (talk) 11:01, 31 January 2014 (UTC)[reply]

    Pictogram voting question.svg Question How will you handle situations where a craft orbits more than one body? --Tobias1984 (talk) 11:13, 31 January 2014 (UTC)[reply]
     Support If necessary, it is possible to use a qualifier to indicate each body that has been orbited. Maybe type of orbit (P522) or create a new one.--Micru (talk) 17:41, 8 March 2014 (UTC)[reply]
     Support Danrok (talk) 18:54, 15 July 2014 (UTC)[reply]


       Not done
    Descriptionlanguage independent or context-specific label
    Data typeItem
    Domainmeta / all
    Allowed valuesall items
    Example<Jacques Chirac (Q2105)> described by source (P1343) <Lentapedia (Q17290934)> with qualifier label <Ширак, Жак>
    Proposed byMicru (talk) 17:12, 8 July 2014 (UTC)[reply]

    Motivation. Sometimes it is needed to use a label as a qualifier of another property, or there is a label that only was valid during a certain period (to be used with qualifier "start/end date"). I also think that it could be used to model systems since inputs and outputs usually require a way of name the signals.--Micru (talk) 17:12, 8 July 2014 (UTC)[reply]

    contained in

       Not done
    Descriptiona material entity located inside another one with no parts in common
    Data typeItem
    Domainmeta / all
    Allowed valuesall items
    Examplethylakoid (Q337889)=>chloroplast membrane (Q5103158) // lung (Q7886) => thoracic cavity (Q994554)
    Proposed byMicru (talk) 13:55, 9 July 2014 (UTC)[reply]

    Basic property.--Micru (talk) 13:55, 9 July 2014 (UTC)[reply]

    •  Oppose. contained_in had been a property of the Basic Formal Ontology (BFO), but it's being removed from BFO 2 per the draft specification in favor of a more generalized location (P276) property. I think that's a move in the right direction; we can use located in to capture the meaning of contained in. Emw (talk) 03:56, 10 July 2014 (UTC)[reply]
    • @Emw: are you sure of that? Looking at "3.6.4 Location" it seems that it has been replaced by occupies_spatial_region which is similar to "contained in". The problem of "located in", also identified in BFO2, is that it needs some prepositional modifier to indicate the location relation between both entities. Instead of "contained in", we could think of a qualifier of "located in", maybe "spatial relation" with values "on", "inside", "above", etc.? --Micru (talk) 07:18, 10 July 2014 (UTC)[reply]

    has role

       Done: no label (P1452) (Talk and documentation)
    Descriptionsomething the subject does because of some special natural, social, or institutional set of circumstances
    Data typeItem
    Domainmeta / all
    Allowed valuesall items
    Examplerudder (Q209760)=>articulated steering (Q499159) // <S5.92 (Q15912780)> source of energy (P618) <dinitrogen tetroxide (Q382984)> with qualifier has role <oxidizing agent (Q187689)>
    Proposed by--Micru (talk) 14:32, 9 July 2014 (UTC)[reply]

    Needed to express which function a given part has in a system (it might be different depending on the system). Notifying @Emw:--Micru (talk) 14:32, 9 July 2014 (UTC)[reply]

    Simply put, roles are things that are optionally manifested in other things. Roles are realized in things by virtue of some special natural, social, or institutional set of circumstances. Roles depend on those things for their existence. For example, roles include any of our various occupations, or the articulated steering (Q499159) of a rudder (Q209760) from Micru's example. One would say that "E. coli has role pathogen", for example, because such bacteria do not have that role when outside an animal's gut. In contrast, one would not say "heart has role circulation", because hearts are typically found in bodies pumping one. (However one might say "heart has role dinner" when being eaten by a lion.) Another example is something like "hydrogen has role nutrient", as stated in the BFO ontology ChEBI, the reference ontology for chemicals of biological interest.
    This brings me to how I would change the 'Description' value of this proposed property. It currently reads "the function an element fulfills in a given system". "Function", "element" and "system" all have significant meaning and would cause alignment problems with BFO ontologies as proposed.
    "Function" in BFO is closely related to "role", but not the same. Functions manifest in something by virtue of its physical structure. "Function" differs from "role" in that the former always or almost always manifests in something by virtue of its physical structure, while the latter optionally manifests in something. For example, the means "heart has function circulation" would be preferred over "heart has role function", because hearts almost always circulate blood, and do so by virtue of their physical structure. Same with "hammer has function impact nails".
    "Element" has strong associations with the set theory operator element of (ϵ). The distinction between an "element" and a "set" is what differentiates the basic membership properties instance of (P31) and subclass of (P279) -- these are often analogized to "element of" and "subset of". I would not want the proposed wording to imply that has role is only applicable for instances and not for classes. BFO permits usage of has role on both instances and classes and the proposed property should do the same.
    Finally, "system" is a vague term and seems likely to produce undue confusion. Certainly Coco Chanel (Q45661) has a specific kind of role, the occupation fashion designer (Q3501317), and Escherichia coli (Q25419) has role pathogen (Q170065), but to say that either subject is "in a given system" is rather obtuse. Things that have a role can be part of a system, but that should not be necessary and so shouldn't be in the description. I think we can safely drop the word "system" from the description.
    I would word the description as "something the subject does because of some special natural, social, or institutional set of circumstances", per BFO.
    Micru, thanks for proposing this property. I think it's likely to become a significant positive outcome of the Wikidata:Requests_for_comment/Refining_"part_of" RFC. Emw (talk) 03:49, 10 July 2014 (UTC)[reply]
    Description changed as per Emw's comment.--Micru (talk) 07:22, 10 July 2014 (UTC)[reply]
    That sounds useful. But as the role/function distinction suggests, that may be tricky to work out. Something about which I was wondering: it would be interesting to have a property that would say the areas of intervention of a particular administrative unit (eduction, security, etc.) Would that be possible to say that with this property or would it pave the way to irredeemable messiness ? @Innocent bystander: --Zolo (talk) 09:04, 10 July 2014 (UTC)[reply]
    @Zolo: "Has role" is for entities that change their function depending on the context. I do not think that an organization adopts different roles, it is created with a fixed set of "functions" in mind and it always fulfills those, for that reason I would recommend using use (P366) or proposing a new "organization function" property if p366 seems too bound to artificial objects.--Micru (talk) 07:30, 11 July 2014 (UTC)[reply]
    • @Yair rand: I have been thinking about this, and I reached the conclusion that the confusion would be normal because... they are the same property. Well, character role (P453) is constrained to "interpretative roles", but there is the question if we can generalize character role (P453) for all uses, or if we can exclude "interpretative roles" from this generic one. Would a "Constraint:Item|NOT property=cast member (P161)" work?.--Micru (talk) 07:30, 11 July 2014 (UTC)[reply]
      I'm afraid I don't see the similarities or associations between the two properties at all. --Yair rand (talk) 00:40, 15 August 2014 (UTC)[reply]
     Support I think this property would be very useful in book cataloging, for qualifying the very various roles of the authors : editor, translator, preface, etc.,
    and yes Micru, organizations can have different roles in the same action, depending on many things, time, money, politics, etc. : an organisation can also be authoring a book... as author, as editor, as publisher... and it may also (for administrations), lead a politic, finance it, coordinate, etc. --Hsarrazin (talk) 15:33, 14 August 2014 (UTC)[reply]

    ✓ Done as P1452 (P1452). Similar to character role (P453), but p453 is constrained to people, while this new property can be used for everything. Notification: @Hsarrazin, Yair rand, Innocent bystander, Zolo, Emw:--Micru (talk) 14:47, 20 August 2014 (UTC)[reply]

    @Hsarrazin, Yair rand, Innocent bystander, Zolo, Emw: Ups, I have just noticed that "has role" is redundant with P794 (P794). I will ask for a merge.--Micru (talk) 21:52, 20 August 2014 (UTC)[reply]
    D - I had not seen it either… just added an alias to make it more explicit… --Hsarrazin (talk) 22:18, 20 August 2014 (UTC)[reply]

    number of parts

       Not done
    Descriptionthe number of parts that this object has in total or of a specific type (when qualifying "has part")
    Data typeNumber (not available yet)
    Domainmeta / all
    Allowed valuesall items
    Example<Fregat (Q1453740)> has part <reaction control system (Q1321971)> with qualifier number of parts <12>
    Proposed by--Micru (talk) 14:40, 9 July 2014 (UTC)[reply]

    Different meaning than quantity (P1114).--Micru (talk) 14:40, 9 July 2014 (UTC)[reply]

    • Pictogram voting comment.svg Comment: two-wheeler (Q233040) currently uses has part (P527) with wheel (Q446): quantity (P1114) = 2 --- Jura 17:26, 9 July 2014 (UTC)[reply]
    •  Oppose. I support the spirit of this property -- capturing cardinality constraints -- but I think we should use a generic cardinality property and merge quantity (P1114) and this proposed property into it. (As a practical matter I would redefine P1114 to have the semantics of that generic property.) That's much closer to how OWL and existing major ontologies handle cardinality constraints. "Cardinality" might be too stuffy as a label for such a property -- perhaps we could get by with 'number' or 'has number', and supply a description that indicates it doesn't take phone numbers? Emw (talk) 04:02, 10 July 2014 (UTC)[reply]

    has agent

       Not done
    Descriptionentity that performs this action
    Data typeItem
    Domainmeta / all
    Allowed valuesall items
    ExampleBoolean function (Q942353)=>logic gate (Q170451)

    With aliases: "performed by", "done by", "executed by".--Micru (talk) 14:14, 11 July 2014 (UTC)[reply]

    transliteration type

       Not done
    Descriptionused as qualifier of any transliterated name or label to indicate the system used
    Data typeItem
    Domainmeta / all
    Allowed valuesSee: w:Transliteration and w:List of ISO romanizations
    Example<Hu Jintao (Q15029)> transliteration <Hú Jǐntāo> with qualifier transliteration type <pinyin (Q42222)>
    Proposed by--Micru (talk) 12:12, 15 July 2014 (UTC)[reply]

    Motivation: Wikidata:Project_chat#How_to_handle_native_language_names.3F--Micru (talk) 12:12, 15 July 2014 (UTC)[reply]

    so what was Hu Jintao's birthname? How many birth names did he have (with different transliterations) and are these really all birth names?
    I agree something needs to be done but we need a solution that works for people with multiple names. I need to think about this. Filceolaire (talk) 22:28, 17 July 2014 (UTC)[reply]
    @Filceolaire: Yes, actually you are right, we'd need a generic transliteration (monolingual string) with this property as qualifier (description changed accordingly). It would be easier to have the name as an independent item, because then we'd be able to put all transliterations for that name together, but that again seems to be in the domain of linguistic data... --Micru (talk) 06:56, 18 July 2014 (UTC)[reply]
    Consider Joseph Stalin (Q855).
    • birth name:იოსებ ბესარიონის ძე ჯუღაშვილი (in Georgian)
      • Transliteration:Ioseb Besarionis dze Jughashvili (English)
      • Transliteration:Iósif Vissariónovich Dzhugashvili (spanish)
    • Known as:Ио́сиф Виссарио́нович Ста́лин
      • Transliteration:Iosif Vissarionovich Stalin (English)
    • Known as:Ио́сиф Ста́лин (Russian)
      • Transliteration:Joseph Stalin (English)
      • Transliteration:José Stalin (Spanish)
      • Transliteration:Josef Stalin (Spanish)
    Qualifiers can't have qualifiers so I'm not sure where to put the 'transliteration type' property.
    it does look like we need a 'known as' property and a 'transliteration' (or maybe 'localisation') property. Filceolaire (talk) 11:50, 18 July 2014 (UTC)[reply]
    @Filceolaire: I don't think that we can do that with qualifiers, because also transliterations may have sources. I have tried to replicate with items your structure here: (talk) 12:47, 18 July 2014 (UTC)[reply]
    Still thinking about this. So basically we have 2 solutions:
    • Each time we want to describe a name, we create a separate item. Which means that if we want to get the Arabic transliteration of Richard Wagner we do it through Richard (Q1249148) and Wagner (Q2539239). There are 2 issues first that it may seem unwieldy and requires creating items even for nicknames and given names that were given to only one person. More importantly, it may not always give a correct result. One case in point is transforming transmogrigying a name into Chinese (not sure it would be called transliteration, but it is needed notheless). For instance, Richard appears to be rendered the differently in Chinese in Richard Wagner and in Richard the Lionheart.
    • Using qualifiers. That seems to require create new properties, like, I guess a string-type "first name" property in addition to the current item-type one. It would also create massive redundancy (repeating the 10s of transliteration of the same first name in 1000s of items). And as Micru notes that may also have issues for sourcing (which actually is an issues I have encountered several times with statements containing qualifiers, so there may be something more general at stake here).
    We could also make a mix of the two, but that may not be very intuitive.
    We could also decide that we can let the software do the job for simple cases (providing standard transliterations of Greek or Russian words for English-speaking users iappears to be pretty mechanical and straightforward). But it would not work for all cases, so that would not really solve anything. -Zolo (talk) 13:24, 18 July 2014 (UTC)n[reply]
    @Zolo: you forgot to mention the third option, which is to treat the whole name as an item, and then transcribe/adapt it as a whole, linking to linguistic data when/if available with qualifiers (see linked items).
    I think we should not enforce any solution, just offer a path for the names of an item to grow organically depending on the number of editors interested in the item: first only labels (as now), then use monolingual properties for transliterations all in the same item, and when the item grows too big, then split it into subitems (see my example).--Micru (talk) 13:37, 18 July 2014 (UTC)[reply]
    When we get a 'monolingualtext' datatype I look forward to getting the 'official name' property and will be proposing a number of other name properties. Also proposing that the datatype for the 'birth name' be changed from 'string'.
    When we get a multilingual text' datatype I will be proposing a 'localised version' property with multilingual text datatype, to use as a qualifier for the name properties. For some languages there may well be more than one commonly used localisation for famous names (e.g. for languages using more than one script). For many small languages the wikidata transliteration will effectively become the default name for that person/country/company since no one else has ever bothered to write about that item in that language before.
    Note that there is still no place to put a 'transliteration type' property but we do get a 'language' qualifier for free for each 'localised version'. Filceolaire (talk) 19:29, 18 July 2014 (UTC)[reply]

    Food plant

       Not done
    DescriptionThe type of plant eaten by a species of animal
    Data typeItem
    Template parameter?
    Allowed valuesPlant species, genus or families
    ExamplePareulype berberata (Q1985138) => Berberis (Q158503)
    Format and edit filter validationPlants
    SourceWikipedia list articles, external databases, etc.
    Robot and gadget jobsPossibly
    Proposed byAndy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits

    Many species, particularly invertebrates, feed exclusively on one species genus or family of plants. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:21, 5 September 2014 (UTC)[reply]

    @Pigsonthewing: Why not use main food source (P1034) to express that? Tobias1984 (talk) 17:09, 5 September 2014 (UTC)[reply]
    Yes, that should do (I looked, but couldn't find anything). Please consider this proposal withdrawn. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:53, 5 September 2014 (UTC)[reply]

    parents of hybrids

    Descriptionparents of hybrids (animal, plant, at whatever rank)
    Data typeItem
    Template parameter|hybrid= in en:Template:Infobox cultivar
    Domainliving things (organisms)
    Allowed valuesnames of organisms
    ExampleMagnolia × soulangeana (Q731443) => Magnolia denudata (Q794061) × Magnolia liliiflora (Q942031)
    Robot and gadget jobsPossible, from infoboxes, list articles or external databases
    Proposed byAndy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits

    Widely used in horticulture. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:34, 8 September 2014 (UTC)[reply]

    The scope should not resticted to plants and species. Thus rename the property to hybrid parents (de: Hybrideltern). To model cases like mule (Q41692) = Equus asinus × Equus caballus and hinny (Q217584) = Equus caballus × Equus asinus we will need an additional qualifer to denote sex (Q290) with the allowed values male organism (Q44148) and female organism (Q43445). --Succu (talk) 16:08, 8 September 2014 (UTC)[reply]

    Achim Raschka (talk)
    Andrawaag (talk)
    Brya (talk)
    CanadianCodhead (talk)
    Dan Koehl (talk)
    Daniel Mietchen (talk)
    Eewilson (talk)
    FelixReimann (talk)
    Hyperik (talk)
    Infomuse (talk)
    Infovarius (talk)
    Jean-Marc Vanel
    Joel Sachs
    Klortho (talk)
    Lymantria (talk)
    Magnefl (talk)
    Mellis (talk)
    Michael Goodyear
    Mr. Fulano (talk)
    Myrmoteras (talk)
    Nis Jørgensen
    Peter Coxhead
    Andy Mabbett (talk)
    Prot D
    Rod Page
    Strobilomyces (talk)
    Stuchka (talk)
    Succu (talk)
    TiagoLubiana (talk)
    Tommy Kronkvist (talk)
    Totodu74 (talk)
    Tris T7 TT me
    William Avery
    Mike Krüger
    GoEThe (talk)
    JJ Ford BHL
    The Anome
    Pictogram voting comment.svg Notified participants of WikiProject Taxonomy --Succu (talk) 18:39, 20 September 2014 (UTC)[reply]

    Succu has a point. - Brya (talk) 19:03, 20 September 2014 (UTC)[reply]
     Support with the amendment of Succu. Tobias1984 (talk) 08:19, 21 September 2014 (UTC)[reply]
     Support as Tobias1984. - Brya (talk) 05:58, 22 September 2014 (UTC)[reply]
    Now of course  Support --Succu (talk) 21:03, 22 September 2014 (UTC)[reply]

    Demonym (item)

       Not done
    Descriptionproperty for item that has the demonym as label. The main item linked should use the .. form as English language.
    Data typeItem
    Template parameter
    "population_demonym" in en:template:infobox settlement
    "gentilé" in fr:Modèle:Infobox_Subdivision_administrative
    Allowed valuesitems with demonyms (see sample)
    Cuba: "Cuban" as currently as English label on item Q16910682, see Wikt:Cuban#English
    Cuba: "cubain" as currently as French label on item Q16910682, see Wikt:cubain#French
    Calgary: "Calgarian" as currently as English label on item Q16910680, see Wikt:Calgarian#English
    SourceWikipedia infoboxes

    Make it easier to find and use them. For each language, the form used as label for items should be defined (e.g. singular masculine adjective). Additional items with qualifiers could be used to store alternate forms. --- Jura 11:44, 19 May 2014 (UTC)[reply]

    Edited the above proposal. --- Jura 14:20, 19 May 2014 (UTC)[reply]
     Support. Emw (talk) 19:03, 31 May 2014 (UTC)[reply]
    Pictogram voting comment.svg Comment someone nuked the sample items, but I guess the proposal can be understood without them. --- Jura 07:48, 22 June 2014 (UTC)[reply]
     Oppose Since "demonym items" seem not to fulfill the notability criteria, I don't think it is a good idea to create this property. Better to wait till there is Wiktionary support, and treat them as linguistic data.--Micru (talk) 08:05, 22 June 2014 (UTC)[reply]
    The items would fulfill the notability criteria because they are linked to this property.
    I don't see how Wiktionary support would be of any help, as it's structured by string sequence rather than meaning. @Micru: Maybe you could explain how this would work. --- Jura 08:12, 22 June 2014 (UTC)[reply]
    With lexical data support we would link to the item containing all declinations and lexical forms. I strongly discourage this property. Instead I recommend of the following:
    • facet of (P1269) for existing items like American (Q4742863), understanding that these items are not meant to represent the word, but the concept of being of a certain place and how that concept came into existence.
    • a mono/multilingual string property for demonyms with appropriate qualifiers
    --Micru (talk) 20:24, 23 June 2014 (UTC)[reply]
    Wiktionary doesn't seem to be on the planning horizon, so we need to find another solution.
    Obviously, we do need to be rigorous with the labels, otherwise automated use of the words isn't possible. Current infoboxes in Wikipedia just use one form, so it shouldn't be much of an issue to define a primary form for this either.
    Some sort of a string property could work. If the above solution seems to complicated, we could just start with the existing string format and add qualifiers to define the language used. --- Jura 05:29, 29 June 2014 (UTC)[reply]
    Jura, that seems more reasonable and the monolingual datatype is almost ready (no need to qualify with language). If there are articles about the demonym, they can be linked with the qualifier statement is subject of (P805). With the datatype change I would support the proposal.--Micru (talk) 08:53, 29 June 2014 (UTC)[reply]
    This property was already created and was later deleted: [4]. 08:34, 26 June 2014 (UTC)[reply]
    It seem to have been defined in a different way, attempting to use existing items for Wikipedia articles. The infoboxes mentioned above generally don't use these. --- Jura 05:29, 29 June 2014 (UTC)[reply]
    Symbol oppose vote oversat.svg Strong oppose. This is exclusively lexical data, and thus not (atm) at all within Wikidata's domain. --Yair rand (talk) 20:55, 1 July 2014 (UTC)[reply]
    @Yair rand: Wouldn't {{Wait}} be a better template to use then? --Jakob (talk) 21:59, 1 July 2014 (UTC)[reply]
    Since the suggested datatype is "Item", I'm not sure it would make sense to consider an eventual property targeting lexical data to be the same as the one proposed here. --Yair rand (talk) 02:56, 2 July 2014 (UTC)[reply]
    Ahh...okay. I  Support the general idea, but we need monolingual text, not an item. --Jakob (talk) 13:06, 2 July 2014 (UTC)[reply]
    • Withdrawn, replaced by the proposal below. --- Jura 13:19, 7 September 2014 (UTC)[reply]

    category for people born here

    Descriptionlink to category entity contains people born in this human settlement
    Data typeItem
    Template parameterborn place (categorization)
    Domainhuman settlement (Q486972) (human settlement)
    Allowed valuesWikimedia category (Q4167836)
    ExampleMoscow (Q649)Category:Births in Moscow (Q16363460)
    Format and edit filter validation
    Sourcewikipedia categorization
    Robot and gadget jobs
    Proposed byVlsergey (talk)

    need for people autocategorization in ruwiki (and other wikis) Vlsergey (talk) 00:39, 4 September 2014 (UTC)[reply]

    category for people died here

    Descriptionlink to category entity contains people died in this human settlement
    Data typeItem
    Template parameterdeath place (categorization)
    Domainhuman settlement (Q486972) (human settlement)
    Allowed valuesWikimedia category (Q4167836)
    ExampleMoscow (Q649)Category:Deaths in Moscow (Q6378606)
    Format and edit filter validation
    Sourcewikipedia categorization
    Robot and gadget jobs
    Proposed byVlsergey (talk)

    need for people autocategorization in ruwiki (and other wikis) Vlsergey (talk) 00:41, 4 September 2014 (UTC)[reply]

    phone number / toll free phone number / fax (en)

       Done: no label (P1244) (Talk and documentation)
    Data typeURL
    Proposed byGZWDer (talk) 10:39, 18 June 2013 (UTC)[reply]
    I don't understand your proposal. Why do you want to use the "mulitlingual text" datatype? For telephone numbers the IRI-Datatype should be used as far I know. Otherwise I would use the string datatype, because there is nothing multilingual in phone numbers. --Pyfisch (talk) 13:37, 18 June 2013 (UTC)[reply]
    My mistake.--GZWDer (talk) 02:41, 19 June 2013 (UTC)[reply]
    Pictogram voting comment.svg Comment phone numbers having nothing to do with IRIs. --Danrok (talk) 00:19, 17 October 2013 (UTC)[reply]
    •  Support IRI can be used to express a phone number. In the far future, a "phone number" datatype would be great, since Wikivoyage often gets invalid phone numbers: many people insert a homepage URL instead, even though we have a separate field for the homepage URL (and these URLs are valid IRIs, so just IRI would not catch the mistake). Nicolas1981 (talk) 08:32, 28 November 2013 (UTC)[reply]
    •  Support--Konggaru (talk) 16:33, 3 January 2014 (UTC)[reply]
     Support -- DerFussi 11:26, 24 January 2014 (UTC)[reply]
    •  Support --NatiSythen (talk) 07:59, 28 January 2014 (UTC)[reply]
    •  Support Maybe it makes sense to distinguish between land-line and mobile phone and fax numbers, too. In any case we need a rule how to handle multiple numbers. And how to write it to the database, so that the numbers can be "translated" in individual language branches. A "phone number" datatype would be useful. --RolandUnger (talk) 13:10, 2 February 2014 (UTC)[reply]
    • @GZWDer, Pyfisch, Filceolaire, Danrok, Nicolas1981, DerFussi: @NatiSythen, RolandUnger: ✓ Done, but I cannot seem to add the property to an item. --Jakob (talk) 13:44, 8 April 2014 (UTC)[reply]
      Me neither, it seems the tel: URL scheme is not supported. Maybe it's better to change the datatype to string. Mushroom (talk) 10:29, 10 April 2014 (UTC)[reply]
      It should not be to difficult to add tel: sheme, maybe someone can just ask the developers to add it. Best case they only need to change a single line of code. --Pyfisch (talk) 07:09, 19 April 2014 (UTC)[reply]
    • Pictogram voting comment.svg Comment: In the meantime, shouldn't this be created as a string property? Property:P1244 was created over a month ago, but can't be used. --- Jura 06:47, 16 May 2014 (UTC)[reply]
    Pictogram voting comment.svg Comment Later a plain string property got created:
    Status:    Done

    Wikivoyage IsPartOf

       Not done
    Descriptionto port data from Q14205332.
    Data typeItem
    ExampleUnited Kingdom (Q145)=>British Isles (Q38272)
    Robot and gadget jobscan be imported from Wikivoyage and located in the administrative territorial entity (P131).
    Proposed byGZWDer (talk)

    Motivation. GZWDer (talk) 12:25, 20 January 2014 (UTC)[reply]

     Support --Zhuyifei1999 (talk) 12:29, 20 January 2014 (UTC)[reply]
     Oppose The example is faulty; several dominions that are part of the United Kingdom are not part of the British Isles. This information is also not specific to Wikivoyage. Thanks, GerardM (talk) 12:39, 20 January 2014 (UTC)[reply]
    So this will only be used in Wikivoyage.--GZWDer (talk) 12:43, 20 January 2014 (UTC)[reply]
     Support -- DerFussi 11:24, 24 January 2014 (UTC) - But the information has to be specific to each language version. -- DerFussi 05:23, 31 January 2014 (UTC)[reply]
    •  Oppose You can not do that: the Control of Contents in WV-de belongs to WV-de and must be discussed there, not here in English. I am Bbb of WV-de and have no guilty account here.-- 12:30, 30 January 2014 (UTC)[reply]
      The usage of that information is still the decision of the WV-de community. We can, but dont have to us it on de. -- DerFussi 05:21, 31 January 2014 (UTC)[reply]
    •  Support IsIn and IsPartOf are identical in a technical sense. Maybe the naming IsIn is better because this feature can be used not only in the main namespace. This information has to be specific to each language branch. Maybe it is useful to distinguish between one primary and obligatory IsIn and a random number of secondary IsIns. For instance Garmisch-Partenkirchen is primarily in Bavaria and secondarily in the Alps. The Alps are in Europe, and it is not possible to put it in a single bread-crumb trail with Bavaria or Germany. --RolandUnger (talk) 13:10, 2 February 2014 (UTC)[reply]

    wikinews portal (en)

       Not done
    Descriptionthe name of the relevant news portal on wikinews for this item
    Data typeMultilingual text (not available yet)
    Examplescience (Q336) => Science and technology (English)
    Sourcetemplates in wikipedia articles
    Robot and gadget jobsbots can find this in templates within articles on wikipedia
    Proposed byDanrok (talk) 21:53, 30 August 2013 (UTC)[reply]

    Motivation. Danrok (talk) 21:53, 30 August 2013 (UTC)[reply]

    Work period (from) (en)

       Not done
    Data typePoint in time
    Example 1MISSING
    Example 2MISSING
    Example 3MISSING
    Proposed byGZWDer (talk) 10:48, 18 June 2013 (UTC)[reply]
    •  Support Filceolaire (talk) 23:20, 18 August 2013 (UTC)[reply]
    •  Oppose use 'occupation' with start date/end date. Filceolaire (talk) 06:15, 25 February 2014 (UTC)[reply]
    •  Not done No consensus; discussion has been open for almost a year. --Jakob (talk) 15:27, 11 April 2014 (UTC)[reply]

    Work period (until) (en)

       Not done
    Data typePoint in time
    Example 1MISSING
    Example 2MISSING
    Example 3MISSING
    Proposed byGZWDer (talk) 10:48, 18 June 2013 (UTC)[reply]

    GUI toolkit / GUI framework

    Descriptionframework or toolkit a program uses to display windows
    Data typeItem
    Allowed valuesitems should be instance/subclass of GUI framework
    ExampleGIMP (Q8038) => GTK (Q189464)
    Format and edit filter validationsingle value
    Sourceofficial websites of programs
    Proposed by-Tobias1984 (talk) 22:59, 18 June 2014 (UTC)[reply]
    • @Jakec: - Most of the toolkits already have items (en:List of widget toolkits). And I think connecting the programs with the toolkits is notable too. It is a frequent discussion in programmer circles which toolkit is used for which program. And none of the toolkits is 100 % plattform independent. Platforms are going for a native look and usage paradigm which some toolkits do better than others. Some examples:
    •  Support. I changed "gui" to "GUI" to reflect that it's an acronym, and thus typically capitalized in English. Emw (talk) 23:16, 10 July 2014 (UTC)[reply]
    •  Support --[[kgh]] (talk) 13:09, 11 July 2014 (UTC)[reply]
    • ✓ Done -Tobias1984 (talk) 16:26, 11 July 2014 (UTC)[reply]


       Not done
    Descriptiondescription how heavy an vehicle is
    Data typeNumber (not available yet)
    Template parameterput Wikipedia infobox parameters here, if existing; ex: "population" in en:template:infobox settlement
    Domainaircraft (Q11436) (aircraft), BMW F 800 (Q796496)(MotorcycleModel), motorcycle (Q34493)(Motorcycle), motor car (Q1420) (automobile)
    Allowed valuespounds or kg
    Examplethe motorcycle has a mass of 200kg
    Format and edit filter validation(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
    Robot and gadget jobsShould or are bots or gadgets doing any task with this? (Checking other properties for consistency, collecting data, etc.)
    Proposed by194.121.90.163

    Every vehicle has a mass therefore we should have a matching property. Also the themplate has a mass property 13:43, 1 July 2014 (UTC)[reply]

    see also -- 10:15, 18 July 2014 (UTC)[reply]


       Done: no label (P1328) (Talk and documentation)
    DescriptionAmount people working for an entity
    Data typeNumber (not available yet)
    Allowed valuesPositive whole number
    Example23,056 employees
    Robot and gadget jobsGlobal Economic Map Bot can do this
    Proposed byMcnabber091 (talk) 19:01, 14 April 2014 (UTC)[reply]
     SupportMcnabber091 (talk) 10:11, 16 May 2014 (UTC)[reply]
     Support --- Jura 14:24, 23 May 2014 (UTC)[reply]
     Support Amir (talk) 21:32, 23 May 2014 (UTC)[reply]
    Pictogram voting comment.svg Comment This was a duplicate of employees (P1128). It seems people are not aware of properties. Why? — Finn Årup Nielsen (fnielsen) (talk) 09:42, 9 June 2014 (UTC). Edited: added a 'not' — Finn Årup Nielsen (fnielsen) (talk) 15:24, 14 July 2014 (UTC)[reply]
     Oppose. Duplicate property.-CCD (talk) 18:01, 5 September 2014 (UTC)[reply]

    Maximum glide ratio / beste Gleitzahl / Finesse

    DescriptionA measure of the glide path in calm air.
    Data typeNumber - unitless.-invalid datatype (not in Module:i18n/datatype)
    ExampleBoeing 767-200 with fuel exhaustion = ~12
     Support - all flyable things need this in my opinion. --Tobias1984 (talk) 14:01, 16 May 2013 (UTC)[reply]
     Support - except baloons. Could go ahead now, as it's unitless. 20:55, 12 September 2014 (UTC)[reply]

    reporting mark

       Done: reporting mark (P1471) (Talk and documentation)
    DescriptionA reporting mark is an alphabetic code of one to five letters used to identify owners or lessees of rolling stock and other equipment used on certain railroad networks.
    Data typeString
    Template parameter"marks" in en:template:infobox SG rail
    Domainrailroads, railways
    ExampleNew York, New Haven and Hartford Railroad (Q1472738) => NH
    Proposed byDanrok (talk) 01:08, 9 January 2014 (UTC)[reply]
    How about using station code (P296) instead? Is there difference between them? --Konggaru (talk) 04:39, 19 January 2014 (UTC)[reply]
    Pictogram voting comment.svg Comment they are different. Danrok (talk) 20:42, 26 January 2014 (UTC)[reply]
     Support --Konggaru (talk) 09:16, 28 January 2014 (UTC)[reply]
     Support Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:06, 4 September 2014 (UTC)[reply]

    cause of

       Done: has effect (P1542) (Talk and documentation)
    Descriptioneffect of this cause. When causally immediate effects exist, use 'immediate cause of' instead, and use this property for causally distant effects.
    Data typeItem
    Template parameter'result' in Template:Infobox_military_conflict, possibly others. Notify participants in property proposal discussion before any automatic import.
    DomainNone. Often but not always events or processes. See examples and discussion.
    Allowed valuesAny. Often but not always events or processes. See examples and discussion.
    ExampleAmerican Civil War (Q8676) cause of Thirteenth Amendment to the United States Constitution (Q175613), Reconstruction Era (Q623429)
    SourceExternal references
    Proposed byEmw (talk) 03:53, 15 September 2014 (UTC)[reply]

    This would be an inverse property of has cause (P828). To make that apparent, the label would follow the 'has X' and 'X of' pattern used in properties like has part (P527) and part of (P361). It would have the aliases 'has effect', 'has result', 'outcome of', and others.

    Like other inverse properties, this property, 'immediate cause of' and 'contributing factor of' would require editors to use common sense. For example, this means not putting 15 million 'cause of' claims in Big Bang (Q323) or physics (Q413). When not used in clearly inappropriate ways this generic property to capture causally downstream information would be quite useful. Emw (talk) 03:53, 15 September 2014 (UTC)[reply]

    •  Support John Vandenberg (talk) 03:09, 19 September 2014 (UTC)[reply]
    •  Support Danneks (talk) 19:30, 20 September 2014 (UTC)[reply]
    • Pictogram voting comment.svg Withdrawing proposal. Sorry, it just occurred to me that for consistency in how P828 was relabeled from 'medical causes' to 'has cause' to facilitate a broader approach to modeling causes, so P780 would make sense to relabel from 'symptoms' to 'cause of'. That would make this property proposal redundant. Please see Property_talk:P780#symptoms_-.3E_cause_of. Emw (talk) 14:14, 21 September 2014 (UTC)[reply]
    • Pictogram voting comment.svg Update. One of the editors in the linked discussion prefers keeping the property labeled 'symptoms'. See Property_talk:P780#Emw_2014-09-23-1219 for an overview. Please comment there. Emw (talk) 12:25, 23 September 2014 (UTC)[reply]

    Orphanet ID

       Done: Orphanet ID (P1550) (Talk and documentation)
    DescriptionID in the database
    Data typeString
    Template parameter?
    Domainpharmaceutical drugs and diseases
    Allowed valuesany valid ID
    Examplepneumocystosis (Q7205993) = "723" (ORPHA723)
    Format and edit filter validationUnique ID/Single ID
    Robot and gadget jobsIf possible.
    Proposed by-Tobias1984 (talk) 06:27, 1 October 2014 (UTC)[reply]


       Done: UBERON ID (P1554) (Talk and documentation)
    DescriptionID from UBERON ontology
    Data typeString
    Template parameter?
    Domainanatomy (Q514), body (Q170494)
    Allowed valuesall valid IDs
    Examplehead (Q23640) = '0000033' (UBERON:0000033)
    Format and edit filter validationUnique ID (UBERON:7 digits)
    Robot and gadget jobsthere will be
    Proposed byEmitraka (talk)
    Discussion Emitraka (talk) 18:09, 6 October 2014 (UTC)[reply]

    • ✓ Done Created it sooner, because it has enough support from the WikiProject. -Tobias1984 (talk) 06:21, 7 October 2014 (UTC)[reply]

    MacTutor id

    Descriptionid of a person in MacTutor History of Mathematics archive
    Data typeString
    DomainQ5 ?
    Proposed byZolo (talk)

    Widely used, see Template:MacTutor (Q10623208) Zolo (talk) 08:38, 5 October 2014 (UTC)[reply]

    •  Support, reliable sources for biographies of mathematicians, full list here: [5], Bezik (talk) 09:43, 5 October 2014 (UTC)[reply]
    •  Support, useful for references. Danneks (talk) 12:52, 6 October 2014 (UTC)[reply]
    •  Support Per Danneks. --Eurodyne (talk) 00:45, 10 October 2014 (UTC)[reply]
    @Bezik, Danneks, Eurodyne: done: MacTutor biography ID (P1563)--Zolo (talk) 08:00, 13 October 2014 (UTC)[reply]

    At the Circulating Library - Author (or AtCL Author)

    DescriptionAuthor information on At the Circulating Library database of victorian fiction, 1837-1901.
    Data typeString
    Allowed valuesnumbers
    ExampleFrank Barrett (Q18190913) => 343 (
    Proposed byLugusto (talk)

    Found while researching for birthday/date of death data for en.Wikisource. Contains lots of minor writers without information on similar sites. The example above, for example, was not found on VIAF. Lugusto (talk) 18:54, 5 October 2014 (UTC)[reply]

     Support --Kolja21 (talk) 19:55, 5 October 2014 (UTC)[reply]
     Support --Eurodyne (talk) 19:44, 6 October 2014 (UTC)[reply]
    @555, Kolja21, Eurodine: done: At the Circulating Library ID (P1564)--Zolo (talk) 08:01, 13 October 2014 (UTC)[reply]

    end reason

       Done: end cause (P1534) (Talk and documentation)
    Descriptionqualifier used to supplement "end date"
    Data typeItem
    Allowed valuesall
    ⟨ Goldie Hawn (Q188459)  View with Reasonator View with SQID ⟩ spouse (P26) View with SQID ⟨ Bill Hudson (Q2742575)  View with Reasonator View with SQID ⟩
    with qualifier end reason divorce (Q93190)
    SourceWikipedia infobox or article text
    Proposed byCbrown1023 talk

    The main inspiration for this qualifier proposal is including an explanation of why a marriage ended (e.g., divorce (Q93190), legal annulment of marriage (Q701040)), something often kept track of in infoboxes but not included in Wikidata so far. We can definitely broaden it to include any more information about why something started or ended though, such as if someone transferred schools, dropped-out, etc. Cbrown1023 talk 02:30, 6 August 2014 (UTC)[reply]

    If this is to be used for many properties, I don't think that "reason" is the right word here. --Yair rand (talk) 22:31, 6 August 2014 (UTC)[reply]
    What do you suggest? I couldn't come up with a better name either, but "end reason" is the best one we came up with. If we agree it's a good idea, though, the good thing about Wikidata being a wiki is that we can always change the name or add new aliases. Cbrown1023 talk 01:32, 7 August 2014 (UTC)[reply]

    has quality

       Done: has quality (P1552) (Talk and documentation)
    Descriptionthe entity has an inherent or distinguishing characteristic that cannot be migrated
    Data typeItem
    Allowed valuesall
    Exampleposthumous birth (Q2106128) => posthumous (Q1395509)
    Proposed byMicru (talk)

    used by

       Done: used by (P1535) (Talk and documentation)
    Descriptionthe person or organization that uses an object or a facility
    Data typeItem
    Allowed valuesall
    ExampleMudra barracks (Q15121187) => Personalamt der Bundeswehr (Q575684)
    Proposed byMicru (talk)
    • Generic property for linking the entity being used and agent using it (the above "has agent" links an action with the agent performing it).--Micru (talk) 11:04, 17 August 2014 (UTC)[reply]
     Support. Emw (talk) 17:58, 7 September 2014 (UTC)[reply]
     Support --Eurodyne (talk) 22:27, 24 September 2014 (UTC)[reply]

    immediate cause of

    Descriptioneffect of this cause. When causally immediate effects exist, use 'immediate cause of' instead, and use this property for causally distant effects.
    Data typeItem
    Template parameter'result' in Template:Infobox_military_conflict, possibly others. Notify participants in property proposal discussion before any automatic import.
    DomainNone. Often but not always events or processes. See examples and discussion.
    Allowed valuesAny. Often but not always events or processes. See examples and discussion.
    ExampleAmerican Civil War (Q8676) immediate cause of assassination of Abraham Lincoln (Q1025404)
    SourceExternal references
    Proposed byEmw (talk) 03:53, 15 September 2014 (UTC)[reply]

    Inverse of 'has immediate cause'. See notes in property proposals for 'has immediate cause' and 'cause of' above. Emw (talk) 03:58, 15 September 2014 (UTC)[reply]

    contributing factor of

    Descriptionthing that is significantly influenced by this cause, but does not directly result from it
    Data typeItem
    Template parameterNone.
    DomainNone. Often but not always events or processes. See examples and discussion.
    Allowed valuesAny. Often but not always events or processes. See examples and discussion.
    Exampleobesity (Q12174) contributing factor of heart disease (Q190805)
    SourceExternal references
    Proposed byEmw (talk) 04:10, 15 September 2014 (UTC)[reply]

    Inverse of 'has contributing factor'. See discussion in 'has immediate cause' and 'cause of' above for details. Emw (talk) 04:10, 15 September 2014 (UTC)[reply]

    Demonym (monoligual text)

       Done: demonym (P1549) (Talk and documentation)
    DescriptionDemonym. For English, the singular masculine adjective form is preferred. In English, this should start with a Capital letter. Add a qualifier if another form is added.
    Data typeMonolingual text
    Template parameter
    "population_demonym" in en:template:infobox settlement
    "gentilé" in fr:Modèle:Infobox_Subdivision_administrative
    Allowed valuesdemonym
    Cuba: "Cuban" (en), see Wikt:Cuban#English
    Cuba: "cubain" (fr), see Wikt:cubain#French
    Calgary: "Calgarian" (en), see Wikt:Calgarian#English
    SourceWikipedia infoboxes
    • This replaces the previous proposal of a property with item datatype. --- Jura 13:19, 7 September 2014 (UTC) @Jakec, Emw:[reply]
    1.  SupportAyack (talk) 19:40, 7 September 2014 (UTC)[reply]
    2. Pictogram voting comment.svg Comment This property will be more suited as a multilingual text for me. Demonym are not necessarelly frozen in one language, expecially for a country. --Fralambert (talk) 23:24, 7 September 2014 (UTC)[reply]
    3.  Oppose. This is Wiktionary's domain. This should wait until we have a system for dealing with complex (very complex) linguistic data. --Yair rand (talk) 02:16, 8 September 2014 (UTC)[reply]
      • Wikipedia generally includes this information (see infobox samples above). --- Jura 06:08, 11 September 2014 (UTC)[reply]
        • I am aware of that. Still, this data is well beyond what Wikidata is currently capable of. Demonyms are words, and words are complicated. Some languages can have hundreds of different forms for a demonym, and there are thousands of languages we can cover, and these are interlinked in strange ways. Simple monolingualtext won't work well at all. --Yair rand (talk) 23:38, 15 September 2014 (UTC)[reply]
    4. Pictogram voting comment.svg Comment I came here to propose this myself. It's relevant to us in addition to Wiktionary, as evidenced by the fact that it's used in infoboxes (for example, see w:en:Halifax, Nova Scotia and its unusual demonym "Haligonian"). But Fralambert does make a good point that it would have to be multilingual, so now I'm torn. --Arctic.gnome (talk) 15:55, 12 September 2014 (UTC)[reply]
      • Good question indeed. For many places demonyms might only exists in one or two local languages. This property can easily hold entries for several languages. A solution that requires translation of the label in every conceivable language might not be that practical. This was the disadvantage of the earlier "item" property. A multilingual property is likely to have the same drawback. --- Jura 08:41, 13 September 2014 (UTC)[reply]
        • I guess most places (other than countries and world-cities) will only have a commonly-used demonym in the local language. Where demonyms exist in two or three local languages, we could add both of them with the language as a qualifier, like I did with legal citation of this text (P1031) on Supreme Court Act (Q7644846). However, countries and world-cities have demonyms in every language, so maybe we should change the name of this parameter to "local demonym"; either that or we wait until compatibility with Wiktionary so we can link to their list of all demoyms. --Arctic.gnome (talk) 20:17, 13 September 2014 (UTC)[reply]
    5.  Support In Wikipedia in Spanish we are creating a Infovox only gets values ​​Wikidata. This is necessary for such a system is feasible --Miguillen (talk) 08:33, 13 September 2014 (UTC)[reply]
    6. GA candidate.svg Weak support Not totally support, I would prefert a multiligual text property. --Fralambert (talk) 22:08, 28 September 2014 (UTC)[reply]

    motto description

       Done: motto (P1546) (Talk and documentation)
    Descriptiondescription of a subject's motto (Q42470)
    Data typeItem
    Template parameteren:Template:Infobox country, field national_motto (see any Template:Infobox country (Q5621162))
    DomainInstances of administrative territorial entity (Q56061) and other organization (Q43229)
    Allowed valuesInstances of motto (Q42470)
    ExampleCanada (Q16) motto description A Mari Usque Ad Mare (Q41423)
    Proposed by LaddΩ chat ;)

    Similar to has seal, badge, or sigil (P418), but for linking to the description of a motto. The latter will be soon created as well, since the monolanguage datatype is about to be deployed. LaddΩ chat ;) 00:17, 16 August 2014 (UTC)[reply]

    • If we can link to the actual item about the motto, why couldn't we use motto text (P1451) on the motto item, rather than on the place item itself? --Ricordisamoa 02:07, 1 September 2014 (UTC)[reply]
    @Ricordisamoa: Indeed, property motto text (P1451) could well be used on motto items themselves, but how many mottos do actually have an item? I checked a few users of en:Template:Infobox country and all were displaying the motto in text, with no associated item. LaddΩ chat ;) 03:50, 1 September 2014 (UTC)[reply]
    There are 170 articles in en:Category:Mottos and subcategories. I dislike the "X description" label format; pointing to the item is supposed to be referring to the entity itself. If we need separate properties for these, this should be "motto" and P1451 should be "motto text". --Yair rand (talk) 02:15, 8 September 2014 (UTC)[reply]

    Exceptionnal heritage of Wallonia ID

    DescriptionThe unique identifier of the fr:Patrimoine immobilier exceptionnel de la Région wallonne.
    Data typechaîne de caractères-invalid datatype (not in Module:i18n/datatype)
    Template parameterfr:Modèle:Exceptionnel Wallonie 4=
    DomainExceptional heritage property of Wallonia (Q2218240)
    Allowed valuestype(s) d'éléments liés (modèle Q ou texte), liste ou intervalle de valeurs permises, format de chaîne de caractères...
    ExampleCharleroi City Hall (Q2477151) => 52011-PEX-0001-02
    Proposed byFralambert (talk)

    When I was asking for Classified properties and protected areas of Wallonia ID (P1133), I believed that the exceptional and classified properties avec the code. I realised that the web link are different. Fralambert (talk) 16:35, 1 September 2014 (UTC)[reply]

    •  Support --- Jura 22:15, 27 September 2014 (UTC)[reply]
     Support -Tobias1984 (talk) 21:22, 28 September 2014 (UTC)[reply]
    @Jura1, Tobias1984: Done, tanks for your approval. --Fralambert (talk) 23:10, 6 October 2014 (UTC)[reply]

    Russian cultural heritage register identifier

    DescriptionID of the cultural heritage on
    Data typeString
    Domaincultural heritage site in Russia (Q8346700)
    Allowed values10-digits numbers with leading nulls, single value, unique
    ExampleRussian Museum (Q211043) => 7810526000
    Format and edit filter validation\d{10}
    Source, ru:Шаблон:Культурное наследие РФ, ru:Шаблон:Культурное наследие РФ 2, ru:Шаблон:Культурное наследие РФ 3, ru:Шаблон:Культурное наследие РФ 4
    Robot and gadget jobsPBot will import values from Russian Wikipedia
    Proposed by--putnik

    The site is not really the official database of the cultural heritage of Russia, but it was created with the support of the goverment and contains the most complete database available for non-specialists. As identifier of the cultural heritage is commonly used this one. --putnik 08:59, 18 September 2014 (UTC)[reply]

    •  Support but it's better to import data from Wikipedia, not from site itself (due to legal rights) -- Vlsergey (talk) 09:02, 18 September 2014 (UTC)[reply]
    •  Support As other monument database. --Fralambert (talk) 12:28, 18 September 2014 (UTC)[reply]
    @Putnik, Vlsergey: ✓ Done --Fralambert (talk) 22:12, 21 September 2014 (UTC)[reply]

    Glad identifier

    Descriptionvery similar to Mérimée ID (P380) and Palissy ID (P481) but for another database : Q18107552, wich have ~100000 items but limited to the listed in the general inventory of cultural heritage (Q16739336) in Brittany (Q12130)
    Data typeString
    Template parameternone, but there is fr:Modèle:Glad
    Domainlisted in the general inventory of cultural heritage (Q16739336)
    Allowed valueslength : 17 caracters ; allways start by MERIMEEIA or PALISSYIM and followed by 8 digits (sometimes the number are the same as the Mérimée ID (P380) and Palissy ID (P481) but not always): MERIMEEIA\d{1,8}|PALISSYIM\d{1,8}
    Exampleall the ~100000 listed in the general inventory of cultural heritage (Q16739336) in Brittany (Q12130) (listed in the general inventory of cultural heritage (Q16739336) is a very wide concept, it could be buildings or objects):
    Robot and gadget jobssame kind of constraint Property talk:P380 : {{Constraint:Format}}, {{Constraint:Unique value}}, {{Constraint:Item|property=P31|items={{Q|16739336}} }} (has to be discussed for other constraints)
    Proposed byVIGNERON (talk)

    Could be very useful for listed in the general inventory of cultural heritage (Q16739336) in Brittany (Q12130) ; the Q18107552 is much more precise than Base Mérimée (Q809830) and Palissy database (Q2886424). VIGNERON (talk) 07:22, 22 September 2014 (UTC)[reply]

    1.  SupportAyack (talk) 10:22, 22 September 2014 (UTC)[reply]
    2.  Support It can also be usefull for registered historic monument (Q10387575) of monument historique classé (Q10387684) like providing a source for coordinates. VIGNERON Does it indicate if a element is part of a Historical Monument (Q916475) or site naturel classé ou inscrit (Q18002388)? --Fralambert (talk) 13:11, 22 September 2014 (UTC)[reply]
      @Fralambert: : yes, here you can see the list of items in Glad by protection. Cdlt, VIGNERON (talk) 14:04, 22 September 2014 (UTC)[reply]
    3.  SupportMathieudu68 talk 16:33, 22 September 2014 (UTC)[reply]

    @VIGNERON, Fralambert, Mathieudu68: ✓ Done: Gertrude identifier (P1529)Ayack (talk) 16:56, 22 September 2014 (UTC)[reply]

    manifestation of

    Descriptionthis item is a physical embodiment of an item representing information
    Data typeItem
    Allowed valuesmainly works
    Examplebanknote (Q47433) => money (Q1368)
    Proposed byMicru (talk)

    Necessary property for modeling w:FRBR levels. Not necessary all the time, but in some cases it might be useful to model the manifestation separately from the expression. --Micru (talk) 08:23, 19 August 2014 (UTC)[reply]

    Could you show at least one such case, please (for works)? Danneks (talk) 18:17, 19 August 2014 (UTC)[reply]
    @Danneks: There are no items like that at the moment, because it is for special cases, but I created one for this example:
    • work item: Iliad (Q8275), all Iliad versions
      • edition item: Iliad (Q17559055), a Catalan translation by Conrad Roure
        • manifestation item 1: Qxxx, the original manuscript by Conrad Roure kept at the library
        • manifestation item 2: Qxxx, a print run of 300 exemplars with 300 pages
        • manifestation item 3: Iliada (Q17559097), special print run of 30 exemplars with 336 pages
    This is of course hypothetic, normally it would be enough having only the work and edition item, only when more than one manifestation is added, then it would be needed to split it.--Micru (talk) 19:41, 19 August 2014 (UTC)[reply]
    @Micru: Thanks. I'm not sure that we need distinct items for different manifestations of the same edition, we are not a library after all :) Alternatively, we can create "text" property with links to Wikisource, like image (P18) and document file on Wikimedia Commons (P996). And I think that text can be translated unambiguously into many more languages than manifestation of. Danneks (talk) 09:32, 20 August 2014 (UTC)[reply]
    @Danneks: Yes, that could be another option, but if it is just to link with the "Wikisource index page" (the text is a sitelink) I wouldn't call it "text", but perhaps simply "Wikisource index page" which is what it is :) I'm also thinking that such a property could be used as qualifier of "scan file".--Micru (talk) 11:15, 21 August 2014 (UTC)[reply]
    And what about non-bibliographic items? How do we link classes like banknote (Q47433) => money (Q1368)? Do we need a more user-friendly name, like "representation of"? Or can we perhaps (ab)use "instance of"? Pinging @TomT0m: to see if he has some idea.--Micru (talk) 11:15, 21 August 2014 (UTC)[reply]
    Well, money is defined as any object generally accepted as payment, and banknotes are clearly a subclass of such objects. Danneks (talk) 11:47, 21 August 2014 (UTC)[reply]
    There are two "money" concepts, one is "object accepted as payment", and the other "record accepted as payment", I guess you are referring to the first one which is clear, but I am more interested in the second. Or better said, how to connect both concepts in a better way than having both merged in one item.--Micru (talk) 12:25, 21 August 2014 (UTC)[reply]
    Sorry, but I don't understand you at all. Isn't record an object (Q488383)? Or maybe you can give reference to the fact that there are two money concepts... Danneks (talk) 12:46, 21 August 2014 (UTC)[reply]
    See the lead text of enwp: the main functions of money are distinguished as: a medium of exchange; a unit of account; a store of value; and, occasionally in the past, a standard of deferred payment. Any item or verifiable record that fulfills these functions can be considered money. By "item" it is meant a physical entity, and by "record" it is meant an information entity.
    Information is what happens when you perceive something, and a physical entity happens independently of perception. To put it briefly, "a banknote" is always happening, but the "realization that the banknote is worth 10EUR" only happens when someone perceives it and decodes its message.
    object (Q488383) has the connotation of something perceived by an observer, not of something happening independently of perception.--Micru (talk) 13:09, 21 August 2014 (UTC)[reply]
    Well, it seems that for different philosophers, "object" has different connotations... Personally, I like the opinion "object=entity". So I think that "banknote" does not exist unless someone will say that it is a banknote, and its "atoms" do not exist unless someone will say that they are atoms... We can say that they "correspond to matter", but the problem is that we can't say anything about matter itself... And I can say that "money" correspond to some matter too, I have some material money :). So I think that having the "instance of" property would be enough for all purposes, unless you will show a convincing counter-example. Danneks (talk) 15:12, 21 August 2014 (UTC)[reply]
    @Micru: This property seems a little bit abstract to be OK with only one example to me. I'm not sure I understand this property. For the money problem, I would say it's easier to say that each coin is an object with a facial value, the facial value would be a number with a currency type. It seems easier to model, a property facial value, a qualifier property for the currency (or converse). For editions, how is this different from using instance of (P31) ? TomT0m (talk) 16:40, 21 August 2014 (UTC)[reply]
    @TomT0m: Well if you are fine using instance of to link class-to-class, then I'm fine with <banknote (Q47433)> instance of <money (Q1368)>. Another example could be <human> instance of <person>, or <teacher> instance of <profession>.--Micru (talk) 09:54, 22 August 2014 (UTC)[reply]
    @Micru: You're one of the ones who like controversies, aren't you ? :) You can't do that without reflexion. Both human and persons are class of real world objects. This means if you take the set of all persons and the set of all humans, the set of all humans is a subset of all persons. This means subclass of (P279) must be used. Now if you take a concept like clade, this is different. The set of all clade is not a set of real world objects, like the set of all humans. Actually <human> is an example of a clade. So the set of all clades is a set of sets of organisms, I will call this kind of class a metaclass. Classes can be members of metaclasses, but not subsets of metaclasses. We should think "member of the set of = instance of (P31)" and "subset of = subclass of (P279). When I talk of artwork, I use a little trick allowed in the OWL language: I use the same item for a written work and for the sets of his, if I understand well, the sets of its printed embodiment. We can do that if we can always tell if we are talking of the item as an artwork or if we are talking of the item as a set of books. This seems cool to me as we do not have to introduce a new concet like embodiment If we are linking an instance of a book object to an artwork item, we can know for sure it is as a class opf books. Now I don't know how that relates to money :) Maybe a property who links abstract things to cncrete one is cool. Can we say a material measure (Q1329540)  View with Reasonator View with SQID is an embodiment of its measure unit ? this would be cool but can't cover the new abstract definitions of the International system for example. TomT0m (talk) 10:33, 22 August 2014 (UTC)[reply]
    @Micru: A set can be an element of another set... a type can be a term in a universe type, which in turn can be a term in another universe type... a category (Q719395) can be an object of another category... This is not controversial, compared to "real objects" and "manifestations". Danneks (talk) 11:57, 22 August 2014 (UTC)[reply]
    @TomT0m: Neither I like controversies nor I dislike them, they just happen :) The example of "person" is complex, because there are different views [6], but if you say that <human> instance of <clade> to mean "member of the set of", that is fine for me.
    A <standard (class)> is definitely a manifestation of <measure unit (class)>. If you think the words "manifestation", or "embodiment" are too abstract, "execution", "rendering", or "interpretation" are also similar. OTOH, if we can use "instance of" for that purpose, I am also fine with it.--Micru (talk) 12:23, 22 August 2014 (UTC)[reply]
    Maybe "used to define" would be more accurate? There is the definition of metre (Q11573), and this definition mentiones light — quite a definite relationship. Danneks (talk) 14:44, 22 August 2014 (UTC)[reply]
    @Danneks: <definition of metre (1983)> manifestation of <metre>. A metre is just a concept, and it has been materialized in different ways through the years.--Micru (talk) 15:44, 22 August 2014 (UTC)[reply]
    @Micru: And here I'd have used instance of. "metre (1983)" is also a concept: I suspect that only humans know about metres, it requires a lot of abstraction. Danneks (talk) 15:58, 22 August 2014 (UTC)[reply]
    @Danneks: Exactly, both of them are mind-made, but as you were saying "metre (1983)" mentions light, which puts it closer to physical world than the other one. I'm also fine using <instance of>, but one has to bear in mind that "metre (1983)" is a class.--Micru (talk) 16:03, 22 August 2014 (UTC)[reply]
    Don't follow you guys. This discussion seems to go way out of control. If meter is a class, then which definition would you give for its instances ? the metre class is the set of all objects of precisely one meter ? Do not seem very useful. TomT0m (talk) 16:15, 22 August 2014 (UTC)[reply]
    Well, we can say: "this distance is a meter". Danneks (talk) 16:34, 22 August 2014 (UTC) I mean, the distance between given two points. Of course, we can say this about many pairs of points... whether it is useful or not. Danneks (talk) 16:51, 22 August 2014 (UTC)[reply]
    I see it as this:
    • <metre> concept gestated from 1668 to 1771 and that is still alive
    • <metre (1983)> one of many implementations of the concept
    • <metre (physical measure)> element built or measured following the 1983 guidelines
    Depending on how many of those abstraction levels you represent in Wikidata, some statements make more sense than others. For instance if we represent all 3 of them, to say "<metre (physical object)> instance of <metre (concept)>" conveys less meaning than to say "<metre (physical object)> instance of <metre (1983)>". It is important to realize that there are no "fix" classes. Like the lense of a microscope we can adjust the abstraction level of any class to at least those three levels. Perhaps it would be more clear if we use a distinct property.--Micru (talk) 17:03, 22 August 2014 (UTC)[reply]
    Fortunately, we would never have an item <metre (physical measure)>, because now it is impossible to build that physical element. And without its building, it will remain an abstract concept, the same as <metre (1983)>. Danneks (talk) 17:20, 22 August 2014 (UTC)[reply]
    Technically speaking, every time that you measure something you are instantiating a class based on <metre (1983)>. And with Commons joining in, maybe someone wants to create an item to represent the object that this image depicts. So never say never :) --Micru (talk) 17:31, 22 August 2014 (UTC)[reply]
    This image depicts a metre bar, which is not an instance of a metre, but of a metallic bar. All we can say is that a certain definition of metre uses that bar... maybe sometime we will be able to say in which way the definition uses it, and it would not involve any "manifestations", only the procedures which were used to compare measuring instruments with that bar. Danneks (talk) 18:04, 22 August 2014 (UTC)[reply]
    This image depicts a metre bar according to your mind, and which is not an instance of a metre according to your mind :) Without the mind there cannot be known if there is or there is not, because the mind precedes all that can be known. Without it you wouldn't be able to codify things and much less think about if there such a thing or not. It is impossible to get rid of "manifestations" (transformation of concepts into physical matter/processes) because it is impossible to get rid of the mind without getting rid of the idea that there are things. You can hide that as much as you wish, or pretend that it is not like that, but as soon you scratch a little bit this fact reappears, which in fact it is just a co-lateral effect of having a smaller system codifying a larger container system and building a model out of it.--Micru (talk) 07:25, 23 August 2014 (UTC)[reply]
    Yes, and Wikidata is a database of human knowledge. Danneks (talk) 07:58, 23 August 2014 (UTC)[reply]
    BTW, I don't think that to get rid of the mind is a healthy desire :) It's like "no man, no problem". Danneks (talk) 08:13, 23 August 2014 (UTC)[reply]
    So that's why there will always be concepts that can happen only in the mind (like mathematics), concepts that happen in the mind inferred from contact with reality, and things that happen independently from the mind but that we wouldn't know if we didn't have a mind to interpret them. For me all this conversation was about defining the relationship between the 1st and the 2nd group, if it is not clear I think it is better to let this discussion aside for a while. Oh, yes I agree about your last point... better than the desire for no-mind is the wish for understanding and control over it. Perhaps a good motto for the site would be "Wikidata - the mind you can edit" :) --Micru (talk) 09:35, 23 August 2014 (UTC)[reply]
    Of course I understand that you want to distinguish the concepts of the first and the second kind, the only problem is that they could not be clearly distinguished. Yesterday aether was "real", today it is "abstract"... Wikidata is needed as a reliable source of knowledge, not as a source of philosophical guesses of its editors. Danneks (talk) 10:14, 23 August 2014 (UTC)[reply]
    "Wikidata is needed as a reliable source of knowledge guesses", FTFY :) As expressed in this essay, abstraction is not a boolean parameter, but a continuum. If we wanted we could put thousands of shades and they will never end, but there are some basic categories that are useful for organizing knowledge and for making sense of it.--Micru (talk) 11:14, 23 August 2014 (UTC)[reply]
    OK, if you are sure that this way of making sense does not generate controversy in a multi-cultural setting and we would not get statements like <Muhammad> <manifestation of> <Allāh> all over the place, then I don't object. Danneks (talk) 12:01, 23 August 2014 (UTC)[reply]
    Wikidata has sources to deal with controversial statemants, that actually reads [claim] according to [the source]. We also have qualifiers like disputed by. But I really would recommand not to use this property out of libraries scope. TomT0m (talk) 12:11, 23 August 2014 (UTC)[reply]

    @Micru: FBBR seems to be an accepted model, can you provide an actual usecase of it ? And out of curiosity, why did you extract only one relationship of it when they forms a whole (meter seems clearly out of scope anyway). TomT0m (talk) 10:28, 23 August 2014 (UTC)[reply]

    @TomT0m:My main interest is to keep compatibility with the Library of Congress, The British Library, and other national libraries like DNB that are implementing RDA (the manifestation of FRBR). What others confuse with "philosophical guesses", is in fact a well-established conceptual framework with wide support. By making sense of that framework in our structure and having the necessary properties to model it, then it will be possible to keep interoperability with huge bodies of data. The use case is to automate metadata import and maintenance as much as possible with as many standards and sources as possible. I did extract only one relationship because it is the most significant one that we are missing (see here). Bear in mind that we are taking some liberties in its application, in order to keep it simple at the begining of item creation, but allowing an expansion of the structure whenever needed.--Micru (talk) 11:14, 23 August 2014 (UTC)[reply]
    It seems contradictory to both take liberties to maintain compatibility. When it is up to mass import, what about database laws issues ? It's probably OK for the US database imports, but … last to maintain compatibility with a well defined goal, how about some documentation about the mapping ? All 1/1 properties, with the same intended meaning ?
    Comment on your spreadsheet : I don't get the 3rd and fourth column. Isn't item/manifestation relation a 1/1 mapping ? TomT0m (talk) 11:38, 23 August 2014 (UTC)[reply]
    Item compatibility can be achieved by keeping a clean concept tree and by knowing the abstraction level that an item represents (think of abstraction level as a zoom). External properties can be mapped directly to q-items and inferred which is the closest property we have. And more documentation would be nice, but it requires a lot of time that I don't have. Item and manifestation are different, I gave an example high above this thread, but abstraction levels are approximations, and as such you can collapse them or expand them at will.--Micru (talk) 12:15, 23 August 2014 (UTC)[reply]
    Sorry, if you can't be clearer than that I won't follow this. TomT0m (talk) 12:35, 23 August 2014 (UTC)[reply]
    @TomT0m: If an item is an instance of any of these items, the instantiated item represents a work. If the item has "edition of" we know that it represents a expression of that work, and if another one is an instance of any of these then it represents a manifestation. FRBR-items are not relevant because they will be stored in Commons and linked with an specific property. In the excel file you have the properties required to connect the different abstraction levels (if they exist). Is that clear enough?--Micru (talk) 13:13, 23 August 2014 (UTC)[reply]
    @Micru: seems OK to me at first sight. If I understand well, a lot of their conceps implicitely maps with our concept. And if we choose to use instance of to link for example to link a work to a radiophonic manifestaion of this work or edition, thought as classes of manifestations, we can also map this to their manifestation of concept. In this view an edition can even be thought as a subclass of a work I guess.
    OK, just to recap :
    An artist starts to worw on something. He gets a sequence of intermediary results. This is what we could «work». finally he gets something he wants show other people, this is what we could call « an expression ». An expression is a text, or a sound, or a video. If we read and record the reading of a text, we get another expression of a work. An expression could be seen as a step in the working sequence of the author(s). For that matter, he uses a communication medium like a book or a radio station. He or they can produce any number of those manifestations. This mean that a work by itself in this model is neither textual or vocal by essence. A manifestation is the set of concrete objects storing the same expression. Different supports can store the same expression (CD - casettes), then they refers to the same expression, but belongs to manifestations. Expressions are sets of identical physical objects. A manifestation as a class can be seen as the set of the union all the physical objects of manifestations of the same expression.
    I think there could be a mismatch here. I woud identify an edition in the Wikidata sence with a manifestation, not to an expression. A reprinting of a book with the same text but with different colors is an edition. TomT0m (talk) 15:30, 23 August 2014 (UTC)[reply]
    @TomT0m: On a more practical level it is not as clear-cut as that, but rather:
    I want to model a novel, and I create an item (Qx) to enter some data about it. This item might represent any abstraction level or all at once (down to physical item), depending on how much data I entered. Then someone else sees my item and creates a new item (Qy) for a translation in his/her language and moves some of the data from Qx to Qz which is a new item more specific for part of the data I originally entered. Qx, my original item, now has a more limited scope, it has stopped representing all levels, and now it only represents work. Qy and Qz are now representing the remaining levels. Then a third person comes an adds some sub-items to Qz, which stops representing expression-manifestation-item, and now represents only expression...
    As you can see the items don't have a fix "area of influence", it keeps evolving depending on how much data is fed to them. The apparent mismatch is not only intentionate, but inevitable if we want to keep some flexibility. In practice it is much easier because people in general understand that if there is this much data, the item refers to this, or that if this is an instance of something material, then the item must be material. The only missing link was when we have to connect a class of non-material objects with a class of material objects. If you say that using "instance of" is enough, then it's fine.--Micru (talk) 17:32, 23 August 2014 (UTC)[reply]


    I think the best use case of this is for instances, not classes. The problem I find with classes is the many-to-many problem, so to me it's not clear if we were using it for classes what direction to go in, i.e.

    Reversed property, manifested as

    • Encylopedia (work) manifested as collection of books (manifestation)
    • Encylopedia (work) manifested as website (manifestation)
    • Encyclopedia volume (work) manifested as book (manifestation)
    • Encyclopedia article (work) manifested as book section (manifestation)
    • Encyclopedia article (work) manifested as webpage (manifestation)

    Original property, manifestation of:

    • Book (manifestation) manifestation of encyclopedia volume (work)
    • Website (manifestation) manifestation of encyclopedia volume (work)
    • Book (manifestation)manifestation of proceedings (work)
    • Book (manifestation)manifestation of dictionary (work)
    • Book (manifestation)manifestation of novel (work)

    The far more useful case, IMO, is using it for instances to denote relationships between instances of manifestations and instances of work. So if we wanted to find the work that a manifestation pertains to, or several manifestations of the work (using the original property only):

    • Insect ( Encyclopedia Britannica work)
      • instance of encyclopedia article (work)
      • author Sir Vincent Brian Wigglesworth
    • Insect ( Encyclopedia Britannica webpage)
    • Insect ( Encyclopedia Britannica article)
      • instance book section (manifestation)
      • manifestation of Insect ( Encyclopedia Britannica work)
      • volume II
      • pages 200-202

    In practical terms many of the work and manifestation properties are in the same wikidata item, in which manifestation of isn't useful but manifested as is slightly useful to the properties herein. As both manifesting a webpage and a book section, it will have a url as well as a volume:

    With neither property, using instance of for everything:

    If there is only one work and the manifestations seem to mix properties okay, this is less of an issue. But consider a song: as a work, the composer is generally considered the author, and this would go in the author property. But if the song is published in a book, the "author" is actually usually published as the author of the book. And in a recording of that song? Well the performer is frequently in the author field. If we have a single wikidata item for a song, in the "authors" field do we put the composer, and the author of every book it was in, and every performer that has ever sung it and the information for all those recordings that have been made of it? This is actually what happens to a popular song like Silent Night. It has a semi-large list of performers already. A Nightmare! (Pun intended).

    tl;dr Support, and I'll probably mostly use it for instances to link manifestations of a work to the work. Mvolz (talk) 20:47, 25 August 2014 (UTC)  Support to match FRBR which is the standard for book metadata. Filceolaire (talk) 21:52, 15 September 2014 (UTC)[reply]

     Support, it may be useful in some cases to distinguish between a text and a text with a medium. Danneks (talk) 15:59, 29 September 2014 (UTC)[reply]