Wikidata:Property proposal/Archive/38

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

Market capitalization

Done: market capitalization (P2226) (Talk and documentation)

Description	Market capitalization
Data type	Number (not available yet)
Template parameter	"The total value of issued shares by a publicly traded company"
Domain	Company items
Source	http://finance.yahoo.com/
Robot and gadget jobs	Bots should be doing this task
Proposed by	Mcnabber091 (talk) 19:22, 24 August 2013 (UTC)

Discussion

Support --Filceolaire (talk) 20:51, 24 August 2013 (UTC)
SupportMcnabber091 (talk) 00:17, 25 August 2013 (UTC)
Support --Danrok (talk) 14:32, 10 September 2013 (UTC)
Support Emw (talk) 03:02, 15 October 2013 (UTC)

Metallicity (en)

Done: metallicity (P2227) (Talk and documentation)

Description	en:Metallicity
Data type	Number (not available yet)
Template parameter	en:Template:Starbox detail metallicity
Domain	term
Example	Sirius (Q3409) => 0.50 ([Fe/H])
Proposed by	GZWDer (talk) 11:04, 8 July 2013 (UTC)

Discussion

Support - This property needs a qualifier to specify which type of metallicity. --Paperoastro (talk) 20:27, 20 July 2013 (UTC) P.S.: I changed the example.
Support Filceolaire (talk) 00:44, 20 August 2013 (UTC)
Support it is necessary for infobox. Sunpriat (talk) 19:22, 6 October 2014 (UTC)

FSB Speed

Done: FSB speed (P2150) (Talk and documentation)

Description	CPU Front Side Bus speed in Hz
Data type	Number (not available yet)
Domain	CPU
Example	Pentium D 945 (Q15217736): 800MHz
Source	Intel website, for Intel processors
Robot and gadget jobs	Robots can gather info on the Intel website and fill the property with it.
Proposed by	MisterSanderson (talk)

Discussion

I want to add information to the CPU items, but there aren't enough properties to that. MisterSanderson (talk) 19:47, 20 November 2013 (UTC)

Support --Tobias1984 (talk) 10:49, 21 November 2013 (UTC)

Comment Again this is not so simplistic. FSB speed is variable and has a relationship with the CPU speed. These things are often configurable settings specific to an individual computer. Also, FSB has been superseded by HyperTransport or Intel QuickPath Interconnect, which further complicates matters. It begs the question do we really need this data, what will it be used for? Danrok (talk) 01:43, 27 January 2014 (UTC)

But there is a FSB speed specified by the manufacturer for the CPU. It's the maximum FSB speed without overclock.--189.83.216.41 13:49, 27 January 2014 (UTC)

On hold To be re-opened when the right datatype is available.--Micru (talk) 13:57, 23 May 2014 (UTC)

Clock speed

Done: clock frequency (P2149) (Talk and documentation)

Description	CPU clock frequency in Hz
Data type	Number (not available yet)
Domain	CPU
Example	Celeron 430 (Q15219210): 1800MHZ
Source	Intel website, for Intel processors
Robot and gadget jobs	Robots can gather info on the Intel website and fill the property with it.
Proposed by	MisterSanderson (talk)

Discussion

I want to add information to the CPU items, but there aren't enough properties to that. MisterSanderson (talk) 00:46, 21 October 2013 (UTC)

Should be number datatype.--GZWDer (talk) 14:27, 26 October 2013 (UTC)
Support Changed to number.--Micru (talk) 16:54, 1 November 2013 (UTC)
Comment - This property should use the unit "Hz". Mega and Giga are just prefixes to the cycle rate. --Tobias1984 (talk) 13:12, 19 November 2013 (UTC)

Tobias1984, using Hz is a bad thing. The clock frequency of actual processors are measured in a small number of GHz, or a big number in MHz. The numbers in Hz will be too big... 1.8GHz = 1,800MHz = 1,800,000KHz = 1,800,000,000Hz. This is very unpractical.--MisterSanderson (talk) 12:42, 20 November 2013 (UTC)

What I'm trying to say is that it shouldn't be a unit-less datatype, but have Hz as a unit. The prefixes and the preferred displaying of MHz and GHz will be handled by the user interface. If it is unit-less it will output 1.8 in a query and it will take additional work to put in the GHz. --Tobias1984 (talk) 12:49, 20 November 2013 (UTC)

The Wikidata user interface can show Hz as GHz automatically?--MisterSanderson (talk) 17:20, 20 November 2013 (UTC)

We are just discussing the specifications of the datatype here: Wikidata:Requests for comment/Dimensions and units for the quantity datatype. Everybody is welcome to comment ;) --Tobias1984 (talk) 18:09, 20 November 2013 (UTC)

Comment CPU clock rates are variable on the vast majority of CPUs. Also, there is no need for clock rate to be specific to CPUs, it applies to all types of chip, e.g. GPUs, RAM, etc. Danrok (talk) 01:37, 27 January 2014 (UTC)

But there is a clock rate specified by the manufacturer for the CPU. It's the maximum clock rate without overclock.--189.83.216.41 13:49, 27 January 2014 (UTC)

Support if it is number with units but

Oppose if it is a dimensionless number. Remember there was a time when CPU clock rate was measured in kilohertz and we want to describe them too without having silly long numbers for next years GHz chips. Filceolaire (talk) 05:44, 25 February 2014 (UTC)

On hold To be done when the right datatype is available.--Micru (talk) 22:09, 4 March 2014 (UTC)

Max TDP

Done: thermal design power (P2229) (Talk and documentation)

Description	potency in Watts
Data type	Number (not available yet)
Domain	CPU
Example	Itanium 9015 (Q15226510) : 104 W
Source	Intel website, for Intel processors
Robot and gadget jobs	Robots can gather info on the Intel website and fill the property with it.
Proposed by	MisterSanderson (talk)

Discussion

I want to add information to the CPU items, but there aren't enough properties to that. MisterSanderson (talk) 00:48, 25 November 2013 (UTC)

On hold To be re-opened when the right datatype is available.--Micru (talk) 13:57, 23 May 2014 (UTC)

explosive energy equivalent

Done: explosive energy equivalent (P2145) (Talk and documentation)

Description	amount of energy discharged during an explosion
Data type	number (with dimension)-invalid datatype (not in Module:i18n/datatype)
Domain	Bombs, especially nuclear bombs
Example	Tsar Bomba (Q188596) --> 50 [unit:megatons] Father of All Bombs (Q1049487) --> 44 [unit:tons]
Proposed by	--Jakob (talk)

Discussion

@Tobias1984, Emw, Snipre: Anyone have an opinion? --Jakob (talk) 00:23, 1 March 2014 (UTC)

Support I am just wondering if we should use a more generic name. The page about the Tsar Bomba uses "Blast yield". In German one uses simply "Sprengkraft" which is "blast force". Maybe "energy released" would emphasize the physical unit. I would also prefer Joules (+SI-prefixes) and convert to kilotons for nuclear weapons. The advantage would be that we could also use it for ammuntion, earthquakes, and volcanic eruptions. --Tobias1984 (talk) 09:50, 1 March 2014 (UTC)

This seems like a good idea. --Jakob (talk) 12:51, 1 March 2014 (UTC)

I would prefer something like energy released or energy content. Snipre (talk) 15:24, 6 March 2014 (UTC)

How about enthalpy change (delta-H)? That would certainly be the most generic naming and could take positive and negative numbers. Tobias1984 (talk) 10:56, 21 May 2014 (UTC)

@Tobias1984, Snipre: Sure, why not? --Jakob (talk) 12:12, 25 May 2014 (UTC)

For explosion it is difficult to express the enthalpy change: is it only the thermal change ? Including the volume expansion work ? Snipre (talk) 13:42, 13 June 2014 (UTC)

On hold TBD when the datatype is available. Label changed to match nuclear weapon yield (Q3062538). Notification: @Jakec, Tobias1984, Snipre:--Micru (talk) 08:52, 7 July 2014 (UTC)

@Micru: Put a name without nuclear in order to use that property for other explosives than nuclear bombs. Something like "explosive energy equivalent". Snipre (talk) 15:10, 18 July 2014 (UTC)

@Snipre: Good idea, changed. As per w:TNT equivalent a "megaton" is equal to 4.184 petajoules, so probably it should be available as unit.--Micru (talk) 15:18, 18 July 2014 (UTC)

There must be "TNT" in the title, because without specifying the type of explosive it has wrong units. -- MichaelSchoenitzer (talk) 21:55, 27 August 2014 (UTC)

torque

Done: torque (P2230) (Talk and documentation)

Description	The torque of a machine, a motor/engine/powerplant, or a vehicle
Represents	torque (Q48103)
Data type	Number (not available yet)
Domain	BMW F800S (Q4836040) BMW F 800 (Q796496)
Allowed values	Zahl
Example	Drehmoment => 80 (80 Nm)
Proposed by	194.121.90.163

Discussion: Support Probably needs qualifiers for things that have gears. The question is also if this can be expressed in a single number or if we need another datatype that can handle functions. -Tobias1984 (talk) 11:11, 29 October 2014 (UTC); Support – The preceding unsigned comment was added by F6F (talk • contribs).; Wait till we have a 'number with dimension' datatype. Filceolaire (talk) 15:21, 2 November 2014 (UTC)

On hold Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:46, 16 March 2015 (UTC)

frequency

Done: frequency (P2144) (Talk and documentation)

Description	Eine Frequenz auf der z.B. ein Radiosender empfangen werden kann in Hz
Data type	Number (not available yet)
Domain	Flux FM (Q1267187) radio station (Q14350)
Allowed values	Zahl
Example	Frequenz => 87500000 (87,5 MHz)
Proposed by	WhiteHole (talk)

Discussion

WhiteHole (talk) 21:28, 3 June 2014 (UTC)

Support - "Carrier frequency" would be clearer. Wait for "Hz" datatype. -Tobias1984 (talk) 10:56, 22 June 2014 (UTC)

Support @Tobias1984:- radio station was just an example, I think carrier frequency is needlessly specific to telecom. That said the other use cases I can think of involve a range of frequencies (i.e. violet (Q428124)). Is there a good way to represent ranges in WD yet? Also I was actually thinking of proposing the property "frequency" for events (i.e. annual). Now I guess I will have to call it "Event frequency" although I suppose it is possible to represent "annually" in Hz... :) Mvolz (talk) 22:24, 31 October 2014 (UTC)

Remember that in radio you have a carrier frequency and a signal bandwidth, both in Hz so you probably do need specific properties to distinguish these. Suggest these are put on hold until we have a 'number with dimension' datatype. Filceolaire (talk) 15:27, 2 November 2014 (UTC)

On hold Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:45, 16 March 2015 (UTC)

Genome size

Done: genome size (P2143) (Talk and documentation)

Description	size of the genome in base pairs
Data type	Number (not available yet)
Template parameter	"size" in en:template:infobox genome
Domain	genome (Q7020) (genome)
Example	human genome (Q720988) => 3234830000
Source	infobox (genome)
Proposed by	Dan Bolser (talk)

Discussion

Motivation is that we want to collect this information for all the genomes in wikipedia, or annotate lists of genomes for automatic querying of genomic information and for cross language updates. Dan Bolser (talk) 11:58, 15 April 2014 (UTC)

Support Such basic metrics should be in Wikidata proper, not just in some external resource. --Magnus Manske (talk) 12:45, 15 April 2014 (UTC)

Wait The information is usually measured in "SI-Prefix"+"Basepairs" which is why we should maybe wait for the datatype "number + unit" to become available. For items about Genomes (e.g. human genome (Q720988)) a more generic "information stored"-property would allow us to use the same property for a wide array of items (e.g. also hard drives etc...). Tobias1984 (talk) 15:17, 15 April 2014 (UTC)

Tobias, the value for an "information stored" property in a genome is interesting to consider, but I don't think it's what we'd be looking for in this property. Genomes are conventionally measured in base pairs (or nucleotides), not bits. (Genomes often contain huge tracts of repetitive sequence whose information content can be highly compressed depending on the algorithm, etc.) Emw (talk) 03:25, 16 April 2014 (UTC)

Comment: I agree with Emw on this point. --2001:630:206:4001:F1A1:AEAA:B627:DDA0 11:26, 6 May 2014 (UTC)

Comment: Dan, Magnus, since genomes are commonly treated as sequences, I think it would make sense to at least change "size" to "length" in the proposal. See for example http://www.ncbi.nlm.nih.gov/assembly/GCF_000001405.26/, http://ensembl.org/Homo_sapiens/Info/Annotation#assembly.

Following up on "genome size" vs. "genome length": the English Wikipedia article Genome size is illuminating. It says genome size "is typically measured in terms of mass in picograms (trillionths (10−12) of a gram, abbreviated pg) or less frequently in Daltons or as the total number of nucleotide base pairs". In other words, "genome size" can have units of mass or length, depending on the audience. "Genome length" cuts down the ambiguity significantly. And although "genome size" has more hits than "genome length" in scholarly publications (compare here, here), Ensembl and NCBI databases both seem to prefer "length" when talking about this kind of thing. Emw (talk) 04:01, 16 April 2014 (UTC)

Comment: Template:infobox genome already notes units of "Mb", but I'll note here that "b" (bases) and not "bp" (base pairs) seems like the best unit for genome length. This accounts for organisms with single-stranded DNA genomes like phi X 174 (the first sequenced genome). Emw (talk) 03:25, 16 April 2014 (UTC)

Comment: From the evidence given, I agree that genome length (in bases) is a better property. --2001:630:206:4001:F1A1:AEAA:B627:DDA0 11:26, 6 May 2014 (UTC)

It's also worth considering that there are several types of biological sequence data we'll want to describe the length of on Wikidata: genes (b), proteins (aa) and chromosomes (b). Do we want one length property for each of these, or one property for all of them? As I note in a related discussion, I'm inclined to use one length property for all sequences. Emw (talk) 02:58, 16 April 2014 (UTC)

WikiProject Molecular biology has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. Tobias1984 (talk) 11:57, 16 April 2014 (UTC)

Comment: We couldn't come to any agreement in our local group... are you suggesting something like Length of nucleotide sequence, which would be a property of items of type gene, genome, etc.?

Length of nucleotide sequence sounds fine to me, and better than just genome size or length. --Daniel Mietchen (talk) 12:14, 9 May 2014 (UTC)

'nucleotide sequence' is ambiguous and typically refers to a gene sequence, not an entire genome. Kaldari (talk) 21:59, 30 December 2014 (UTC)

Wait Let's wait until we have units. Kaldari (talk) 21:59, 30 December 2014 (UTC)

On hold pending unit property. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:44, 9 March 2015 (UTC)

Explosive velocity

Done: explosive velocity (P2231) (Talk and documentation)

Description	see explosive velocity (Q908624) and Table of explosive detonation velocities (Q3062430)
Data type	Number (not available yet)
Domain	chemicals
Example	octanitrocubane (Q413940)=>10100m/s
Proposed by	GZWDer (talk)

Discussion

Motivation. GZWDer (talk) 13:08, 10 February 2014 (UTC)

@GZWDer: I'missing a motivation. :) --Succu (talk) 22:21, 23 April 2014 (UTC)

~~Oppose~~

Neutral I know that Explosive velocity is one of the explosive characteristics but for me this is limited to 30-50 components perhaps 100. This is not sufficient for a property at that level of the WD development. Snipre (talk) 11:16, 28 April 2014 (UTC)

Support There's no harm in having a property that doesn't get used much. atomic number (P1086) will also never see much use, but it's a useful property. --Jakob (talk) 11:30, 28 April 2014 (UTC)

Support --LydiaPintscher (talk) 21:25, 10 May 2014 (UTC)

Support Per Lydia. --Eurodyne (talk) 00:47, 10 October 2014 (UTC)

Wait till we have 'number with dimension' datatype then resubmit. Filceolaire (talk) 22:48, 2 November 2014 (UTC)

On hold Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:37, 16 March 2015 (UTC)

reason for deprecation

Done: reason for deprecated rank (P2241) (Talk and documentation)

Description	Qualifier to allow the reason to be indicated why a particular statement should be considered deprecated.
Data type	Item
Domain	any statement that is not of top rank
Allowed values	item from a controlled vocabulary -- items would need to be P31/P279* a class item made for the purpose
Example	lion (Q140):taxon name (P225) = "Felis Leo" -> this qualifier = item for "superseded taxon"
Source	reference, Wikipedia article, etc
Robot and gadget jobs	Certain types of statement -- eg statements of spouses who are no longer current spouses -- could perhaps be picked up automatically. A report could be created for how many not top-ranked items lack the explanatory qualifier. Bots could also check that statements with this qualifier did not have top rank.

Motivation

It seems to me that it would be generally useful to record not just that a statement is deprecated, but why it is deprecated (or not of top rank, if another statement is preferred).

For example, it may be very valuable to distinguish statements that were once considered correct but are no longer accepted, from statements that were never correct (but which may have been widely reported).

Also, it is useful to have a qualifier for this, that then a reference can be specifically attached to, if the reference details the reason for deprecation.

Also, because it could be valuable for queries eg to be able to report all species for which a specifically obsolete classification is noted, as opposed to a more generally incorrect one.

Also, because a qualifier like this would make it more obvious that a statement was deprecated or not top rank, eg when read on the wikidata page or on Reasonator -- something significant because editors sometimes complain that they are wary of adding wrong statements to Wikidata, even with the "deprecated" rank, because they are concerned that readers (or bots) will not notice the little 'deprecated' checkmark. (Add your motivation for this property here.) Jheald (talk) 08:44, 4 October 2015 (UTC)

Discussion

Support --- Jura 08:56, 4 October 2015 (UTC)
Support Thryduulf (talk: local | en.wp | en.wikt) 16:11, 4 October 2015 (UTC)
Comment "statements of spouses who are no longer current spouses" should not be deprecated in my opinion. They should rather add an enddate to them. deprecated is more for false statements that we have a source for. If newspapers tells Pope Francis (Q450675) is married to Madonna (Q1744), but that information has been proven wrong, then it can be turned to deprecated. But if there is proof that they divorced, that statement should have an enddate. But I Support the idea of this property. A question is what kind of items should be used for my example? -- Innocent bystander (talk) 19:47, 4 October 2015 (UTC)
Support The current example is fine. A value that is widely known but is wrong should be deprecated. where there are multiple correct values they should be standard rank except that the current value (husband, population, etc.) should be preferred.A controlled vocabulary for describing deprecated items is an excellent idea. Joe Filceolaire (talk) 19:07, 12 October 2015 (UTC)

Box office

Done: box office (P2142) (Talk and documentation)

Description	box office accumulated by a film
Data type	Number (not available yet)
Template parameter	w:Template:Infobox film box office, also used on w:Template:Infobox concert tour as gross
Domain	films, concert tours, other live events
Allowed values	monetary values
Example	Skyfall => $1,108,561,013
Source	Information is usually taken
Robot and gadget jobs	Yeah, can be done with bots
Proposed by	— ΛΧΣ²¹

Discussion

I think this is needed, and such information is always worthy for a database. — ΛΧΣ²¹ 02:02, 7 May 2013 (UTC)

It's a good idea but currency (Q4917) and country/worldwide should be qualifiers and the unformatted number the value of the property (1108561013) as different languages may use different formatting and customs for giving the currency unit. - Matthew Beta (talk) 18:35, 7 May 2013 (UTC)

I changed the datatype, it is clearly a number with unit. --Zolo (talk) 12:47, 31 May 2013 (UTC)

Support --Zolo (talk) 12:47, 31 May 2013 (UTC)

Cash

Done: cash (P2232) (Talk and documentation)

Description	The value of cash and securities held by a state government
Data type	Number (not available yet)
Template parameter	"State cash and security holdings"
Domain	Country, region and city articles for the Global Economic Map
Source	http://www.census.gov/govs/local/
Robot and gadget jobs	Bots should be doing this task
Proposed by	Mcnabber091 (talk) 19:05, 24 August 2013 (UTC)

Discussion

Support--Filceolaire (talk) 20:51, 24 August 2013 (UTC)
SupportMcnabber091 (talk) 00:17, 25 August 2013 (UTC)
Support Emw (talk) 02:53, 15 October 2013 (UTC)

argument of perihelion

Done: argument of periapsis (P2248) (Talk and documentation)

Data type	Number (not available yet)
Example	Earth (Q2) → 102.94719 deg

Yair rand (talk) 11:21, 14 October 2015 (UTC)

Discussion: Support - Mbch331 (talk) 11:27, 14 October 2015 (UTC); Support - But probably need a change of name to argument of periapsis. -- Innocent bystander (talk) 20:05, 18 October 2015 (UTC)

Refseq Genome ID

Done: RefSeq genome ID (P2249) (Talk and documentation)

Description	Specifies the RefSeq Genome ID.
Data type	String
Domain	genome (Q7020), strain (Q855769)
Allowed values	String
Example	Chlamydia trachomatis D/UW-3/CX (Q20800373) = NC_000117.1 (RefSeq genome accession)
Format and edit filter validation	Characters + underscore + Digits
Source	http://mygene.info
Robot and gadget jobs	ProteinBoxBot will use this for mapping gene to genome assembly

Motivation

While Wikidata includes RefSeq RNA ID (P639) and RefSeq protein ID (P637), RefSeq Genome ID is still omitted. This is a key identifier for genome assemblies, and becomes increasingly important as new species, and strains of microbial species, are added to Wikidata. The RefSeq Genome ID allows mapping of gene to genome. Entrez Gene ID (P351) are gene identifiers specific to the genome of the specimen they were sequenced from. It will be important to include genome assembly identifiers to species and strain items that genes link to. Putmantime (talk) 23:22, 14 October 2015 (UTC)

WikiProject Molecular biology has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead. --Tobias1984 (talk) 11:13, 15 October 2015 (UTC)

Support Will be essential the more reference genomes of different species make it into Wikidata. Sebotic (talk) 21:09, 16 October 2015 (UTC) Support More and more content on different species make it to Wikidata. Andrawaag (talk) 21:21, 20 October 2015 (UTC)
Support Needed in order to keep track of everything that is coming in. Emitraka (talk) 21:32, 20 October 2015 (UTC)

Done --Tobias1984 (talk) 19:29, 21 October 2015 (UTC)

Life expectancy

Done: life expectancy (P2250) (Talk and documentation)

Description	life expectancy for this group or species
Data type	Number with dimension 'year'-invalid datatype (not in Module:i18n/datatype)
Template parameter	likely many
Domain	biology
Allowed values	0-9, time periods
Example	Pug (Q38698) => 12-15 years
Proposed by	Pikolas (talk)

Discussion

I don't know what datatype to add, though. Pikolas (talk) 00:38, 28 December 2013 (UTC)

Comment where would this data be sourced from? Danrok (talk) 20:25, 29 January 2014 (UTC)

Scientific papers for non-human animals; statistical studies published by research institutes such as Instituto Brasileiro de Geografia e Estatística (Q268072). Pikolas (talk) 23:44, 29 January 2014 (UTC)

Comment datatype should be years. Can be moved to pending properties. --Tobias1984 (talk) 18:47, 4 February 2014 (UTC)

Life expectancy for some organisms is less than one year; measured in days, weeks or months

Support datatype fixed. Filceolaire (talk) 05:54, 25 February 2014 (UTC)

Support --- Jura 18:11, 2 July 2014 (UTC)

Comment you can define an average life expectancy or you can give a range that includes a lower and a higher value. Humans can live between 60 and 110 years. The figures depend on the circumstances, nutrition, lifestyle and medical care. Males in many cases have a shorter life expectancy. --Giftzwerg 88 (talk) 14:27, 13 November 2014 (UTC)

Done --Almondega (talk) 22:47, 21 October 2015 (UTC)

Orbit apsis (two properties)

Not done

Description	the apsides of the elliptical orbits of astronomical bodies
Data type	Number (not available yet)
Template parameter	en:template:Infobox planet aphelion, perihelion, periastron, apoastron, periapsis, apoapsis
Domain	all astronomical bodies with an orbit
Allowed values	positive numbers
Example	<Earth> apsis <152,098,232 km (aphelion)>, <147,098,290 km (perihelion)>
Proposed by	Paperoastro (talk)

Motivation. One of the orbit parameters described by Wikipedia templates. Astronomers uses different names (see the list here), basing on the "major body" (e.g. aphelion and perihelion for bodies around Sun, perigee and apogee around the Earth, and so on...). Qualifiers can distinguish the different types. If the community prefers to have two property, one for "periapsis" and one for "apoapsis", for me is the same. Paperoastro (talk) 14:26, 18 April 2013 (UTC)

Support I think I would use separate properties (periapsis & apoapsis) so the number can of periapsis can have qualifiers for "mean value", "maximum", "minimum" which I recall does wobble a bit for bodies like the moon. --Tobias1984 (talk) 08:32, 19 April 2013 (UTC)

See Wikidata talk:Astronomy task force#Uncertainty. -- Lavallen (block) 14:43, 20 April 2013 (UTC) (And

Support of course)

To be more clear. "Qualifiers can distinguish the different types" as Paperoastro tells above. But the periapsis can also wobble, true, then we can use the "Uncertainty". -- Lavallen (block) 11:12, 21 April 2013 (UTC)

@Tobias1984: if I was too much "cryptic" above. Astronomers use different names for the apsides of orbits, depending on the major of the two bodies: see this table. My idea is use qualifiers to distinguish between aphelion and perihelion, or perigee and apogee, or periastron and apastron, and so on. Periapsis and apoapsis have, by definition, one value, and cannot have "minimum", "maximum" or "mean". --Paperoastro (talk) 16:34, 21 April 2013 (UTC)

Support it is necessary for infobox. Sunpriat (talk) 19:37, 6 October 2014 (UTC)

@Paperoastro, Tobias1984, Lavallen, Sunpriat: Not done Use apoapsis (P2243) and periapsis (P2244). Archived --Almondega (talk) 20:00, 22 October 2015 (UTC)

Electrical resistivity (en) / elektrischer widerstand (de)/ résistivité électrique (fr)/ la resistividad eléctrica (es)

Not done

Description	The electrical resistivity of a substance
Data type	Number (not available yet)
Template parameter	"electrical reistivity" in en:template:infobox element
Domain	Any substance
Example	lithium (Q568) --> 92.8 barium (Q1112) --> 332
Source	w:Electrical resistivities of the elements (data page)
Robot and gadget jobs	Possible, but not needed with ~80 to 90 items where this will go.
Proposed by	--Jakob (Scream about the things I've broken) 17:18, 27 August 2013 (UTC)

Discussion

Support Danrok (talk) 17:34, 27 August 2013 (UTC)
Support but it will go on a lot more than 90 items. chemical tables record that value for thousands of materials. --Tobias1984 (talk) 14:52, 28 August 2013 (UTC)
Oppose It is the inverse of electrical conductivity which has already be approved see Wikidata:Property_proposal/Pending#Electrical_conductivity_.28en.29. Snipre (talk) 17:05, 29 August 2013 (UTC)

But we still need it, because we have to fill in what the source says. --Tobias1984 (talk) 17:10, 29 August 2013 (UTC)

Ok, I agree with the idea of putting the value as defined in the sources but I really doubt that 2 properties for the same concept are necessary: do we have more information about conductivity than for resistivity or the inverse ? Are we really missing some opportunities by having only one property ? From my point of view it is better to keep the number of properties as low as possible but to put effort to find values for all chemical elements or chemicals. I am a little afraid that having different properties for the same concepts will dilute the data import and reduce the possibility to create complete lists or to perform powerful queries. Snipre (talk) 17:41, 7 November 2013 (UTC)

@King jakob c 2, Danrok, Tobias1984, Snipre: Not done Convert to electrical conductivity (P2055). Archived --Almondega (talk) 20:28, 22 October 2015 (UTC)

Particle Life expectancy

Not done

Description	Characteristic property of any elementary particle
Data type	Number (not available yet)
Template parameter	de:Vorlage:Infobox Teilchen
Domain	particles
Allowed values	Number or infinity (infinity for stable particles)
Example 1	MISSING
Example 2	MISSING
Example 3	MISSING
Source	scientific publications
Proposed by	Svebert (talk) 21:02, 15 March 2013 (UTC)

I don't think there is a possibility to set a number to infinity – we would probably need to a "special" (like 999999999999 Years) number, use this only for in stable particles or do a feature request. Other Question: Lifeexpectancy or half-time? They only differ by a constant factor, so it would be redundant to introduce both. -- MichaelSchoenitzer (talk) 01:22, 20 March 2013 (UTC)

Support - can we just use an item for infinite? --Tobias1984 (talk) 17:01, 6 May 2013 (UTC)

Or "No value" for infinite? - Soulkeeper (talk) 14:26, 10 May 2013 (UTC)

the devs need to allow for a value of positive infinity and negative infinity. Filceolaire (talk) 22:29, 16 May 2013 (UTC)

Comment The life expectancy of stable particles is always described as "stable", not "infinity". There can always be an external force of decay and within an infinite time span this will eventually occur. —★PοωερZ_{_talk} 22:39, 16 May 2013 (UTC)

Comment I've never heard the term "life expectancy" in particle physics. "lifetime" (average time before decay), "decay rate" or half-life are the generally accepted terms. ArthurPSmith (talk) 15:56, 26 August 2015 (UTC)

Note that half-life (P2114) has been created. Pamputt (talk) 04:35, 13 October 2015 (UTC)

@Svebert, MichaelSchoenitzer, Filceolaire, ArthurPSmith, Pamputt: Not done Convert to half-life (P2114) ( half-life = ln2 * mean lifetime ) Archived --Almondega (talk) 20:44, 22 October 2015 (UTC)

total debt

Not done

Description	The accumulated debt of a nation or company
Data type	Number (not available yet)
Template parameter	"Debt"
Domain	Country, region and city articles for the Global Economic Map
Allowed values	Number with 'currency' dimension
Example 1	MISSING
Example 2	MISSING
Example 3	MISSING
Source	http://www.usdebtclock.org/ http://www.census.gov/govs/local/ for USA
Robot and gadget jobs	Bots should be doing this task
Proposed by	Mcnabber091 (talk) 19:05, 24 August 2013 (UTC)

Discussion

Support--Filceolaire (talk) 20:51, 24 August 2013 (UTC)

SupportMcnabber091 (talk) 00:17, 25 August 2013 (UTC)

Support Suggest Total Debt as the label. Danrok (talk) 20:14, 26 August 2013 (UTC)
Support This is a fundamental value to understanding a country's economy. Blue Rasberry (talk) 11:27, 20 September 2013 (UTC)
Done Approved, waiting for number datatype. --Jakob (Scream about the things I've broken) 15:54, 10 October 2013 (UTC)

@Mcnabber091, Filceolaire, Danrok, King jakob c 2: Archived Not done Use total debt (P2133). --Almondega (talk) 21:53, 22 October 2015 (UTC)

number of academic staff

Not done

Description	Number of faculty / academic staff members.
Data type	Number (not available yet)
Template parameter	"academic_staff" in en:template:infobox university
Domain	organization (Q43229) (organization)
Allowed values	positive number
Example	Stanford University (Q41506) => 2043
Format and edit filter validation	none
Source	infobox
Robot and gadget jobs	bots can do
Proposed by	Danrok (talk) 00:05, 14 July 2014 (UTC)

Discussion

Wouldn't it be preferable just to have a property for number of staff in general? --Yair rand (talk) 23:46, 15 September 2014 (UTC)
+1. And a better definition of staff is needed: are the PhDs working as assistants counted in the staff of a university ? Snipre (talk) 16:21, 16 September 2014 (UTC)
Support there is already employees which captures the number of employees in an organization, but the count of academic staff is pretty important in the realm of science. Considerations regarding counting PhD as academic staff or not seems not to be a problem, since usually the organizations by themselves provide such statistics. Moreover this value is used in other Wikipedias than the English one (e.g. Polish) so these data doesn't seem to be controversial. Apohllo (talk) 23:18, 15 January 2015 (UTC)

Oppose use contributor to the creative work or subject (P767) with qualifier --Pasleim (talk) 21:39, 21 January 2015 (UTC)

What kind of qualifier? Apohllo (talk) 09:51, 22 January 2015 (UTC)

@Apohllo: I guess quantity (P1114)

qualifier, which could be used, like for example

⟨ university X ⟩ contributor to the creative work or subject (P767) ⟨ docent (Q462390)  

 ⟩
quantity (P1114) ⟨  80 ⟩

, but @Pasleim: this does not seem to fit for this property, as it's defined to be the different persons who worked on an artwork, not employees ... TomT0m ^(talk) 21:14, 14 June 2015 (UTC)

Sorry, contributor to the creative work or subject (P767) is completly wrong here. I meant employees (P1128) with qualifier. --Pasleim (talk) 21:37, 14 June 2015 (UTC)

Question - Is there a reason employees (P1128) with qualifier has use (P366) -> faculty member (Q5428874) would not work for this? Josh Baumgartner (talk) 18:31, 27 February 2015 (UTC)
Definitely, not. First, faculty member (Q5428874) is for academic staff in universities only and cannot be applied to research centers. And second, it is Europe/USA-centric as there is no such thing in other countries. Artem Korzhimanov (talk) 21:45, 12 June 2015 (UTC)

So you create a new item for "academic staff" or use researcher (Q1650915) for research centres. Research centres are just another big organisation.

<employees (P1128):#### (applies to part (P518):researcher (Q1650915)).

Universities are not just another big organisation so there is (in my opinion) a case for a dedicated property for the faculty member (Q5428874) at a university. Joe Filceolaire (talk) 14:20, 3 August 2015 (UTC)

Oppose. Use employees (P1128):#### (applies to part (P518):researcher (Q1650915) or other applicable faculty type). If we encode the type of faculty in the property definition, we'll be forever deciding the boundaries of who is and is not a faculty member. Instead, use the more general property and qualify it with entities specific to the country and/or educational system. For example, Harvard University (Q13371) has 6673 faculty and graduate assistants (per http://nces.ed.gov/collegenavigator/?q=harvard&s=all&id=166027#general), but that's broken down into 4175 full-time faculty, 454 part-time faculty, and 2044 graduate assistants, and then instructional versus research/public service within each one of those. And using https://nces.ed.gov/ipeds/datacenter/InstitutionProfile.aspx to look up Harvard, we can find an even more granular list including area of teaching and tenure status. Let's make entities for the types of academic staff we want to count. Then not only can we qualify employees (P1128) but we can also say, for instance, that a particular person is a full-time tenured research faculty member. Runner1928 (talk) 20:24, 23 September 2015 (UTC)
Not done due to no consensus. Archived Josh Baumgartner (talk) 22:19, 22 October 2015 (UTC)

allies

Not done

Description	the faction's allies in the war
Represents	political faction (Q1393724)
Data type	Item
Template parameter	"allies" in en:template:infobox war faction
Domain	organization (Q43229)
Allowed values	organization (Q43229)
Example	Boko Haram (Q212372) => Al-Qaeda (Q34490)
Format and edit filter validation	(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
Source	infobox
Robot and gadget jobs	bots can gather from infoboxes
Proposed by	Danrok (talk)

Discussion

Motivation: Create relationship links between organizations. Danrok (talk) 14:24, 10 January 2015 (UTC)

Comment Is is possible to represent that efficiently? n allied organizations require n*(n-1)/2 (or even n*(n-1) statements. Apohllo (talk) 00:43, 16 January 2015 (UTC)

Comment It would be nice to have this for people to. I have done some work on Italian Nobles and have thought it would be nice to have a way to indicate which people they fought along side and who they opposed, etc. Antrocent (talk) 08:15, 3 March 2015 (UTC)

Comment. Danrok The info boxes listed in Category:War and conflict navigational boxes (Q8170912) allow for two (or more) belligerents, each with a list of allied units/countries/participants. Many of these have the option of listing the number of combatants,, deaths, casualties etc for each unit. This would seem to required qualifiers on qualifiers, which wikidata cannot do. The way round this is to list the <belligerents>, then list all the <units> with each of these having qualifiers for <allied with>, <no. of combatants>, <Commander/leader> number of deaths (P1120), number of casualties (P1590), etc. Will this proposed property work as the <allied with> property to link units with the side they are on? Filceolaire (talk) 18:07, 13 March 2015 (UTC)

I'm not sure that "allied with" and belligerents are quite the same thing. It is possible for two groups to be fighting against the same opponent, but not as allies. Strictly speaking if you were making a list of belligerents you would only have one list of all participating groups. I'm not sure how to come up with a system that works in all situations, because not all wars are the same, and things can get a little complicated, e.g. Italy and Finland during WW2. Danrok (talk) 16:09, 14 March 2015 (UTC)

I meant "allied with" as a qualifier to "belligerent". As we cannot have qualifiers on qualifiers it seems to me we need to have one list of all participating groups so we need to use qualifiers to separate them into 2 (or three) sides/alliances. A system that works in all cases does seem ambitious but at the least we need a system that can capture the same info that can be included in the Category:War and conflict navigational boxes (Q8170912). Until we have a proposal for a set of properties that can do that we should probably hold off on creating new properties related to conflicts. Filceolaire (talk) 12:26, 28 April 2015 (UTC)

Weak oppose Notable alliances can have their own item, and each ally can use member of (P463) to indicate their membership (with qualifiers to cover dates of shifting loyalties). conflict (P607) can link the alliance with the war/conflict it is part of. Josh Baumgartner (talk) 18:01, 5 June 2015 (UTC)

Coding the info contained in the infobox for a battle is not easy.

We need a property to identify the various sides. We have participating team (P1923) for sport. What is the equivalent for a battle?

We need a property for each of the components of the sides in the battle - like participant (P710) for sport and candidate (P726) for elections. Can we use these for a battle or do we need a new property? "belligerent"?

We need qualifiers for each of the components of each side - what side they fought on (allied with?), how many of them were there, deaths, injured, who was their leader etc.

We need to identify the two (or more) sides - the name at the top of each column. Do we call each of these an "alliance", "side"? What works for the Union Army at Bull Run, The Allies at Normandy and for The Wood Elves at the battle of the Five Armies?. These statements will also have qualifiers for leaders, casualties etc.

Notice that if you list the components of each alliance in qualifiers to the alliance statement then there is nowhere to put the leaders and casualties of the component. Because of this I

Oppose this property proposal. We do need an "allied with" property as a qualifier to each component, alongside qualifiers for leaders, casualties, numbers etc.

At least that is how I see it. Joe Filceolaire (talk) 14:55, 3 August 2015 (UTC)

Not done due to lack of consensus. Archived Josh Baumgartner (talk) 22:30, 22 October 2015 (UTC)

students count

Done: count of students (P2196) (Talk and documentation)

Description	the number of students of any type in an educational organization
Data type	Number (not available yet)
Template parameter	enrollment in en:Template:Infobox university
Domain	educational institution (Q2385804)
Allowed values	positive natural number
Example	Jagiellonian University => 51601
Source	official web pages of universities
Proposed by	Apohllo (talk)

Discussion

Students count is useful when comparing the current size of universities as well as for showing historical data. Apohllo (talk) 22:00, 14 January 2015 (UTC)

domain university (Q3918) should be replaced by educational institution (Q2385804) --Чаховіч Уладзіслаў (talk) 00:18, 25 January 2015 (UTC)

Done Apohllo (talk) 23:13, 26 January 2015 (UTC)

Support --Чаховіч Уладзіслаў (talk) 16:56, 27 January 2015 (UTC)
Comment A more general property makes sense. Besides students count being a very awkward wording, this need is really more akin to population (P1082), but for organizations instead of places. Something signifying that this university has this many students enrolled, this organization has this many members, etc. I'm blanking on the generic word, unless it's simply count, used with the qualifier of (P642). So Jagiellonian University (Q189441) would have count => 51601 of (P642) => student and count => XXXX of (P642) => faculty, etc. Some organizations may have a need to clarify between undergrad/grad students, or on-campus/online students, too, which qualifiers could help distinguish. (And Mensa International (Q184194) would have count 120,000 of (P642) => members, and so on.) Perhaps? Sweet kate (talk) 18:15, 19 March 2015 (UTC)
- Comment I do not think that this proposal may refer to the number of members and has already created a separate proposal below. Perhaps it's really enough to indicate the current number of students. However, there are other property related specifically to the students, it's the number of graduates (by years). It has a much better supported by source. —putnik 16:06, 4 August 2015 (UTC)
Comment Please see my comment at Wikidata:Property proposal/Generic#number of entries/articles. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:02, 24 April 2015 (UTC)
Support --Derfesl (talk)
Support --- Jura 05:45, 8 August 2015 (UTC)
Archived Josh Baumgartner (talk) 22:31, 22 October 2015 (UTC)

Official symbol/insignia

Done: official symbol (P2238) (Talk and documentation)

Description	Official symbol of an administrative entities
Represents	national symbol (Q1128637)
Data type	Item
Domain	Subclass of administrative territorial entity (Q56061)
Example	Utah (Q829) → Rocky Mountain elk (Q742914) with qualifier P794 (P794) → animal (Q729)
Source	Wikipedia articles, often has lists like list of Utah state symbols (Q4436907)

Motivation

Many administrative units have one or many symbols/insignia officially adopted, for example the US states have symbol animal, bird, tree, flower and so on (see Template:Infobox U.S. state symbols (Q10861166)). As it would be ridiculous to add a property for every kind of symbol kind, the P794 (P794) should be used to specify the actual symbol class. Ahoerstemeier (talk) 11:22, 21 July 2015 (UTC)

Discussion

~~Oppose~~. I think we should have separate properties for official state flower, animal, song, fossil, book, beverage, etc. Use "subproperty of (P1647)" to link them to a property like this. We should also have a property for the official symbol of organisations - most political parties have one and many companies do too and these are not always trademarked. e.g. the US Republican party Elephant. Joe Filceolaire (talk) 16:34, 22 July 2015 (UTC)

Second thoughts.

Support We should have specific properties but we should have a generic property too for strange cases which don't match any of the specific properties. Joe Filceolaire (talk) 15:00, 3 August 2015 (UTC)

Which specific properties would you propose? I plan to use this property especially for the province of Thailand (Q50198), which have a tree, flower and aquatic animal, and if I understand the Thai infobox correctly, also a color. It would avoid some extra bot work if I would then use the specific properties directly. Ahoerstemeier (talk) 09:11, 5 August 2015 (UTC)

I suggest you propose them all and see what gets approved. Others may have a different opinion from me. Joe Filceolaire (talk) 16:33, 5 August 2015 (UTC)

Support I like it. Antrocent (talk) 23:30, 3 August 2015 (UTC)
Support --Pasleim (talk) 16:12, 25 August 2015 (UTC)
Comment The property is already created, so it's to late to make any larger changes. One problem I see here is the word "official". The provinces here have a lot of symbols attached to them, but I cannot say that they are "offical" in any way. -- Innocent bystander (talk) 09:04, 20 October 2015 (UTC)
Done by Mbch331 (talk • contribs • logs)
Archived Josh Baumgartner (talk) 22:44, 22 October 2015 (UTC)

number of members

Not done

Description	total number of members of a organization at a given point in time
Data type	Number (not available yet)
Template parameter	en:infobox organization membership
Domain	organization (Q43229)
Example	Mensa International (Q184194) → 121000±1000

Motivation

For non-profit organizations property employees (P1128) is not suitable. Property number of participants (P1132) close in meaning, but refers to events. Therefore we need a separate property for the number of members of the organization. —putnik 14:54, 4 August 2015 (UTC)

Discussion

Oppose Use has part(s) (P527) member (Q9200127) (or, if that's overly specific, create another item for "ordinary member"), with qualifier quantity (P1114), like this. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:24, 6 August 2015 (UTC)
Oppose This property: member count (P2124) was just created and should work for the example. --Tobias1984 (talk) 11:47, 29 September 2015 (UTC)
Marking as Not done because member count (P2124) has already been created via a different proposal. - Nikki (talk) 12:08, 1 October 2015 (UTC)
Archived Josh Baumgartner (talk) 22:46, 22 October 2015 (UTC)

World Heritage Site type

Withdrawn

Description	the type of World Heritage Site (allowed values are only "Cultural site", "Natural Site" or "Mixed Site")
Data type	Item
Template parameter	"designation1_type" in Infobox: designation list - example Wikipedia article
Domain	Any World Heritage Site (list of World Heritage Sites)
Allowed values	Must be one of 3 items: "Cultural site", "Natural site" or "Mixed site" (items to be created)
Example	Sikhote-Alin (Q155322) → "Natural site"
Source	UNESCO World Heritage list
Robot and gadget jobs	Import from, or check for consistency with Wikipedia infobox values. Check that only the 3 allowed items have been used and flag up incorrect values.

Motivation

For example, this will allow lists of Natural, Cultural and Mixed World Heritage sites to be generated automatically (to be used in Wikipedia articles etc). NavinoEvans (talk) 20:13, 26 September 2015 (UTC)

Discussion

Oppose use heritage designation (P1435) instead. --- Jura 22:30, 28 September 2015 (UTC)

@Jura1: Note that we are already using heritage designation (P1435) to indicate that it's a World Heritage Site, so I don't think it will look right to then use the same property for an additional statement about the type of World Heritage Site - we could add a qualifier to all of the existing heritage designation (P1435) = World Heritage Site (Q9259) statements? but we would still need a property to be able to make the statement. NavinoEvans (talk) 17:20, 29 September 2015 (UTC)

There are three ways of doing it:

Sikhote-Alin (Q155322) → "Natural WHS site" (replacing the WHS statement)
Sikhote-Alin (Q155322) → "Natural WHS site" (in addition to the WHS statement)
Sikhote-Alin (Q155322) → World Heritage Site (Q9259) with qualifier P794 (P794) "Natural WHS site"

Which one do you prefer? --- Jura 06:02, 30 September 2015 (UTC)

Option 3 looks good to me. I hadn't thought of using P794 (P794), many thanks. I'm happy to withdraw this property request and run with the suggestion. NavinoEvans (talk) 21:12, 30 September 2015 (UTC)

I marked it as withdrawn based on your statement. Mbch331 (talk) 16:15, 9 October 2015 (UTC)

Archived Josh Baumgartner (talk) 22:49, 22 October 2015 (UTC)

stage in competition

Not done

Description	use for recording the stage in a competition of an individual match
Data type	Item
Template parameter	None - just part of the title usually
Domain	individual matches in any sport - should be subclass of (P279) sporting event (Q16510064)
Allowed values	valid competition stage - e.g. final (Q1366722), semi-final, quarter-final, qualifying round, group stage (presumably other values need adding to this list after discussion, and items need creating where they do not already exist)
Example	1955 FA Cup Final (Q4568367) → final (Q1366722)
Format and edit filter validation	(sample: 7 digit number can be validated with edit filter Special:AbuseFilter/17)
Robot and gadget jobs	Check for contradictions or add statements based on item Label (e.g. all items containing FA Cup Final in label should be 'finals')

Motivation

It would be very difficult to query for say "all World Cup final matches" or "All Wimbledon semi-finals" without this property. At present, you would need to rely on the title of the item which is obviously not ideal, and leaves no option at all for languages missing that do not have a label for the items. NavinoEvans (talk) 11:32, 13 June 2015 (UTC)

Discussion

Comment I use instance of (P31), like

⟨ 1998 FIFA World Cup Final (Q585295)  

 ⟩ instance of (P31) ⟨ list of FIFA World Cup finals (Q2899571)  

 ⟩

. This makes your query trivial :) CLAIM[31:(TREE[2899571][][279279])]. I think this make sense as all competitions or tournaments have their own rules, which make the list of valid competition stages quite hard to complete or easy to misuse. What we actually mean is a property to put a hierarchy on stages items, maybe follows (P155), and a property like opponents who made it to the next step ... That's a model where we got everything we need and that would work with any kind of competitions without problems, without the need to add a new property. To complete

⟨ Foot Worldcup final ⟩ part of (P361) ⟨ Football world cup ⟩

. TomT0m ^(talk) 12:16, 13 June 2015 (UTC)

I can how that would work, but I must say I strongly disagree with that structurally. Interpreting instance of (P31) as "is a" makes the statement

⟨ 1998 FIFA World Cup Final (Q585295)  

 ⟩ instance of (P31) ⟨ list of FIFA World Cup finals (Q2899571)  

 ⟩

completely wrong (it's not a list, it's an individual world cup final match). It would be more appropriate to say

⟨ 1998 FIFA World Cup Final (Q585295)  

 ⟩ instance of (P31) ⟨ FIFA World Cup final ⟩

, which is what I have been doing for FA Cup Final matches until now.

However, although it's technically correct to use

⟨ 1954 FA Cup Final (Q10858417)  

 ⟩ instance of (P31) ⟨ FA Cup Final (Q4484477)  

 ⟩

, it seems IMHO much better to use

⟨ 1998 FIFA World Cup Final (Q585295)  

 ⟩ instance of (P31) ⟨ association football match (Q16466010)  

 ⟩

. Thinking about how we use instance of (P31) human (Q5) for people (rather than call them instances of poets, scientists, disabled people, or any other potential sub-classification), it would appear more consistent to stop at the level of 'football match' and use other properties to specify the other more specific attributes of the match. I'm aware that this is really a matter of opinion and very much related to how instance of (P31) should be used - of course I'm happy to back down if nobody agrees with me :)

Yes, this is a list page but the case for list pages versus classes is still a mess, as a list is often just a list of instances, people like Filceolaire just think they should be merged with classes items :) ... Did you read Help:Classification ? This explains how classification works in languages such as OWL. Their is a strong relations beetween classes and queries in those languages, as a query can define what a class means, beeing a worldcup match can be defined as beeing a football match in the worldcup competition for example. In Wikidata this could mean that, if a user says that an item is an instance of world cup final, a bot could just as well fill the part of <World cup> statement just by knowing the query ... If the item is tagged as an instance but does not appear in the query, then we would know there is missing informations in this item. I think such possibilities should not be avoided just because a few number of classes like disabled people are not really acceptable. Is see no problem in the poet example, as the class of all poets can be defined by queries about whether or not they are authors of a poem item, or if they has poet as an occupation. If you're talking about peoples, just classifying them as people is not much of a classification :) TomT0m ^(talk) 13:17, 13 June 2015 (UTC)

Many thanks for the clarification... I have no real experience with classification beyond what I've picked up from using Wikidata, so will happily go along with your recommendations as you clearly have the experience to back up your suggestions :) I see what you mean about the lists vs classes case, it seems like one of the many areas where we're a little way of anything concrete being decided. I had read the Classification help page before, but found it very useful to read again. I would personally love to see a Wikidata task force set up for classification as I keep feeling it's something that the layman has little chance of entering in a consistent way, but that's obviously a discussion for another place..

I'm happy to carry on using instance of (P31) FA Cup Final (Q4484477) (or whatever other type of match is applicable) as that's what I've been using so far anyway - it doesn't match how other editors seem to have been entering data, but we'll get there in the end. It does indeed mean there is no need for the suggested property, unless anyone else raises valid disagreement with the points made here? NavinoEvans (talk) 20:19, 13 June 2015 (UTC)

Oppose, per above discussion, instance of (P31) seems sufficient for this purpose. Josh Baumgartner (talk) 17:32, 17 June 2015 (UTC)

Oppose create an item for each stage and then use instance of (P31) --Pasleim (talk) 15:51, 25 August 2015 (UTC)

For English soccer we have FA Cup Final (Q4484477) and list of FA Cup finals (Q769746) but in most cases we just have a "list of" article like en:List of FIFA World Cup finals. This article is a collection of all the FIFA finals so in wikidata terms it is a class. I agree that <'instance of:'List of FIFA World Cup finals'> doesn't make sense in English and that is why the wikidata item associated with this article should be renamed "FIFA World Cup finals" with "List of FIFA World Cup finals" as an alias and it should have a "list of" statement and a "subclass of" statement and it should be used as a target for 'instance of' statements, so that is what I am going to do. Joe Filceolaire (talk) 00:01, 26 August 2015 (UTC)

On a separate issue there is some work to be done to figure out how to show the results of all the matches in knockout competitions like the FA Cup and the final stage of the FIFA World Cup and most tennis tournaments. League tables are already pretty much sorted. Joe Filceolaire (talk) 17:32, 26 August 2015 (UTC)

Not done due to lack of consensus. Archived Josh Baumgartner (talk) 16:16, 23 October 2015 (UTC)

number of trains involved

Withdrawn

Description	number of trains involved in a rail accident
Data type	Number (not available yet)
Template parameter	"trains" in en:template:infobox rail accident
Domain	railway accidents (train wreck (Q1078765))
Allowed values	integer >= 0
Example	Clapham Junction rail crash (Q5125870) -> 3, Hixon rail crash (Q5872894) -> 1
Format and edit filter validation	?
Robot and gadget jobs	import from Wikipedia infoboxes

Motivation

I've been improving the Wikidata entries for rail crashes today and this is something that is missing. Thryduulf (talk: local | en.wp | en.wikt) 18:13, 16 August 2015 (UTC)

Discussion

There is no reason to limit the number involved to trains. We already have number of participants (P1132) to express the number involved in an event, and a train accident is an event. What about a construction like

⟨ the accident ⟩ participant (P710) ⟨ train (Q870)  

 ⟩
number of participants (P1132) ⟨ 3 ⟩

?

Like that if a car and a human was involved we can add

⟨ the accident ⟩ participant (P710) ⟨ train (Q870)  

 ⟩
number of participants (P1132) ⟨ 3 ⟩

⟨ the accident ⟩ participant (P710) ⟨ human (Q5)  

 ⟩
number of participants (P1132) ⟨ 2 ⟩

⟨ car ⟩ participant (P710) ⟨ human (Q5)  

 ⟩
number of participants (P1132) ⟨ 1 ⟩

... author TomT0m / talk page 13:58, 17 August 2015 (UTC)

Hmm, I had assumed that "participants" was restricted to humans. Thryduulf (talk: local | en.wp | en.wikt) 20:57, 20 August 2015 (UTC)

Thryduulf: participant (P710) is already being used for teams in a league table and in principle can be used for an event however it has datatype 'item'. For many events the number of participants may be large and we may not have an item for each participant so there may be a need for a 'number of participants' or 'number involved' property. If you want to amend this proposal so it could do that job you will have my support. As currently proposed it is, I believe, too narrow and specialised. Joe Filceolaire (talk) 17:07, 21 August 2015 (UTC)

I'm happy to broaden. I don't know how to do that and keep a specification of what the type of participant is. Something like "number of <item> involved: n" would be good. Either suggest it here or propose it separately and I'll withdraw this. Thryduulf (talk: local | en.wp | en.wikt) 21:28, 22 August 2015 (UTC)

Question Now units are available I've been using this to achieve the aim, e.g. at British Airways Flight 2276 (Q20962419): number of participants (P1132) → 172 unit human (Q5), 1 unit aircraft (Q11436). At Bridgnorth Cliff Railway (Q4966907): number of participants (P1132) → 2 unit vehicle (Q42889). Should I continue doing this or is there a better way? Thryduulf (talk: local | en.wp | en.wikt) 12:42, 13 September 2015 (UTC)

@Thryduulf: This works pretty well. I think it alleviates the need for this property. Josh Baumgartner (talk) 05:19, 16 September 2015 (UTC)

Wow. I never thought of doing that.

Oppose. We don't need this property now we can do that. Joe Filceolaire (talk) 03:21, 23 September 2015 (UTC)

Withdrawn - per the above this is not needed now we have units. Thryduulf (talk: local | en.wp | en.wikt) 14:36, 24 September 2015 (UTC)
Not done as not needed any more. Archived Josh Baumgartner (talk) 16:19, 23 October 2015 (UTC)

Properties for specific levels of administrative units

Currently, items of populated places (villages, towns, cities, city districts etc.) are equiped with located in the administrative territorial entity (P131) which should contain the immediately superior unit. Such a statement can be (hypotetically) supplemented with a qualifier which would define which kind (level) of the administrative unit is linked.

However, the invoke function is not (and will be not in the foreseeable future) able to work with qualifiers, is not able select the requested value by qualifier and is not able to work with a multilevel statement tree structure (imitating the categorization structure) and extract the appropriate value from them using structured conditioning. Said simply: when we have a specific item of any place - we cannot invoke the municipality, the district, the region or the land where the place is. The only level of administrative unit which have its specific property and is widely used for populated places is country (P17).

The new function of Wikidata:Arbitrary access can be very helpful e.g. in monument lists created for Wiki Loves Monuments. They can help to fulfill specifying entries in the list tables as well as prefilled at the file description pages of images uploaded using campaing link to UploadWizard as well as to help with precise automatic categorization of such images. However, the statements need to be specific enough to be usable for practical work.

I propose to create specific properties for specific levels of administrative units. In case of my country (Czechia), we need properties:

municipality
district
region
land

Format and usage should be analogous to the existing country (P17):

Data type: item
Infobox parameters: corresponding (can be simply imported)
Domain: place
Examples:
- municipality (P?) of Matějov (Q12036136) → Sedlec-Prčice (Q368888)
- district (P?) of Matějov (Q12036136) → Příbram District (Q852457)
- region (P?) of Matějov (Q12036136) → Central Bohemian Region (Q188399)
- land (P?) of Matějov (Q12036136) → Bohemia (Q39193)
- country (P17) of Matějov (Q12036136) → Czech Republic (Q213)

--ŠJů (talk) 07:25, 24 August 2015 (UTC)

Do you propose to create separate properties for each level of adminunits?--Ahonc (talk) 07:34, 24 August 2015 (UTC)

Yes. As explained above. --ŠJů (talk) 07:42, 24 August 2015 (UTC)

@ŠJů: Can you work with Lua ? It's not really hard to make an equivalent of Invoke in lua using Wikidata API, who can manipulate qualifiers. I'll encourage you to poke the lua developers of the Wikipedia of your language. By knowing the instance of (P31) of a division, you can pretty much know which kind of division is the upper level of that item. Note that the administrative model of division differs from one country to another, which would make a mess of properties ... Other projects are trying to work this "upper division" model, for project consistency sake I'd say it's better that all the administrative division local project stay aligned with this. author TomT0m / talk page 10:05, 24 August 2015 (UTC)

But why we need qualifiers? We should use P131 for Q12036136, then P131 for Q852457, then P131 for Q188399.--Ahonc (talk) 10:18, 24 August 2015 (UTC)

Another problem with the model ŠJů propose above, is that some administrative divisions maybe have the same "name" in one language, but completly different names in other languages. In Swedish we use different words for a county in US, UK, Denmark, Norway, Sweden and Russia. We would then need one property for each type of administrative level in every present and historical country. We would soon reach 1000 properties only for administrative units. The disadvantage of the other model is that the simple hierarchy is sometimes missing, and in Sweden is it only present in the relation between municipalities and counties. -- Innocent bystander (talk) 11:29, 24 August 2015 (UTC)

Why? We may create universal properties for this, e.g. Level1AD, Level2AD, Level3AD, Leve4AD, Level5AD, Level6AD etc.--Ahonc (talk) 11:36, 24 August 2015 (UTC)
Sweden: Then I don't get how using specialized properties makes things really for Sweden :) If there is no hierarchy or a flat one, then ... whatever you do things will remains flat. author TomT0m / talk page 11:37, 24 August 2015 (UTC)

Oppose. We used to have 'type of administrative territorial entity' and we replaced it with instance of (P31) and that is what we can use for this. <instance of (P31):municipality of the Czech Republic (Q5153359)>. There is a reason we do not put the higher levels on each item and that is because these hierarchies change from time to time. If the info is in one place only then it is much easier to keep it accurate. This does mean that you need to go to a different item to get the links to the higher levels and that is something we live with.

Taxa used to have something similar and the specialised properties for upper levels were deleted so now we only have parent taxon (P171). Just be thankful country (P17) is still around. Joe Filceolaire (talk) 13:56, 24 August 2015 (UTC)

Oppose per Joe. --Pasleim (talk) 15:39, 25 August 2015 (UTC)

Oppose per Filceolaire (talk • contribs • logs). Josh Baumgartner (talk) 05:11, 16 September 2015 (UTC)

Not done per no consensus.

Archived Josh Baumgartner (talk) 16:54, 23 October 2015 (UTC)

Road length

Description: length of the road
Datatype: number value
Links:
Infobox parameter example: w:en:Template:Infobox road
Comments: unit in kilometer per default
Will unit conversions be available? --Rs chen 7754 02:17, 3 February 2013 (UTC)
Could be used for rivers also. --Eric-92 (talk) 00:22, 8 February 2013 (UTC)
Also railways, hiking trails, etc. --Avenue (talk) 14:06, 9 February 2013 (UTC)
If it is used for roads, rivers, railways... shouldn't it be called just "length"? Or will we have "road length", "river length", "railway length", "hiking trail length"...? Nikola (talk) 04:44, 8 March 2013 (UTC)

Support --Faux (talk) 19:19, 18 February 2013 (UTC)

Support Agree - needs to be for rivers and similar. Danrok (talk) 20:06, 21 February 2013 (UTC)

Support --Daniel749 talk 09:21, 23 February 2013 (UTC)

I'm a bit concerned about putting these in kilometers by default - we shouldn't be giving preference to one unit of measurement over another. --Rs chen 7754 07:15, 25 February 2013 (UTC)

What would you suggest? We could have two properties; length in km and length in miles. Danrok (talk) 18:18, 25 February 2013 (UTC)

This should really be in some sort of convertible field that accepts both formats, but we'll have to see what the final product is. --Rs chen 7754 23:35, 25 February 2013 (UTC)

There's no need to agree on one unit, Wikidata will have an datatype "Number with unit" witch has an internal unitconversion.

Support as Number with unit. -- MichaelSchoenitzer (talk) 22:16, 13 March 2013 (UTC)

Support to MichaelSchoenitzer. --Ricordi samoa 11:40, 31 March 2013 (UTC)

Support if number with unit datatype is used. —Scott5114 ↗ [EXACT CHANGE ONLY] 02:57, 19 April 2013 (UTC)

Support per above. --Rs chen 7754 03:00, 19 April 2013 (UTC)

~~Support~~ but I think it should just be 'length', that way it can be applied to anything with a length property without having to submit a new property proposal for each type of item used for. ~~Units should be convertible units of distance, but we will have to see how that works as support for this datatype is built.~~ (changed as we now have units and a length property, so no need for this one) Joshbaumgartner (talk)

Oppose if we have generic property length as proposed above, we don't need special properties for roads, rivers, railways...--Giftzwerg 88 (talk) 10:33, 24 January 2015 (UTC)

Oppose, this seems redundant after two years. Sjoerd de Bruin (talk) 20:10, 9 September 2015 (UTC)

Oppose even though I proposed it. --Rs chen 7754 21:41, 20 September 2015 (UTC)

Sorted to Place for further discussion, now that units are available. Josh Baumgartner (talk) 06:36, 29 September 2015 (UTC)

Oppose Use:

⟨ road (Q34442)  

 ⟩ located on linear feature (P795) ⟨ object or value ⟩

Josh Baumgartner (talk) 06:36, 29 September 2015 (UTC)

Not done per lack of consensus.

Archived Josh Baumgartner (talk) 16:58, 23 October 2015 (UTC)

production volume

Done: production rate (P2197) (Talk and documentation)

Description	Amount of a given material produced in a given time
Data type	number with dimension-invalid datatype (not in Module:i18n/datatype)
Domain	Materials
Example	iron (Q677) --> 2,400,000,000 [unit:tons per year]
Proposed by	--Jakob (talk)

Discussion

@Filceolaire, Mcnabber091: Thoughts? --Jakob (talk) 18:49, 19 March 2014 (UTC)

Sounds good Jakob, What is the source? Mcnabber091 (talk) 09:53, 26 March 2014 (UTC)

@Mcnabber091: As far as I know, there's no one source for all materials. Nature's Building Blocks An A-Z Guide of the Elements New Edition (Q15925426) is a good source for chemical elements. --Jakob (talk) 11:58, 26 March 2014 (UTC)

SupportMcnabber091 (talk) 10:15, 16 May 2014 (UTC)

Support --Eurodyne (talk) 21:10, 11 October 2014 (UTC)

On hold Pending units data type. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 08:24, 23 March 2015 (UTC)

@Jakec, Mcnabber091, Mcnabber091, Eurodyne, Pigsonthewing: Please final thoughts, now that we have the datatype. --Tobias1984 (talk) 20:09, 15 September 2015 (UTC)
Done by GZWDer (talk • contribs • logs)
Archived Josh Baumgartner (talk) 17:06, 23 October 2015 (UTC)

maximum operating altitude

Done: maximum operating altitude (P2254) (Talk and documentation)

Description	ceiling or maximum density altitude specified for an aircraft to operate at
Represents	ceiling (Q1221006)
Data type	Number (not available yet)
Template parameter	"ceiling main" in en:template:Aircraft specifications
Domain	aircraft (Q11436)
Allowed values	quantity with units of distance
Example	⟨ Stout 2-AT Pullman (Q7620606)    ⟩ ceiling Search ⟨ 15000 feet ⟩

Motivation

Basic aircraft specification. Josh Baumgartner (talk) 07:06, 16 September 2015 (UTC)

Discussion

Support. but I prefer the label "maximum altitude"to avoid confusion with architecture. Joe Filceolaire (talk) 04:06, 20 September 2015 (UTC)
- Comment change made. Josh Baumgartner (talk) 00:38, 20 October 2015 (UTC)
Oppose use a more generic property instead. --- Jura 22:28, 28 September 2015 (UTC)
- Question @Jura1: Do you have one in mind? Josh Baumgartner (talk) 06:26, 29 September 2015 (UTC)
  - Several are possible, but I let you choose or propose one. --- Jura 07:30, 29 September 2015 (UTC)
Support per Joe Filceolaire. I see significantly less benefit in requiring complex statements using one or more generic properties than a single dedicated property that makes data entry and reuse simple. Thryduulf (talk: local | en.wp | en.wikt) 18:53, 30 September 2015 (UTC)
Support per Joe Filceolaire. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:34, 8 October 2015 (UTC)
Done with suggested changes. Archived Josh Baumgartner (talk) 18:45, 23 October 2015 (UTC)

Debrett's People of Today ID

Done: Debrett's People of Today ID (P2255) (Talk and documentation)

Domain	people
Robot and gadget jobs	Could be imported into the mix 'n' match tool.

Motivation

Much like Who's Who, Debrett's is a reference guide for individuals in British society, such as political figures. Rock drum (talk) 17:16, 13 October 2015 (UTC)

Discussion

@Jdforrester, Andrew Gray: May be of interest. Rock drum (talk) 17:16, 13 October 2015 (UTC)

I'm in favour. James F. (talk) 16:08, 17 October 2015 (UTC)
Support Josh Baumgartner (talk) 20:20, 22 October 2015 (UTC)

@Rock drum, Jdforrester, Andrew Gray, Joshbaumgartner: Done Debrett's People of Today ID (P2255) Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:11, 23 October 2015 (UTC)

Archived Josh Baumgartner (talk) 23:10, 26 October 2015 (UTC)

coined by

Not done

Description	To indicate the person (or maybe sometimes the organisation / group of people) who has used a specific word for a concept for the first time.
Data type	Item
Domain	any concept, usually a newly 'named' term, movement...
Allowed values	human (Q5)
Example	action painting (Q217213) → Harold Rosenberg (Q959768); relational art (Q2566149) → Nicolas Bourriaud (Q1979892), stipule (Q304216) → Carl Linnaeus (Q1043)
Source	external reference

Motivation

I've encountered many cases where a specific word for a concept was first used or coined by someone. No property exists for this yet. Quite usual in the case of artistic and cultural movement, see examples above. discoverer or inventor (P61) doesn't seem correct here. Spinster (talk) 19:05, 14 June 2015 (UTC)

Discussion

@Spinster:

Weak oppose Mmm Wikidata is not about terms but about concepts, and is mutilingual, so this does not fits well. This could maybe be applied as a qualifier to properties like official name or likes, but definitely not to items. I don't understand why discoverer or inventor (P61) would not fit. TomT0m ^(talk) 19:16, 14 June 2015 (UTC)

See the examples above. You are correct that I should not say these are terms - they ARE concepts, regardless of their translation. I updated the proposal to reflect this. For instance, the Finnish Wikipedia article about action painting, Toimintamaalaus, mentions Harold Rosenberg. Movements and concepts like action painting (Q217213) and relational art (Q2566149) and stipule (Q304216) have definitely not been discovered or invented by those people (usually critics, researchers). They were "seen" and named by them for the first time, that's a different thing. Spinster (talk) 19:26, 14 June 2015 (UTC)

Comment More than 6,000 mentions of the exact phrase 'coined by' on enwp. Spinster (talk) 19:26, 14 June 2015 (UTC)

@Spinster: The person that created a new name for that concept ... did not invent that term ? There still is a confusion beetween the concept and the term :). A man can discover or describe a concept, like a taxon, then create a name for this new concept. Actually we already have a property taxon author in taxonomy I could very well see generalized. But I think a construction

⟨ the stuff ⟩ name (or original name) (or late name) Search ⟨ thename ⟩
author (P50) ⟨ the person ⟩

is pretty good, it does not talk of the concept, just of the name (plus the inventor of a treasure, or a planet, for example, obviously did not create them, hence the name of the property). I don't see a need for this property. Statements with this pattern will just have to be labelled coined by in infoboxes since it's a common way to express this in english. TomT0m ^(talk) 19:42, 14 June 2015 (UTC)

@Spinster: stipule (Q304216) was coined by Carl Linnaeus (Q1043). I have some doubts. --Succu (talk) 21:16, 14 June 2015 (UTC)

Oppose, refers to terms instead of the actual entities. This is Wiktionary's domain. --Yair rand (talk) 10:02, 15 June 2015 (UTC)
Support though I think 'first described by' might be a better name, or should, at least, be an alias. Filceolaire (talk) 18:12, 18 June 2015 (UTC)
Perhaps this should be a qualifier on a mono-lingual text property? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:28, 20 June 2015 (UTC)
Oppose this is very problematic, as it really is a qualifier for the label, something that is not supported. Perhaps if we had a property 'referred to as' (monolingual type) with qualifier 'by' that could get the job done, but if you add a 'coined by' directly to the item based on one label, and the label is later changed, then the 'coined by' statement is no longer valid. Josh Baumgartner (talk) 17:29, 15 September 2015 (UTC)

@Spinster, TomT0m, Filceolaire, Pigsonthewing:

Not done Archived Josh Baumgartner (talk) 18:05, 27 October 2015 (UTC)

population density

Not done

Description	The relationship between population and surface of an administrative division (city, region, state, etc).
Represents	population density (Q22856)
Data type	Quantity
Domain	Islands, countries, cities, regions...
Example	Esino Lario (Q43049) → 42.11
Robot and gadget jobs	Bots can estract this value from wikipedia templates and from the relationship between population (P1082) and area (P2046)

Motivation

Useful to indicate the population density of a Country, city, province, region, islan, etcetera. Yiyi .... (talk!) 15:40, 10 September 2015 (UTC)

Discussion: Comment Not sure it's a usefull property, since we can calculate it with the land area and the propulation.

This has been proposed before. See WD:Property_proposal/Archive/10#Population density / تراکم جمعیت. Oppose. --Yair rand (talk) 23:05, 10 September 2015 (UTC)

Oppose per Yair. Joe Filceolaire (talk) 14:21, 14 September 2015 (UTC)

Not done Can be calculated. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:19, 20 September 2015 (UTC)

@Yair rand, Pigsonthewing, Yiyi, Filceolaire: et al. I have a source in my knee: Småorter 1990 (Q20087097). It tells that Korsbacken (Q2791237) in 1990 had a population of 68 in an area of 5 hectare (Q35852) with a population-density of 1,278 km^-2. It is obvious here that a "calculation" would give the result 1360 km^-2, but that is not what the source tells. Should we rely on sources or our own calculations? -- Innocent bystander (talk) 18:07, 21 October 2015 (UTC)

A source may have known the area to more significant figures than were displayed in the source. Thus the sources calculation may be more accurate than redoing the calculation based on the published population and area. So the population density from the source should be used. Jc3s5h (talk) 18:59, 21 October 2015 (UTC)

Archived Josh Baumgartner (talk) 19:06, 27 October 2015 (UTC)

military unit

Not done

Description	Military units to which the subject was attached.
Represents	subdivision (Q14946396)
Data type	Item
Domain	Person
Allowed values	Item should be an instance of human (Q5) or fictional character (Q95074); linked item should be an instance of military unit (Q176799), or child thereof.
Example	Guy Gibson (Q1385145) → No. 617 Squadron Royal Air Force (Q1158122) Ty Carter (Q16208557) → 61st Cavalry Regiment (Q4641796), 4th Infantry Division (Q231293) & 7th Infantry Division (Q261179)
Source	External references, Wikipedia biographies etc

Motivation

To allow items about soldiers, etc, to cover their military career; complementing the high level military branch (P241) and military or police rank (P410). We have commander of (DEPRECATED) (P598), which is similar, but this property could cover all members of a unit rather than just the commanding officer. - AdamBMorgan (talk) 10:56, 5 June 2015 (UTC)

Discussion

Oppose. We can use ~~part of (P361) (or employer (P108))~~member of (P463) with qualifying dates to detail a military career without needing to create a too-specific property. Josh Baumgartner (talk) 17:52, 5 June 2015 (UTC)
I don't think part of (P361) is appropriate here. I can see employer (P108) but I'd put that at the same level of military branch (P241); a soldier is technically employed by, for example, the United States Army and not by 4th Infantry Division or 61st Cavalry Regiment. (S/he could even be said to be employed by the state or the DoD/MoD equalivalent of that state). - AdamBMorgan (talk) 22:10, 5 June 2015 (UTC)

@AdamBMorgan: I can see the point on employer (P108) but can you elaborate on why you don't see part of (P361) as useable? Josh Baumgartner (talk) 15:48, 9 June 2015 (UTC)

Oppose. I agree with @Joshbaumgartner:, though I would propose member of (P463) as a better answer than part of (P361). In other non-military topic areas, the group consensus has been to stay with less specific properties. member of (P463) is not limited to a single value, and can still be qualified with start and end dates. Sweet kate (talk) 19:40, 25 July 2015 (UTC)

member of (P463) does seem to be the better option; I've amended my initial opposition to reflect it. Josh Baumgartner (talk) 21:01, 20 October 2015 (UTC)

Comment how will it be different to military branch (P241)? --Pasleim (talk) 22:13, 6 June 2015 (UTC)

@Pasleim: military branch (P241) is a very high level, very broad category: just country and branch, such as Fooian Army, Fooian Navy, etc (practical examples would be items like Royal Navy (Q172771), Royal Air Force (Q165862), United States Marine Corps (Q11218), United States Army (Q9212) and so forth) — at least in theory and by design; it might be misapplied in practice. This property would be for a lower level, giving specific regiments, squadrons etc (such as Grenadier Guards (Q772054), 45 Commando (Q4638181), 101st Airborne Division (Q165256), 2nd Foreign Infantry Regiment (France) (Q2703367) etc). As such, military branch (P241) is supposed to be limited to a single value, because it is unlikely for a person to transfer between, for example, the Army and the Navy; while this property could potentially have many values, each qualified with start and end dates, because transferring between units within a branch is common. - AdamBMorgan (talk) 11:00, 9 June 2015 (UTC)

@AdamBMorgan, Pasleim: I agree about branch. Even if a person does serve in multiple branches (I served alongside some who did) that is fundamentally different than tracking the actual unit they are part of and needs to be maintained as its own piece of data. For example, a Navy corpsman may be assigned to a Marine unit; they are still Navy, but the unit they are assigned to is not. Josh Baumgartner (talk) 15:48, 9 June 2015 (UTC)

Ok, I see. There was apparently a German translation error, that's why I was confused. --Pasleim (talk) 07:43, 11 June 2015 (UTC)

Support per discussion above. Filceolaire (talk) 22:59, 10 June 2015 (UTC)
Support --- Jura 11:27, 12 June 2015 (UTC)
Support Antrocent (talk) 20:15, 8 July 2015 (UTC)
Comment What's wrong with member of (P463) that we can't use it? —putnik 16:14, 4 August 2015 (UTC)
Comment same as putnik — What's wrong with member of (P463) that we can't use it? it's not an ID, so member of (P463) should be enough :) --Hsarrazin (talk) 19:58, 6 August 2015 (UTC)

Oppose. member of (P463) should just be used. Kharkiv07 (T) 02:30, 8 August 2015 (UTC)
Oppose - I agree we should record this, and I think it will be very useful, but member of (P463) is the most natural solution - indeed, it's already been discussed and approved on Property talk:P463 & Wikidata:Project_chat/Archive/2014/05#Military units. Andrew Gray (talk) 22:15, 3 September 2015 (UTC)
Comment given that "member of sports team", "member of monastic order", "member of political party" and "employer" are separate from member of (P463) or "part of", I don't think we should mix this. --- Jura 10:12, 13 September 2015 (UTC)
Not done due to lack of consensus. Archived @Jura1, Andrew Gray, Kharkiv07, Hsarrazin, Filceolaire, Pasleim: Josh Baumgartner (talk) 19:51, 27 October 2015 (UTC)

NGA artist id

Done: National Gallery of Art artist ID (P2252) (Talk and documentation)

Description	identifier assigned to an artist by the National Gallery of Art in Washington DC
Data type	String
Domain	humans who are artists
Allowed values	\d+
Example	Gerrit Dou (Q335927) → 1239
Source	http://www.nga.gov/content/ngaweb/Collection/artists.html.html
Formatter URL	http://www.nga.gov/content/ngaweb/Collection/artist-info.$1.html
Robot and gadget jobs	Should be added to Mix'n'Match

Motivation

Nice and clean database like Smithsonian American Art Museum person/institution ID (P1795) & National Gallery of Victoria artist ID (P2041), but this is in a league of it's own. They keep information about artists as part of their NGA Online Editions. Very good background information. Multichill (talk) 18:45, 6 October 2015 (UTC)

Support --Jane023 (talk) 18:48, 6 October 2015 (UTC)

Support. Joe Filceolaire (talk) 14:47, 7 October 2015 (UTC)
Support Jonathan Groß (talk) 15:09, 7 October 2015 (UTC)
Support as proposed. Josh Baumgartner (talk) 22:31, 21 October 2015 (UTC)
Support--Oursana (talk) 15:05, 26 October 2015 (UTC)
Done by Multichill (talk • contribs • logs)
Archived @Multichill, Jane023, Filceolaire, Jonathan Groß, Oursana: Josh Baumgartner (talk) 19:56, 27 October 2015 (UTC)

age group

Not done

Description	used for sports teams to describe their age group, such as under XX national teams with XX being a number.
Data type	Item
Domain	youth sports teams and youth national teams
Example	Germany national under-21 football team (Q314851) → under-21 sport (Q21152744)
Robot and gadget jobs	only items that define age groups / age classes are reasonable as values.

Motivation

Sports: To determine the age group different youth club teams or youth national teams are in (see example). Currently, this is only determinded via the title. Yellowcard (talk) 17:35, 19 October 2015 (UTC)

Discussion

Weak oppose I think it's a team classifications issue, so instance of (P31) can deal with it as a classification property. Example

⟨ PSG less than 18 ⟩ instance of (P31) ⟨ less than 18 football team ⟩

The <less than 18 football team> item could have statements like :

⟨ less than 18 football team ⟩ subclass of (P279) ⟨ football team ⟩
⟨ less than 18 football team ⟩ max member age Search ⟨  18 ⟩
⟨ less than 18 football team ⟩ min number age Search ⟨  16 ⟩

for example. author TomT0m / talk page 17:56, 19 October 2015 (UTC)

@TomT0m: Sounds convincing. Is there any "min number age" and "max number age" properties yet? Yellowcard (talk) 18:07, 19 October 2015 (UTC)

After more thoughts about it: I don't think this solution would be the best one. Using items like under-21 sport (Q21152744) with a new property "age group" would be better and could be used as qualifier for instance of (P31) → national association football team (Q6979593) instead of a new statement. This way we would still have the direct relation of the under XX national teams to national association football team (Q6979593) which has certain advantages escpecially for the direct access from Wikipedia versions. Yellowcard (talk) 23:46, 19 October 2015 (UTC)

Oppose this can be done with competition class (P2094), which was specifically designed for this purpose:
⟨ Germany national under-21 football team (Q314851)    ⟩ competition class (P2094) ⟨ under-21 sport (Q21152744)    ⟩
Josh Baumgartner (talk) 22:59, 22 October 2015 (UTC)

@Joshbaumgartner: Nice, thank you! I will use that one. This proposal can be closed, then. Yellowcard (talk) 10:27, 24 October 2015 (UTC)

@Yellowcard, TomT0m:

Not done as withdrawn by proposer.

Archived Josh Baumgartner (talk) 16:53, 28 October 2015 (UTC)

frequency of event

Done: event interval (P2257) (Talk and documentation)

Description	The standard nominal interval between scheduled recurring events
Data type	quantity with units-invalid datatype (not in Module:i18n/datatype)
Template parameter	?
Domain	recurring event (Q15275719)
Allowed values	positive time values
Example	Olympic Games (Q5389) → 4 years; Gordon Bennett Cup in ballooning (Q1537962) → 2 years; Jersey Battle of Flowers (Q539642) → 1 year; Labour Force Survey (Q6467417) → 3 months (qualifier: applies to part (P518) United Kingdom (Q145))

Motivation

I couldn't find any way of representing this with existing properties. Thryduulf (talk: local | en.wp | en.wikt) 19:47, 30 September 2015 (UTC)

Discussion

Support I'm sure we have this, but "no, we don't". --- Jura 12:01, 1 October 2015 (UTC)

Comment We have instance of (P31)=annual event (Q18574946) etc. --Pasleim (talk) 12:12, 1 October 2015 (UTC)

Do we have (or want) items for events that occur every 10 years, 1 week, 3 months, 18 months, 90 days, etc, etc? Thryduulf (talk: local | en.wp | en.wikt) 14:55, 1 October 2015 (UTC)

Support. Joe Filceolaire (talk) 23:41, 6 October 2015 (UTC)
Support per Jura, this one certainly seems overdue. Josh Baumgartner (talk) 16:49, 23 October 2015 (UTC)
Support --AmaryllisGardener ^talk 15:39, 26 October 2015 (UTC)
Done Archived Josh Baumgartner (talk) 20:06, 28 October 2015 (UTC)

Mobile country code

Done: mobile country code (P2258) (Talk and documentation)

Description	The ITU-T Recommendation E.212 defines mobile country codes as well as mobile network codes
Represents	mobile country code (Q978801)
Data type	String
Domain	mainly countries
Allowed values	###
Example	Afghanistan (Q889) → 412
Source	Wikipedia article
Robot and gadget jobs	Yes but only once

Discussion

--GZWDer (talk) 12:09, 4 October 2015 (UTC)

Support --- Jura 12:13, 4 October 2015 (UTC)
Support --Pasleim (talk) 20:02, 4 October 2015 (UTC)
Done Archived Josh Baumgartner (talk) 20:30, 28 October 2015 (UTC)

maximum capacity

Withdrawn

Description	maximum number of people that an object or place can hold.
Data type	Number (not available yet)
Allowed values	quantity
Example	Magic Kingdom (Q1324340) →100,000
Source	external reference

Useful info IMO. --AmaryllisGardener ^talk 17:06, 24 October 2015 (UTC)

Discussion

Support. Yes, useful for aircraft, ships, venues, and probably other things too. Thryduulf (talk: local | en.wp | en.wikt) 02:13, 25 October 2015 (UTC)
Support. Joe Filceolaire (talk) 21:11, 25 October 2015 (UTC)
Oppose. How does this differ from maximum capacity (P1083) "number of people allowed for a venue or vehicle"? Bovlb (talk) 21:18, 25 October 2015 (UTC)

@Bovlb, Filceolaire, Thryduulf: I was not aware of that property, since "maximum capacity" wasn't listed as an alias, my bad. Withdrawn. --AmaryllisGardener ^talk 15:33, 26 October 2015 (UTC)

Archived Josh Baumgartner (talk) 20:32, 28 October 2015 (UTC)

Mobile Network Code

Done: mobile network code (P2259) (Talk and documentation)

Description	The ITU-T Recommendation E.212 defines mobile country codes as well as mobile network codes.
Represents	mobile network code (Q31623)
Data type	String
Domain	companies
Allowed values	##, ###
Example	China Mobile (Q741618) → 00, 02, 07
Source	Wikipedia article
Robot and gadget jobs	Yes

Discussion

--GZWDer (talk) 12:08, 4 October 2015 (UTC)

Support. I have two follow-up questions. First, what happens when many companies have the same MNC? That is, are mobile network codes a qualifier of mobile country codes or countries? Many mobile companies have the MNC of 01, but only one per country. Should we treat MCC and MNC as interrelated to make each pair distinct? And second, how do we deal with one company with many MNCs, like worldwide telecom Orange? Do we simply give it many MNC values, or create items for each national subsidiary with one MNC each, or something else? Runner1928 (talk) 15:27, 4 October 2015 (UTC)

Comment I suppose you'd have to add the country code as a qualifier. --- Jura 22:20, 4 October 2015 (UTC)

Done Archived Josh Baumgartner (talk) 22:12, 28 October 2015 (UTC)

subproperty (external)

Done: external subproperty (P2236) (Talk and documentation)

Description	all resources related by that property from an external vocabularies are also related this property
Data type	URL
Example	http://rdf.freebase.com/ns/book.book.genre → genre (P136)

Motivation

The goal of this property is to allow to store mapping from external vocabularies into Wikidata. For example http://rdf.freebase.com/ns/book.book.genre → genre (P136) states that the property /book/book/genre from Freebase should be mapped to the Wikidata property genre (P136). A first use case of this property may be to store in Wikidata the mapping used for the Freebase to Wikidata migration. Tpt (talk) 19:11, 15 June 2015 (UTC)

Discussion: Support Will help with mappings. --Denny (talk) 22:45, 17 June 2015 (UTC); Support Symmetric with the naming proposal for "subproperty of external property", I would call this one "superproperty of external property". --Markus Krötzsch (talk) 09:46, 18 June 2015 (UTC)

Support. Filceolaire (talk) 21:30, 18 June 2015 (UTC)

Comment If it is intended for perfect mappings, I would still recommend equivalent property (P1628). --Hjfocs (talk) 11:01, 19 June 2015 (UTC)

units used for this property

Done: no label (P2237) (Talk and documentation)

Description	...
Data type	Item
Template parameter	display in Template:Property documentation
Domain	property with numbers and units
Allowed values	items for units or Q21027105
Example	temperature (P2076) → degree Celsius (Q25267), kelvin (Q11579), degree Fahrenheit (Q42289)

Support makes it easier to find out what item to use. Inspired by Wikidata:Units. --- Jura 08:01, 28 September 2015 (UTC)

scheme of external identifier of this property

Withdrawn

Description	defines the uniqueness of an external identifier as used by a property (issuing authority, naming standard)
Data type	Item
Template parameter	display in property documentation
Domain	properties for external identifiers
Allowed values	items for specific schemes
Example	Wikimedia language code (P424) → unique identifier issued by single authority; Mobile Network Code (P3000) → identifiers issued by multiple authorities, unique within authority, defined by standard

Support helps group external identifiers --- Jura 01:45, 17 October 2015 (UTC)
Support. Joe Filceolaire (talk) 18:22, 19 October 2015 (UTC)
withdrawn, used other property instead. --- Jura 16:31, 28 October 2015 (UTC)

usage of "no value" for this property

Withdrawn

Description	clarify if and how no value is being used in combination with this property
Data type	Item
Template parameter	display in property documentation
Domain	properties
Allowed values	items for specific uses
Example	inventory number (P217) → "no value means: database checked, but no entry found"; instance of (P31) → "do not use no value"

Support clarifies use of this special value --- Jura 01:45, 17 October 2015 (UTC)
Support. Joe Filceolaire (talk) 18:23, 19 October 2015 (UTC)
withdrawn, can stay on talk, I guess. --- Jura 16:31, 28 October 2015 (UTC)

subproperty of (external)

Done: external superproperty (P2235) (Talk and documentation)

Description	all resources related by this property are also related that property from an external vocabulary
Data type	URL
Example	father (P22) → http://schema.org/parent

Motivation

The goal of this property is to allow to have something less strong than equivalent property (P1628) in order to do mapping with external vocabularies. For example there is no "father" property in schema.org but only a "parent" property. With this new property we would be able to declare directly in Wikidata that father (P22) "subproperty of" http://schema.org/parent . The main use case of such mappings would be to create tools that outputs Wikidata content directly in these vocabularies. Tpt (talk) 19:11, 15 June 2015 (UTC)

Discussion: Support I would rename it to 'superproperty' though, but that can be done later too. --Denny (talk) 22:44, 17 June 2015 (UTC); Support But I would prefer "subproperty of" since this is already used in RDFS, OWL, and Wikidata (subproperty of (P1647)). However, to clarify the distinction from subproperty of (P1647), I would suggest to call this one "subproperty of external property". --Markus Krötzsch (talk) 09:46, 18 June 2015 (UTC); Comment A general concern with all of these external mapping relations is that Wikidata has its own data values that are often more complex than plain RDF (XML Schema) values. Therefore, in a strict sense, it does not make sense to declare a Wikidata property to be (sub/super/equivalent)Property with an external RDF property that takes completely different types of values. The documentation of these mapping properties should explain how exactly the value of an external property is to be converted from a certain XML Schema datatype to Wikidata and vice versa. For example, if you state father (P22) "subproperty of" http://schema.org/parent then it involves the assumption that the Wikidata items that are values of P22 are converted to URIs before being taken as values for http://schema.org/parent. We can do this for all Wikidata property values based on our RDF mapping but for complex values this mapping has several forms for a value. For super properties it is more complicated since the mapping might depend on the concrete use. For example, if some external property "http://example.org/father" would be declared to be a subproperty of Wikidata's P22, then this would probably mean: "If an external object S has an item Sitem in Wikidata, and there is an external statement "S http://example.org/father O", and O is a URI that has an item Oitem in Wikidata, then the statement "Sitem P22 Oitem" can be imported. Making this mapping requires a lot more background information on how external URIs (or other XSD values) are mapped to Wikidata. There can also be cases where an external triple, even though it uses a property that is mapped to a Wikidata property, cannot be related to Wikidata in any way, since there is no meaningful mapping of the subject and object (for example, a birthdate should use some time-value as its object; if it uses some other thing then it makes no sense). This is why all of these mapping properties are very different from the standard "subproperty of" et al. relations in RDFS and OWL. --Markus Krötzsch (talk) 09:46, 18 June 2015 (UTC); Support Suitable for non-perfect and hypernymic mappings. I would still recommend equivalent property (P1628) for perfect mappings though (for instance: genre (P136) → http://mappings.dbpedia.org/index.php/OntologyProperty:Genre). --Hjfocs (talk) 10:40, 19 June 2015 (UTC)

issuing body

Not done

Description	...
Data type	Item
Template parameter	none known
Domain	published works
Allowed values	any of class organization (Q43229)
Example	Commentarii Mathematici Helvetici (Q5152387) → Swiss Mathematical Society (Q684691) Rendiconti del Circolo Matematico di Palermo (Q2535964) → Circolo Matematico di Palermo (Q516426)
Source	specific to the item
Robot and gadget jobs	none so far

Motivation

Most associations produce some kind of periodical publication for their members. Especially those with the goal of elevating public knowledge like historical societies, learned societies or professional associations produce scientific journals or yearbooks intended for a general audience beyond the circle of the actual members. The rôle of the association is usually that of issuance and taking responsibility and varies: Smaller associations select a printer's shop but perform most of the work by volunteer labor (roughly comparable to self-publishing (Q1568650) without a proper publishing company), huge associations like American Mathematical Society (Q465654) probably have publishing departments with full-time staff comparable to any commercial publishing venue.

Interestingly enough the usual German term for this role is "Herausgeber" (a corporate counterpart to editor (P98) here reserved for persons) and the English one is "publisher" (identical to publisher (P123) reserved for commercial publishing firms). Commentarii Mathematici Helvetici (Q5152387) is an internationally known journal, from en:Commentarii Mathematici Helvetici we can learn that it is edited by a person (as editor-in-chief) and publishing (probably the "usual" actions like organizing the printing process, distribution, subscription handling, fee collection, possibly also the infrastructure for the peer reviewing processes) is outsourced to the much bigger European Mathematical Society. However the journal still is most closely connected to the Swiss Mathematical Society as the issuing (corpoarate) body, according to en:WP the Society "owns and operates" the journal (I imagine: holds the rights on the name, implements the board of editors, interacts with the publisher, and most importantly takes financial and moral liability for failure or misconduct, so in theory the Society is also the collector of excess revenue generated in the publishing of the journal ;-).

Many of these periodicals like Rendiconti del Circolo Matematico di Palermo (Q2535964) carry their issuing body akready in the name, but I don't think that named after (P138) would fit the particular role of the association with respect to the publication. Loosely related properties are maintained by (P126) (IIRC introduced for railways and bridges but now broader in scope), printed by (P872) and distributed by (P750) as properties for subprocesses of the publishing chain, standards body (P1462) (quite exactly the role we need but with focus on standards, not publications in general), commissioned by (P88) (rather far-fetched, but in case there is a separate publishing company involved it's clearly the task of the issuing body to commision the work to that company) and the rather unspecific contributor to the creative work or subject (P767) which encompasses "co-creators". A concept of "Corporate authorship" (author (P50) or creator (P170)) IMHO does not apply here, the journals may sometimes contain official communications of the association but the bulk content consists of scholarly or scientific or educational or recreational articles with personal authorship like in any other journal. Gymel (talk) 22:31, 12 June 2015 (UTC)

Discussion

Oppose. publisher (P123) is meant to cover this. "Publisher" in English can be a big organisation or a small group of amateurs. Some professional societies work with a magazine publisher where the society is the editor (P98) (i.e. they select the articles) while the publisher sells the advertising and organises the printing and distribution. I don't see the need for another property. Filceolaire (talk) 21:48, 18 June 2015 (UTC)

Well, editor (P98) is constrained to humans only. But I see that I may have been fooled by the German description of publisher (P123): All other languages formulate it in terms of "responsibility" for the publication, the German one however goes along the lines of "the media company which puts it on the market". But "ordinary publications" are cited by giving place and year of publication and publisher's name: That "publisher" is clearly the publishing house, i.e. the company you or your bookseller would have to contact in order to obtain a copy. So I'm inclined to assumue that the German description got it right and the wording for all other languages does not reflect the intended purpose and acutal usage of the property.

Anyway: The "Commentarii Mathematici Helvetii" example above shows (a human editor-in-chief and) two organizations with distinct roles. Putting the "responsible" Swiss Mathematical Society into the publisher (P123) slot following your suggestion leaves the question where the European Mathematical Society acting as "publishing house" belongs: Also in publisher (P123) but with a distinguishing qualifier for "function"?. -- Gymel (talk) 22:33, 18 June 2015 (UTC)

What about <owned by (P127):Swiss Mathematical Society (Q684691)> since the Swiss own the journal and just subcontract the publishing to the Europeans? Filceolaire (talk) 14:30, 12 July 2015 (UTC)

Indeed ownership (actually possession (Q829909) and property (Q6422240)) usually lie with the society and not with the publishing house. But I fear nobody would understand that the economic "ownership" would be intended to extend to the bibliographic role of creator or corporate editor which consist of organizing, running and being responsible for the editorial process of content selection. I checked the articles on en.wiki for major journals like Nature (Q180445), Science (Q192864), and The Lancet (Q939416) and they are described as owned by their respective publishing companies or scientific societies. The infoboxes always declare a "publisher", in the split case stereotypically as "X on behalf of Y". en:Journal of High Energy Physics would be an example for that, see JHEP Home page for details: "The Journal of High Energy Physics (JHEP) is an international, peer-reviewed, online-only, scientific journal owned by the International School for Advanced Studies (SISSA - Trieste, Italy) and published by Springer".

That means owned by might well work for the English language and two questions remain:

is owned by (P127) for all languages sufficiently broad in meaning (I tend to question that for the German language)
even if, would we prefer to have a distinct poperty for "bibliographic" usage (we are here talking about the subclass "journals" of creative works and aspects of intellectual responsibility, ownership of physical copies or works of art would be in the same domain but something very different). -- Gymel (talk) 10:56, 20 August 2015 (UTC)

Oppose, publisher (P123) works fine. If the act of publishing the work is a collaboration, it seems feasible to list all parties with publisher (P123). As for editor (P98), it should not be limited to people, but instead broadened to allow organizations as well. owned by (P127) is correct to represent the economic relationship, and can be used in addition to publisher (P123) but is not a substitute for it. The following should do the job for the examples:

⟨ Commentarii Mathematici Helvetici (Q5152387)  

 ⟩ publisher (P123) ⟨ Swiss Mathematical Society (Q684691)  

 ⟩

⟨ Rendiconti del Circolo Matematico di Palermo (Q2535964)  

 ⟩ publisher (P123) ⟨ Circolo Matematico di Palermo (Q516426)  

 ⟩

Josh Baumgartner (talk) 07:30, 20 September 2015 (UTC)

Archived Josh Baumgartner (talk) 23:54, 29 October 2015 (UTC)

OSHA liquid class

Withdrawn

Description	classification of flammable liquids
Data type	Item
Domain	liquids
Allowed values	Class IA flammable liquid (Q21009053); Class IB flammable liquid (Q21009055); Class IC flammable liquid (Q21009056); Class II combustible liquid (Q21009057); Class IIIA combustible liquid (Q21009058); Class IIIB combustible liquid (Q21009059);
Example	octane (Q150681) → Class IB flammable liquid (Q21009055)
Format and edit filter validation	The only valid values are the six items listed above that are a part of the OSHA flammable/explosive liquid classification scheme
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This refers to a standard classification scheme used by Occupational Safety and Health Administration (Q746186) to organize liquids by their flash point and/or boiling point. The Pocket Guide to Chemical Hazards includes the flash point and boiling point of a chemical where known, but it also includes the OSHA liquid class. James Hare (NIOSH) (talk) 18:27, 24 September 2015 (UTC)

Discussion

Comment Is this still valid with the introduction of the GHS classification ? Is this not redundant with the NFPA system ? Snipre (talk) 23:12, 27 September 2015 (UTC)

Snipre:

With respect to NFPA: This is indeed the same system used by the NFPA. However, I looked up Wikidata properties with "NFPA" or "National Fire Protection Association" in the name and could not find any properties that corresponded to this classification scheme (IA/IB/IC/II/IIIA/IIIB). So this would still be a novel property for this specific classification scheme.

With respect to the GHS: I did some research and OSHA indeed harmonized their hazard communication system to align with the GHS. The regulation covering flammable liquids in the workplace is 29 CFR 1910.106 which uses the GHS system. However, confusingly, OSHA still has materials on their website using the outmoded system; see e.g. this resource page on flammable liquids. And if you do some Google searching, you'll find that people still use these terms, even if from a regulatory standpoint they're no longer applicable. I absolutely agree we need GHS codes for flammability but from a public health perspective it may be worthwhile to include these other codes as well if they are codes people use and know. James Hare (NIOSH) (talk) 15:35, 1 October 2015 (UTC)

@James Hare (NIOSH): We already have properties for NFPA classification, see NFPA Health (P993), NFPA Fire (P994), NFPA Instability (P995) and based on heuristic (P887). I just want to be sure we don't duplicate the properties. For GHS, I asked the question because I know some organizations plan to switch from their system to the GHS like the SIMDUT classification of Canada. If I mention GHS this is just to be sure we don't create properties for a classification system which will disappears in some years because of the adoption of GHS. If OSHA has its own classification which will be maintained in the next years so we can speak about specific properties. Snipre (talk) 10:01, 5 October 2015 (UTC)

Snipre, those are other NFPA classification schemes, but not this specific one. James Hare (NIOSH) (talk) 14:24, 5 October 2015 (UTC)

Comment Why having properties for this when we know the classes themselves are liquid classes, that we can maek them as member of some classification like

⟨ Class IA flammable liquid (Q21009053)  

 ⟩ instance of (P31) ⟨ OSHA classification class ⟩

(or something more specific) ? Generic classification and metaclasses does not require specific classification properties, and if we want to filter classes on some classification, the SPARQL endpoint allows to do it pretty easily. See Help:Classification for a rationale. author TomT0m / talk page 10:27, 5 October 2015 (UTC)

TomT0m, is your idea to have different substances be considered instance of (P31) the liquid class, rather than having a standalone property? James Hare (NIOSH) (talk) 14:24, 5 October 2015 (UTC)

@James Hare (NIOSH): Exactly. Classification is something very common in Wikidata, we don't need a property for each classification, while with a good use of instance of (P31), subclass of (P279) and classes/metaclasses we can achieve the same thing and use any classification scheme we want. author TomT0m / talk page 14:47, 5 October 2015 (UTC)

I am happy with that then if others are. James Hare (NIOSH) (talk) 15:15, 5 October 2015 (UTC)

I withdraw this proposal; we will go ahead with using instance of (P31). James Hare (NIOSH) (talk) 15:08, 21 October 2015 (UTC) @James Hare (NIOSH), Snipre, TomT0m: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

chemical exposure route

Withdrawn

Description	the process by which one is exposed to a chemical hazard
Data type	Item
Domain	chemicals
Allowed values	Items that describe bodily contact with something; examples include ingestion (Q1663054) and inhalation (Q840343). More items may need to be created to accommodate other forms of exposure such as skin contact.
Example	methyl bromide (Q421758) → inhalation (Q840343) (among others)
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. James Hare (NIOSH) (talk) 22:48, 24 September 2015 (UTC)

Discussion

Oppose Redundant with route of administration (P636). The name of this property is not describing clearly the current proposal but the principle is the same. Snipre (talk) 23:15, 27 September 2015 (UTC)

I understand the logic here, but the name "route of administration" suggests the intended route for a deliberate exposure to a chemical (i.e. a pharmaceutical), which is different from the multiple pathways by which a substance can harm a body. "Route of administration" indicates how to do something; this proposed property indicates "watch out for this", which are two rather different things. James Hare (NIOSH) (talk) 15:42, 1 October 2015 (UTC)

I've been thinking about this and I think it's important to separate route of exposure and route of administration, especially in the context of medications and their manufacture. Just as one example, aspirin is usually administered by mouth, but those working in aspirin manufacturing can be exposed by breathing in aspirin dust. These should be distinguished. Emily Temple-Wood (NIOSH) (talk) 03:09, 8 October 2015 (UTC)

Notified participants of WikiProject Medicine --Tobias1984 (talk) 17:48, 4 October 2015 (UTC)

Support --Tobias1984 (talk) 09:58, 8 October 2015 (UTC)

Withdrawing proposal; route of administration for dedicated chemical exposure appears to work. James Hare (NIOSH) (talk) 14:18, 23 October 2015 (UTC)

@Tobias1984, James Hare (NIOSH): If the use route of administration (P636) is a problem under the current label, we can modify the label in order to enlarge its use to intended and unintended exposure. If you have proposition, just give them. Snipre (talk) 15:17, 23 October 2015 (UTC)

@James Hare (NIOSH), Snipre, Emily Temple-Wood (NIOSH), Tobias1984: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

symptoms from overexposure

Withdrawn

Description	a symptom caused by overexposure to a chemical
Represents	symptom (Q169872)
Data type	Item
Domain	chemicals
Allowed values	any item that describes a medical symptom such as frostbite (Q1350326)
Example	methyl bromide (Q421758) → frostbite (Q1350326); dizziness (Q10916362); et. al.
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. It is very similar to symptoms and signs (P780) but different in that exposure to these chemicals does not inherently lead to these symptoms (as an illness would), but rather, symptoms are caused by exposure at unsafe levels (i.e. above the exposure thresholds set by National Institute for Occupational Safety and Health (Q60346) or Occupational Safety and Health Administration (Q746186)). Exposure at unsafe levels can come from immediate exposure at lethal levels, or from daily exposure over a period of months or years. James Hare (NIOSH) (talk) 22:48, 24 September 2015 (UTC)

Discussion

@James Hare (NIOSH): Have you thought about creating items for all the overexposures. We have arsenic poisoning (Q730249), cobalt poisoning (Q5138720), ... - The advantage would be that we could use the normal symptoms and signs (P780) property. And I have the feeling that there is more interesting data that we could attach to such items. Plus it would keep substances and medical conditions more separate concepts on Wikidata. --Tobias1984 (talk) 23:12, 7 October 2015 (UTC)

@Tobias1984:, that is certainly one technique we could try. Is there any way we could link to the exposure item from the main item, e.g. cobalt (Q740) to cobalt poisoning (Q5138720)? Could I do cobalt (Q740) → has effect (P1542) → cobalt poisoning (Q5138720), and then cobalt poisoning (Q5138720) → has cause (P828) → cobalt (Q740)? In any case I appreciate your help with modeling this data. James Hare (NIOSH) (talk) 00:15, 8 October 2015 (UTC)

Withdrawing proposal; the generic symptom property on dedicated chemical exposure items is working. James Hare (NIOSH) (talk) 14:19, 23 October 2015 (UTC)

@James Hare (NIOSH), Emily Temple-Wood (NIOSH), Tobias1984: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

organ affected by chemical exposure

Withdrawn

Description	body part harmed from unsafe contact with chemicals
Data type	Item
Domain	chemicals
Allowed values	any item describing an organ or organ system, such as nervous system (Q9404) or skin (Q1074).
Example	methyl bromide (Q421758) → human eye (Q430024); skin (Q1074); respiratory system (Q7891); central nervous system (Q47273)
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. James Hare (NIOSH) (talk) 22:48, 24 September 2015 (UTC)

Discussion

@James Hare (NIOSH): I just remembered that we had a similar dicussion a while back and found it in the archive (link). afflicts (P689) should actually be used to show if a disease afflicts a human/horse/cat etc... Adding a qualifier for the organ would be one possibility. But I don't know if a good qualifier has been found yet. Otherwise this proposal would offer a very explicit approach and would be a good idea. - In any case for some statements it is not easy to decide what should qualify what. Maybe separate statements are better? --Tobias1984 (talk) 23:19, 7 October 2015 (UTC)

@Tobias1984:, if we have an existing afflicts (P689) that would allow a qualifier for organs, that would work fine. For what it's worth, I only have affected organ data for humans, but for exposure symptoms, I have data for humans and for "animals" not further specified. Will that create an issue? James Hare (NIOSH) (talk) 00:09, 8 October 2015 (UTC)

Withdrawing proposal; have instead opted for the "afflicts" property for dedicated chemical exposure items. James Hare (NIOSH) (talk) 14:19, 23 October 2015 (UTC)

@James Hare (NIOSH), Tobias1984: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

NIOSH recommended exposure limit

Withdrawn

Description	chemical exposure limit believed to best protect worker safety and health
Represents	Recommended exposure limit (Q7302715)
Data type	Number (not available yet)
Template parameter	en:Template:Chembox Hazards
Domain	chemicals
Allowed values	numbers with a unit of measurement such as gram per litre (Q834105) and a qualifier property "type of exposure limit" (see proposal below) with values carcinogen-related exposure limit (Q21010877), short-term exposure limit (Q7501690), ceiling concentration (Q21010844), or time-weighted average concentration (Q21010836)
Example	ethylene glycol monomethyl ether (Q903362) → 3.0E-7 gram per litre (Q834105) (type of exposure limit: time-weighted average concentration (Q21010836))
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. James Hare (NIOSH) (talk) 23:51, 24 September 2015 (UTC)

Discussion

I am withdrawing this proposal. James Hare (NIOSH) (talk) 18:55, 21 October 2015 (UTC)

@James Hare (NIOSH): Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

OSHA permissible exposure limit

Withdrawn

Description	maximum chemical exposure permitted by U.S. law
Represents	Permissible exposure limit (Q7169331)
Data type	Number (not available yet)
Template parameter	en:Template:Chembox Hazards
Domain	chemicals
Allowed values	numbers with a unit of measurement such as gram per litre (Q834105) and a qualifier property "type of exposure limit" (see proposal below) with values carcinogen-related exposure limit (Q21010877), short-term exposure limit (Q7501690), ceiling concentration (Q21010844), or time-weighted average concentration (Q21010836)
Example	ethylene glycol monomethyl ether (Q903362) → 8.0E-5 gram per litre (Q834105) (type of exposure limit: time-weighted average concentration (Q21010836))
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. James Hare (NIOSH) (talk) 23:51, 24 September 2015 (UTC)

Discussion

Weak oppose There are dozens of organizations providing this kind of limitations (at least one per country). Do we want one unique property and the identity of the source can by deducted from the source data or can we use a qualificer ? These questions should be discussed first. Snipre (talk) 23:05, 27 September 2015 (UTC)

If there is a generic property for regulatory exposure limit amount, with the option to qualify which organization promulgated which regulation, that would work fine. Incidentally, do we want such a property to be strictly legal requirements, or also non-binding recommendations that health organizations may also promulgate? The government agency I work for, NIOSH, issues recommended exposure limits (RELs) that often deviate from the permissible exposure limits (PELs) that OSHA requires. See the proposal above for a NIOSH recommended exposure limit property. James Hare (NIOSH) (talk) 15:48, 1 October 2015 (UTC)

@James Hare (NIOSH): I already started to think about these kind of properties and I had a first draft in my pocket. Please have a look at this table and at the properties

* TLV-TWA (Time-Weighted Average Threshold Limit Value for 8h)

* TLV-STEL (Short-Term Exposure Limit for 15 min)

* TLV-STEL (ceiling limit)

By creating 3 specific properties we can solve the problem of the qualifiers for the application field of each value. Then stay the problem of the mention the organization which fixes that value. So we face 2 solutions one property per organization and 3 qualifiers (8 hours, short duration and ceiling) or 3 specific properties like mentioned above and the mention of the organization as qualifier. For me all organizations are eligible to have their data in WD, we just need a way to differentiate them. Snipre (talk) 09:54, 5 October 2015 (UTC)

This looks reasonable. I have two caveats. First, the time threshold (especially for STELs) may not be the same for each substance; I would have a default time value and then allow for time duration as a qualifier. Also, there are two different types of values: recommended values, which can be suggested by any reasonably knowledgeable organization, and permissible values, which are legally binding requirements. On the WikiProject Chemistry talk page I recommended two separate properties for legally required value vs. recommended the value, with the former having applies to jurisdiction (P1001) as a qualifier and the latter having some other qualifier. How would we fit that into this model: would we create a parallel series of properties for a total of six, or would we find another way to qualify legally binding value vs. recommended value? (Also, threshold limit value should be spelled out.) James Hare (NIOSH) (talk) 14:34, 5 October 2015 (UTC)

I am withdrawing this proposal. James Hare (NIOSH) (talk) 18:56, 21 October 2015 (UTC)

@James Hare (NIOSH), Snipre: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

type of exposure limit

Withdrawn

Description	variant of exposure limit used by the National Institute for Occupational Safety and Health and the Occupational Safety and Health Administration (U.S.) to describe different types of chemical hazard exposure limits
Data type	Item
Domain	as a qualifier to exposure limit-related statements on chemical items
Allowed values	short-term exposure limit (Q7501690), carcinogen-related exposure limit (Q21010877), ceiling concentration (Q21010844), or time-weighted average concentration (Q21010836)
Example	ethylene glycol monomethyl ether (Q903362) → 8.0E-5 gram per litre (Q834105) (type of exposure limit: time-weighted average concentration (Q21010836))
Source	Pocket Guide to Chemical Hazards, published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Part of future mass importation of the Pocket Guide to Chemical Hazards into Wikidata.

Motivation

This is part of a mass importation of data from the Pocket Guide to Chemical Hazards into Wikidata. Note that it is not meant to be used on its own as a statement, but for a qualifier for chemical exposure limit statements. This is because within permissible exposure limits and recommended exposure limits, there are varieties of exposure limits, and different chemicals have different recommendations (ranging from "a little bit during the workday won't hurt" to "avoid if at all possible"). James Hare (NIOSH) (talk) 00:01, 25 September 2015 (UTC)

Discussion

Oppose Better use dedicated properties instead of properties with a lot of qualifiers. Snipre (talk) 01:13, 26 September 2015 (UTC)

Snipre, would this mean dedicated properties for "recommended exposure limit (time-weighted average)", "recommended exposure limit (ceiling concentration)", etc.? Note that a given chemical isn't going to have more than one; it will have a TWA or a ceiling concentration or a STEL or a carcinogen limit (of effectively zero). James Hare (NIOSH) (talk) 15:54, 1 October 2015 (UTC)

@James Hare (NIOSH): Exactly. And no you won't have only one limit because you always have short term and long term exposure limit and depending on the organization some will give you a ceiling concentration instead of a short/long term exposure limit. Snipre (talk) 14:42, 16 October 2015 (UTC)

Right; see my proposal on Wikidata talk:WikiProject Chemistry. If no one objects to that I will proceed with proposing those. James Hare (NIOSH) (talk) 16:33, 16 October 2015 (UTC)

I am withdrawing this proposal in favor of separate properties. James Hare (NIOSH) (talk) 20:26, 21 October 2015 (UTC)

@James Hare (NIOSH), Snipre: Archived -Almondega (talk) 02:27, 30 October 2015 (UTC)

State of matter at STP

Not done

Description	State of a substance at standard temperature and pressure
Represents	phase (Q104837)
Data type	Item
Domain	chemical elements and compounds
Allowed values	One of: solid (Q11438), liquid (Q11435), gas (Q11432)
Example	sodium chloride (Q2314) → solid (Q11438)
Source	Freebase
Robot and gadget jobs	Freebase import

Motivation

Needed to import https://www.freebase.com/chemistry/chemical_compound/phase_at_stp - not to be confused with phase of matter (P515). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:58, 18 June 2015 (UTC)

Discussion

Comment maybe we could use phase of matter (P515) with a qualifier to indicate measurement at STP--Pasleim (talk) 13:12, 22 June 2015 (UTC)
Question Which property use (as qualifier) to say sodium chloride (Q2314) phase of matter (P515) solid (Q11438) at standard temperature and pressure (Q102145) (<== the word in bold) ? Visite fortuitement prolongée (talk) 19:52, 26 June 2015 (UTC)
Oppose I prefer to oppose and to use a most general way using phase of matter (P515) with temperature and pressure as qualifier. All temperature dependent properties will use this pattern. Snipre (talk) 18:53, 22 June 2015 (UTC)
However, phase of matter (P515) at standard temperature and pressure (Q102145) is usefull; even if it is as qualifier. I support allowing standard temperature and pressure (Q102145) as qualifier (in a correct way, of course), since it is a widespread alias for specific temperature and pressure. Visite fortuitement prolongée (talk) 19:15, 22 June 2015 (UTC)
See Wikidata:Property proposal/Pending/2#Temperature (proposed as qualifier) and Wikidata:Property proposal/Pending/2#Pressure (proposed as qualifier). Visite fortuitement prolongée (talk) 19:52, 26 June 2015 (UTC)
Oppose I prefer to oppose and to use a most general way using phase of matter (P515) with temperature (Q11466) and pressure (Q39552) as Wikidata qualifier (Q15720608) Almondega (talk) 18:23, 15 August 2015 (UTC)

Notified participants of WikiProject Chemistry

Not done: use phase of matter (P515) with temperature (Q11466) and pressure (Q39552) as Wikidata qualifier (Q15720608). Archived --Almondega (talk) 02:55, 30 October 2015 (UTC)

electrons per shell

Not done

Description	Number of electrons per shell, of a chemical element. Qualified with the shell number.
Data type	Number (not available yet)
Template parameter	`\|electrons per shell=` in en:Template:Infobox element
Domain	instances of chemical element (Q11344)
Allowed values	integers; max = 18
Example	iron (Q677) → 2, 8, 14, 2 (shells 1-4 respectively)
Source	Freebase
Robot and gadget jobs	Freebase import

Motivation

Qualified with "electron shell number" (see below). Needed to import https://www.freebase.com/chemistry/chemical_element/electrons_per_shell Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:55, 18 June 2015 (UTC)

Discussion

Comment this should either be of datatype string, or if we stay with number, we need a qualifier to indicate which number corresponds to which shell. --Pasleim (talk) 13:07, 22 June 2015 (UTC)

A qualifier is already prosed; see next setion, below. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:20, 30 June 2015 (UTC)

Ah ok, the German translation confused me. --Pasleim (talk) 05:14, 4 July 2015 (UTC)

Oppose Use <has part(s) (P527):electron (Q2225) (quantity (P1114):"8") (applies to part (P518):'shell 2')> Joe Filceolaire (talk) 17:23, 2 August 2015 (UTC)
Oppose I agree with Joe Filceolaire. Almondega (talk) 18:59, 15 August 2015 (UTC)

Notified participants of WikiProject Chemistry

Not done: Use <has part(s) (P527):electron (Q2225) (quantity (P1114):"8") (applies to part (P518):'shell 2')>. Archived --Almondega (talk) 02:55, 30 October 2015 (UTC)

electron shell number

Not done

Description	Number of electron shell, of a chemical element.
Data type	Number (not available yet)
Domain	instances of chemical element (Q11344)
Allowed values	integers; max = 7
Example	iron (Q677) has shells 1-4
Source	Freebase
Robot and gadget jobs	Freebase import

Motivation

Qualifier of "electrons per shell" (see above). Needed to import https://www.freebase.com/chemistry/chemical_element/electrons_per_shell Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:55, 18 June 2015 (UTC)

Discussion

Which purpose has this value?--Kopiersperre (talk) 07:05, 19 June 2015 (UTC)

As I wrote: Qualifier of "electrons per shell" (see above). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:19, 30 June 2015 (UTC)

This should have datatype string or item, as the shell number is just a label of the shell. --Pasleim (talk) 05:12, 4 July 2015 (UTC)

Oppose Use applies to part (P518) with an item for the shell (create items for each shell). Joe Filceolaire (talk) 17:24, 2 August 2015 (UTC)

Notified participants of WikiProject Chemistry

Not done: Use <has part(s) (P527):electron (Q2225) (quantity (P1114):"8") (applies to part (P518):'shell 2')>. Archived --Almondega (talk) 02:55, 30 October 2015 (UTC)

electron configuration

Not done

Description	electron configuration
Represents	electron configuration (Q53859)
Data type	String
Template parameter	`\|electron configuration=`, in en:Template:Infobox element
Domain	Chemical elements
Example	iron (Q677) → [Ar] 3d6 4s2
Source	Freebase
Robot and gadget jobs	Wikidata:Property proposal/Natural science#electrons per shell

Motivation

Needed to import https://www.freebase.com/chemistry/chemical_element/electron_configuration Could be qualified with the item for the corresponding element (argon (Q696) in the example given). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:07, 18 June 2015 (UTC)

Discussion

Andy Mabbett is the example Iron or argon? What does this mysterious string mean? I suspect it can be translated into more useful property statements. Joe Filceolaire (talk) 17:44, 2 August 2015 (UTC)
@Filceolaire: For elements after the first row you always start with the noble gas of the previous row (in this case Argon). And then you count how many electrons you add in each orbital to reach the element (in this case Iron). In this case 6 electrons in the 3d orbital and 2 in the 4s orbital. I guess we could construct the string from a lot of other more semantic statements, but I think saving the string separately is practical and not too much work. That is just one statement more for each element (~118 statements). --Tobias1984 (talk) 21:57, 2 August 2015 (UTC)

So this says that Fe has 6 electrons more than argon in shell 3 and these are in subshell 3d and has 2 electrons in subshell 4s. All this can be expressed by creating an item for each subshell and adding a bunch of statements of the form

has part(s) (P527):electron (Q2225) (quantity (P1114):"2") (applies to part (P518):'shell 1')

has part(s) (P527):electron (Q2225) (quantity (P1114):"8") (applies to part (P518):'shell 2')

has part(s) (P527):electron (Q2225) (quantity (P1114):"2") (applies to part (P518):'subshell 3s')

has part(s) (P527):electron (Q2225) (quantity (P1114):"6") (applies to part (P518):'subshell 3p')

has part(s) (P527):electron (Q2225) (quantity (P1114):"14") (applies to part (P518):'subshell 3d')

has part(s) (P527):electron (Q2225) (quantity (P1114):"2") (applies to part (P518):'subshell 4s')

This gives the same information in a clearer form using our existing properties.

I

Oppose this property. Joe Filceolaire (talk) 23:18, 2 August 2015 (UTC)

I think we can do better. I would prefer

⟨ subject ⟩ has part Search ⟨ subshell 3d ⟩
quantity (P1114) ⟨  14 ⟩

for example. It's a more hierarchical subdivision, the subshell is a part, it's made of electrons, not the opposite. author TomT0m / talk page 07:29, 10 August 2015 (UTC)

TomT0m: the 'C' template is leading you astray. That is why I really dislike it. You have put electron (Q2225) as the item but the item in this example is iron (Q677). I suppose we could write

iron (Q677) <has part(s) (P527):'shell 1'> (has part(s) (P527):electron (Q2225)) (quantity (P1114):'2')

but that looks clumsy to me and looks like the quantity applies to 'shell 1' instead of to 'electron'. That's why I wrote it the other way. Joe Filceolaire (talk) 23:20, 2 September 2015 (UTC)

@Filceolaire: That's not the C template, that's the fact that we use Q number together with the fact that previsualisation is barely usable with the current UI ... This is a tricky question. For example in the RDF partial export, who only exports the main snak if I remember well, the user would get a set of redundant statements "has part" : electron. To me it's better practice to decompose hierarchically an object : an atom has electron shells, the electron shells has electron. This makes the structure more explicit and regular, the fact that an atoms has electrons and that this electrons can be regrouped into shells seems is a less regular scheme that seems more an artefact of the Wikidata data model ... I think a rule of thumb should be that the main snak should by itself add information ... author TomT0m / talk page 08:06, 3 September 2015 (UTC)

TomT0m I agree that we should try and make sure the bare claim (without qualifiers) should give useful info. In this case we should have a claim

<has part(s) (P527):electron (Q2225)> (quantity (P1114):"26") and that should be preferred rank so basic queries only see that and not the other P527 claims. You could have P527 claims for each of the shells as well I suppose, but I think we also need all the separate <has part(s) (P527):electron (Q2225)> claims if we want to codify the distribution of electrons in the shells without having to create separate items for subshell 3d of iron atoms etc. Joe Filceolaire (talk) 22:00, 12 September 2015 (UTC)

Oppose Electron configuration would be much more efficiently handled by having multiple claims such as "has electrons in orbital", which would take in 1-18 (but above 14 only for g orbitals), and be qualified by orbital type (3d etc.).--Jasper Deng (talk) 15:36, 2 October 2015 (UTC)

Notified participants of WikiProject Chemistry

Not done: Use <has part(s) (P527):electron (Q2225) (quantity (P1114):"8") (applies to part (P518):'shell 2')>. Archived --Almondega (talk) 02:55, 30 October 2015 (UTC)

ionization energy

Done: ionization energy (P2260) (Talk and documentation)

Description	minimum amount of energy required to remove an electron from an atom or molecule in the gaseous state
Represents	ionization energy (Q483769)
Data type	Number (not available yet)
Template parameter	en:Template:Infobox element parameter `ionization energy N` [1, 2, 3...]
Domain	chemicals
Allowed values	any integer or floating point number with a measurement unit such as electronvolt (Q83327).
Example	2-heptanone (Q517266) → 9.33 electronvolt (Q83327), series ordinal (P1545): 1
Source	Pocket Guide to Chemical Hazards published by National Institute for Occupational Safety and Health (Q60346)
Robot and gadget jobs	Future importation of Pocket Guide to Chemical Hazards into Wikidata. Check to see if there is a valid unit of measurement?

Motivation

This property is one of the properties included in the Pocket Guide to Chemical Hazards, a U.S. government publication I am working on importing to Wikidata. Although my data only includes the first ionization energy, I think it would make sense to have this property accept multiple values with qualifies for first ionization energy, second ionization energy, etc. James Hare (NIOSH) (talk) 15:59, 24 September 2015 (UTC)

Notified participants of WikiProject Chemistry James Hare (NIOSH) (talk) 16:02, 24 September 2015 (UTC)

Discussion

Support Looking forward to all the pocket-guide data! --Tobias1984 (talk) 18:00, 24 September 2015 (UTC)

Comment There are several ionization energies for each atom. How can we distinguish them ? Qualifiers or several properties ? Snipre (talk) 11:18, 28 September 2015 (UTC)

Qualifiers would make sense; we could use series ordinal (P1545) for this purpose. James Hare (NIOSH) (talk) 23:40, 4 October 2015 (UTC)

@James Hare (NIOSH): Can you create an example item with statments on test.wikidata.org. (The left panel will allow you to create as many items and propties as you need). --Tobias1984 (talk) 12:59, 5 October 2015 (UTC)

I have created a test item using the first, second, and third ionization energies of silicon. James Hare (NIOSH) (talk) 14:16, 5 October 2015 (UTC)

@James Hare (NIOSH): Thanks! That looks like a way to do it. I wonder how it will scale for 100 electrons? @Snipre: What do you think? Create/Wait? --Tobias1984 (talk) 14:31, 5 October 2015 (UTC)

@James Hare (NIOSH), Tobias1984:

Support Snipre (talk) 15:30, 5 October 2015 (UTC)

Notified participants of WikiProject Chemistry

Done ionization energy (P2260). Archived --Almondega (talk) 03:05, 30 October 2015 (UTC)

Average shot length (film)

Done: average shot length (P2208) (Talk and documentation)

Description	cinemetrical measure: en:Shot_(filmmaking)#Duration
Data type	number with units (once available)-invalid datatype (not in Module:i18n/datatype)
Domain	films
Allowed values	duration in seconds or minutes
Example	The Mist (Q695209) → 5.4 seconds; Russian Ark (Q59721) → 96 minutes

Support --- Jura 11:17, 14 August 2015 (UTC)
Support --Jklamo (talk) 19:29, 13 September 2015 (UTC)
Support --Pasleim (talk) 20:06, 4 October 2015 (UTC)
Archived Josh Baumgartner (talk) 17:37, 30 October 2015 (UTC)

Model number part of product name

Not done

Description	model number as included in product name
Data type	String
Domain	products
Allowed values	number possibly with some strings
Example	"iPhone 5C" (Q14883923) → 5C, "Nokia 7280" (Q369565) → 7280

Support series ordinal (P1545) might include this information, but its scope is different. --- Jura 14:20, 24 August 2015 (UTC)
Oppose use software version identifier (P348) with expanded scope. author TomT0m / talk page 15:30, 24 August 2015 (UTC)
- That is tempting, but generally it seems to be used several times on the same item for different versions numbers where as this would be used once per item. --- Jura 05:27, 26 August 2015 (UTC)
  - Still oppose, I don't see either your use case or the example you mention. Is 2000 in "Windows 2000" a numbering part ? It does not make any sense. author TomT0m / talk page 06:33, 26 August 2015 (UTC)
    - It's not meant for software. The software version identifier (P348) is. How do you usually use that? --- Jura 06:35, 26 August 2015 (UTC)
Jura Why have this? Why not just have the product name? Joe Filceolaire (talk) 15:35, 24 August 2015 (UTC)
- The product name should be in the label, but unless the labels are consistent (in every language), sorting on that can be somewhat random. --- Jura 15:50, 24 August 2015 (UTC)
  - I don't think this property would fix that. Mostly you want to sort products on the launch date. Oppose. Joe Filceolaire (talk) 22:54, 25 August 2015 (UTC)
    - No, I don't want that, except if you would add launch date for each item. --- Jura 05:27, 26 August 2015 (UTC)
      - If it deserves a separate item then it deserves a launch date statement. Joe Filceolaire (talk) 18:37, 31 August 2015 (UTC)

Oppose This is an arbitrary breakdown of the official name (P1448) of an item. Is there a source that says "'5c' is the model number portion of the 'iPhone 5c'"? For many products the model number is separate from the name. The iPhone 5c, for example has at least 50 different model numbers associated with it, but none are '5c'. I do think that a general 'manufacturer part number' property is a good idea, but that number may or may not be part of the official name (P1448). If you are looking to sort models within a series, use series ordinal (P1545). Josh Baumgartner (talk) 08:49, 20 September 2015 (UTC)

@Jura1, TomT0m, Filceolaire:

Not done due to lack on consensus.

Archived Josh Baumgartner (talk) 22:35, 30 October 2015 (UTC)

Camera properties from Freebase

Latest comment: 9 years ago97 comments23 people in discussion

Find below a number of properties from http://www.freebase.com/digicams/digital_camera?schema= - the freebase schema for camera properties. I have added this heading for general discussion of these. Joe Filceolaire (talk) 23:41, 2 September 2015 (UTC)

discussion

Do we want wikidata to be a product catalog? Maybe wikidata will become that but I don't think we have the resources to maintain a catalog of every camera model and every other mass produced product as well. Joe Filceolaire (talk) 23:41, 2 September 2015 (UTC)

What is the quality of the Freebase product data for cameras? Was this a hobby project of some guy which will never be checked against anything or is this information good? How can we tell? Joe Filceolaire (talk) 23:41, 2 September 2015 (UTC)

Isn't there a better way to describe cameras using existing properties. (has part(s) (P527)?) How many of these properties can be used on other products? Joe Filceolaire (talk) 23:41, 2 September 2015 (UTC)

Because of the above I lean oppose but I could be convinced. Jura? Joe Filceolaire (talk) 23:41, 2 September 2015 (UTC)

Do you plan to collaborate in the field? If yes, what will you likely be doing? --- Jura 05:05, 3 September 2015 (UTC)

I plan to continue to contribute here on the more strategic issues - like the three I raised above - helping to develop wikidata general policy and properties that support it that rather than properties that only work for one domain. What happens with the Freebase camera data will set a precedent for other product data: lets think it through before we make a decision. Joe Filceolaire (talk) 12:17, 3 September 2015 (UTC)

So you participate in strategic issues such as building the code for Wikibase? Or would you collaborate in building a proof of concept of statements? --- Jura 13:38, 3 September 2015 (UTC)

I'm afraid I'm not a programmer. I have been collaborating on building the data models and the properties to implement them here on Wikidata and will continue to do so while I can. Meet me on my talk page if there is anything I can help you with. Meanwhile can you comment on the three questions above? Joe Filceolaire (talk) 19:59, 3 September 2015 (UTC)

With the properties we create, would you be building a proof of concept (series of items with statements)? --- Jura 04:35, 4 September 2015 (UTC)

I thought that was what you were proposing to do when you proposed these properties - create a series of statements based on info imported from Freebase? Doing some more research I see that en:Comparison of digital SLRs already has a lot of cameras and en:Template:DSLR cameras with movie mode list more and en:Template:Infobox camera already is looking for a lot of this data. To answer my own questions above

It looks like wikipedia is already compiling a digital camera product catalog. Having the info on wikidata will mean infoboxes and comparison tables can be generated from the same data - simplifying maintenance.
There does seem to be enough interest on wikipedia to curate such a catalog and keep it accurate, even if the freebase data were wobbly.
Have you compared the Freebase properties with the 'infobox camera' entries? Is there a match? an overlap?

I am coming around to supporting all these - especially as no one else has any issue with them. I was just a little bothered that adding all these seemed to take us in a new direction that hadn't been discussed. Joe Filceolaire (talk) 02:33, 5 September 2015 (UTC)

Camera ISO

Withdrawn

Description	...
Data type	Item
Domain	cameras
Example	Universe (Q1) → Earth (Q2)
Source	https://www.freebase.com/digicams/camera_iso/cameras
Robot and gadget jobs	Primary Sources Tool

Support Needed for imports from Freebase. --- Jura 10:59, 28 August 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

Camera resolution

Withdrawn

Description	MISSING
Data type	Item
Domain	cameras
Allowed values	items for specific resolutions. Each would have vertical and horizontal resolutions
Example	Universe (Q1) → Earth (Q2)
Source	https://www.freebase.com/digicams/digital_camera/supported_resolutions
Robot and gadget jobs	Primary Sources Tool

Support Needed for imports from Freebase. --- Jura 09:32, 29 August 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

Camera vertical resolution

Withdrawn

Description	Largest possible vertical resolution captured by digital camera in pixels.
Data type	Number (not available yet)
Domain	camera resolutions
Example	Universe (Q1) → Earth (Q2)
Source	https://www.freebase.com/digicams/camera_resolution/vertical_resolution
Robot and gadget jobs	Primary Sources Tool

Support Needed for imports from Freebase. --- Jura 09:32, 29 August 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

Camera horizontal resolution

Withdrawn

Description	Largest possible horizontal resolution captured by digital camera in pixels.
Data type	Number (not available yet)
Domain	camera resolutions
Example	Universe (Q1) → Earth (Q2)
Source	https://www.freebase.com/digicams/camera_resolution/horizontal_resolution
Robot and gadget jobs	Primary Sources Tool

Support Needed for imports from Freebase. --- Jura 09:32, 29 August 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

storage medium

Withdrawn

Description	type of storage medium used (e.g. for digital images in a camera)
Data type	Item
Domain	cameras, possibly other
Allowed values	items for specific storage media
Example	Universe (Q1) → Earth (Q2)
Source	http://www.freebase.com/digicams/camera_storage_type
Robot and gadget jobs	Primary Sources Tool

Support Needed for imports from Freebase. --- Jura 10:46, 31 August 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

camera sensor

Withdrawn

Description	image sensor used for the conversion of light
Data type	Item
Domain	cameras, possibly other
Allowed values	items for specific sensors
Example	Universe (Q1) → Earth (Q2)
Source	http://www.freebase.com/digicams/camera_sensor_type sensor in en:Template:Infobox camera
Robot and gadget jobs	Primary Sources Tool, templatetiger

Support Needed for imports from Freebase, infobox. Might need some fine tuning later. --- Jura 07:32, 1 September 2015 (UTC)
Oppose without valid examples or description. Josh Baumgartner (talk) 22:37, 30 October 2015 (UTC)
Support. Josh Baumgartner see the description at the head of this section. Joe Filceolaire (talk) 04:32, 31 October 2015 (UTC)
Support thanks @Filceolaire:. Josh Baumgartner (talk) 06:12, 31 October 2015 (UTC)

equivalent Italian first name

Withdrawn

Description	related Italian given names
Data type	Item
Domain	first names
Allowed values	Italian first names
Example	John (Q4925477) → Giovanni (Q1158906)
Source	it:Giovanni (nome)#Varianti_in_altre_lingue etc
Robot and gadget jobs	import from itwiki

Support I suppose it could be any language (at least for European names), but itwiki has well structured articles on first names. --- Jura 07:37, 11 October 2015 (UTC)

Oppose A better structure is to introduce a class item for "Given name John and variants", of which both John (Q4925477) and Giovanni (Q1158906) would be instance of (P31) (as well as instances of "male given name"). It would then be easy to query for items that were instances of this class, and then filter for the language of the name being Italian (very easy in SPARQL) -- and would immediately lend itself to any language.

A hierarchical star structure of this kind would be much simpler to maintain and also to query that the structure User:Jura1 proposes (and would not require a multitude of new properties, or indeed any new properties, to implement). Jheald (talk) 13:56, 18 October 2015 (UTC)

Unfortunately, we still haven't seen a suitable reference for that alternate proposal. --- Jura 14:02, 18 October 2015 (UTC)

Oppose per Jheald. Joe Filceolaire (talk) 19:06, 19 October 2015 (UTC)
Oppose per Jheald. -Ash Crow (talk) 11:58, 2 November 2015 (UTC)

Hederich article (1770)

Done: Hederich encyclopedia article (P2272) (Talk and documentation)

Description	entry in Hederich’s encyclopedia of mythology, 3rd edition (1770), via zeno.org
Represents	Gründliches mythologisches Lexicon (Q21139260)
Data type	String
Domain	entities related to Greek, Roman, Egyptian, or Oriental mythology (mythical characters, deities, fictional animals, stones etc.)
Example	Abadir (Q2786349) → Abádir
Source	external reference [1]
Formatter URL	http://www.zeno.org/Hederich-1770/A/$1
Robot and gadget jobs	yes

Motivation

With this property we could start to get a hold of our immense heap of items on entities from Greek mythology. Mix'n'match is a great tool to help with this. Jonathan Groß (talk) 08:48, 18 October 2015 (UTC)

Discussion

Support Marcus Cyron (talk) 13:12, 18 October 2015 (UTC)
Support Magnus Manske (talk) 13:02, 19 October 2015 (UTC)
@Jonathan Groß, Marcus Cyron, Magnus Manske: Done Archived --Almondega (talk) 13:25, 3 November 2015 (UTC)

Heidelberg Academy for Sciences and Humanities member ID

Done: HAdW member ID (P2273) (Talk and documentation)

Description	entry in the list of members of the Heidelberg Academy for Sciences and Humanities
Data type	String
Domain	scientists, scholars, philanthropists
Allowed values	string
Example	Gerhard Ebeling (Q67327) → 1
Format and edit filter validation	1 to 4 digits number
Source	external reference, mix'n'match
Formatter URL	http://www.haw.uni-heidelberg.de/akademie/mitglied.de.html?id=$1
Robot and gadget jobs	yes

Motivation

With the help of Magnus Manske, I already created a matching list for the members of the Heidelberg Academy [2]. By adding this statement to all the items, it would be easier to maintain and qualify the statement member of (P463) -> Heidelberg Academy for Sciences and Humanities (Q833738) Jonathan Groß (talk) 13:26, 12 October 2015 (UTC)

@Tobias1984:: Was hältst du davon? Jonathan Groß (talk) 13:55, 12 October 2015 (UTC)

Discussion

Support Looks like a good addition. --Tobias1984 (talk) 14:10, 12 October 2015 (UTC)

Support --Olaf Kosinsky (talk) 10:04, 14 October 2015 (UTC)

Comment There are formatter URLs for English as well (http://www.haw.uni-heidelberg.de/akademie/member.en.html?id=$1). Perhaps we could incorporate those in some way? Jonathan Groß (talk) 13:49, 19 October 2015 (UTC)

Support Seems stable and straight-forward. I would recommend the label changed to Heidelberg Academy for Sciences and Humanities member ID to be more in line with other similar properties. Josh Baumgartner (talk) 20:18, 22 October 2015 (UTC)

Agree. That leads to a rather clumsy German version ("Identifikator in der Mitgliederliste der Heidelberger Akademie der Wissenschaften"), but at least this is accurate. The IDs are just arbitrary, BTW, but they are stable. Jonathan Groß (talk) 07:46, 23 October 2015 (UTC)

Done --Tobias1984 (talk) 17:37, 3 November 2015 (UTC)

World Health Organisation International Nonproprietary Name

Done: World Health Organisation international non-proprietary name (P2275) (Talk and documentation)

Description	International name for
Data type	monolingual-invalid datatype (not in Module:i18n/datatype)
Template parameter	international nonproprietary name (Q824258)
Domain	drug
Allowed values	String
Example	DL-amphetamine (Q179452): amfetamine, amfetaminum, amfétamine, anfetamina, амфетамин, أمفيتامين, 苯丙胺
Source	http://www.who.int/medicines/services/inn/en/

Notified participants of WikiProject Chemistry, @Almondega, Innocent bystander, Filceolaire:

The property already exists but has the wrong datatype (see P1805 (P1805)): currently a string property is used but the name is provided in at least 4 languages (English, French, Spanish and Latin) and can be translated in Russian, Arabic, Chinese as well. So a monolingual datatype is required. No need of multilingual datatype because the WHO only translates in 7 seven languages and no extension for translation is expected.

The proposal is done in parallel at the deletion request of the current property: Wikidata:Properties_for_deletion#World_Health_Organisation_International_Nonproprietary_Name_.28P1805.29. Snipre (talk) 12:27, 3 November 2015 (UTC)

Support --Almondega (talk) 12:35, 3 November 2015 (UTC)

Stöder

Question Is there a posibility to add a filter/constraint that prevents the use of others than these 7 languages? -- Innocent bystander (talk) 12:37, 3 November 2015 (UTC)

Currently, no. Snipre (talk) 11:04, 4 November 2015 (UTC)

Support. Joe Filceolaire (talk) 23:23, 3 November 2015 (UTC)

Tournament victories

Not done

Description	Tournament victories of a sportsperson
Data type	Item
Domain	tennis players and others
Example	Serena Williams (Q11459) => "WTA Tour (Q2537906)" (=>P1131 (P1131): "64")
Source	ATP/WTA/ITF or Wikipedia articles
Robot and gadget jobs	Not now.
Proposed by	Stryn (talk)

Discussion

Oppose. This is too specific. Can we make it more general so it can be used by sports people participating in the Olympics and the Super bowl and various leagues as well? We can even combine P1130 (P1130) and P1131 (P1131) into one generic property too, that can be used for other sports with qualifier "sport:mens singles tennis" or "sport:womens doubles tennis" etc. when it refers to tennis (if there are separate mens and womens competitions then these are separate sports). Filceolaire (talk) 22:55, 20 March 2015 (UTC)
Oppose. Should be deduced from the data. Update of this kind of properties is a nightmare. Snipre (talk) 12:27, 22 May 2015 (UTC)
  Oppose Use:
⟨ Serena Williams (Q11459)      ⟩ number of wins (P1355) ⟨  64 ⟩
part of the series (P179) ⟨ WTA Tour (Q2537906)      ⟩
Josh Baumgartner (talk) 04:33, 16 September 2015 (UTC)
@Stryn, Filceolaire, Snipre:

Not done due to lack of consensus.

Archived Josh Baumgartner (talk) 20:59, 6 November 2015 (UTC)

Infobox property

Not done

Description	The Wikidata properties which feed into Wikipedia (or sister project) templates
Data type	Property
Domain	Wikimedia project templates
Allowed values	Pn
Example	Template:Infobox architect (Q10973090) => date of birth (P569)
Source	Wikimedia project template documentation
Proposed by	Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits

Discussion

Needed so that we can automate template translation and documentation. Values should have qualifiers (e.g. applies in English Wikipedia (Q328)). May be useful to feed (or replace) TemplateData in Visual Editor. May be read from existing TeplateData. This mailing list conversation refers. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:17, 5 March 2015 (UTC)

Support Useful to know which properties are needed by which templates (and vice-versa, which templates use which properties). What would you use for "applies in" -- stated in (P248) ? Or is there something better ? Jheald (talk) 23:09, 5 March 2015 (UTC)
Support it's similar to Wikidata:Property_proposal/Generic#property_used_in_this_template. --- Jura 12:44, 6 March 2015 (UTC)
Comment how would this be used? How do I know what field in the template corresponds to what property? I doubt this will work and scale. How do you deal with properties being used in language a, but not language b? This should be part of a bigger plan/design. Multichill (talk) 19:30, 8 March 2015 (UTC)
- As ordering statements on items doesn't work well, using P:P1545 to define the (standard) order of display can be used. Obviously, projects can still decide not to use the information given by this property at all. --- Jura 11:34, 9 March 2015 (UTC)
- The language issue is addressed in "Values should have qualifiers (e.g. applies in English Wikipedia (Q328))". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:09, 13 March 2015 (UTC)
Support but needs to be put on hold until we have the "property" datatype which I believe the team is working on. Filceolaire (talk) 23:26, 20 March 2015 (UTC)
@Filceolaire: property data type is already available since 3 months... --Pasleim (talk) 21:02, 24 March 2015 (UTC)
Oppose {{ExternalUse}} works well. Why to change it? --Pasleim (talk) 21:02, 24 March 2015 (UTC)
"Wikipedia User: 'Yeah, templates work fine, why use Wikidata?' ". --- Jura 08:07, 25 March 2015 (UTC)

How does {{ExternalUse}} help us to build template documentation in Wikipedias? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 09:51, 26 March 2015 (UTC)
I'm really not understanding how one uses this property based solely on the provided example. Can I get an explanation? --Izno (talk) 02:36, 28 March 2015 (UTC)
- @Izno: (Sorry for the late reply, I've been travelling) At for example, en:Template:Infobox_person#TemplateData, there are a list of infobox properties used in the infobox which is also known as Template:Infobox person (Q6249834). Many of these have a direct, 1:1 correspondence with the properties used in Wikidata (indeed, the infobox may transclude the corresponding data from Wikidata). The proposed property would be used to list the former properties. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:08, 15 April 2015 (UTC)
  - So, we want the developers to build something which tracks properties at a template-use level. That use seems to be more "useful" per a large number of comments here and above for templates/module tracking of certain properties (and more scalable, and automatically updates, etc. etc.). So in that use I would certainly oppose. This property however would track the use of a property matched to a particular parameter of the template? I think that that property might be interesting. Is that the property you really mean this to be? I think that makes sense. My concern is that (not normally, but sometimes) parameter names change. Thoughts? Also per Multichill, how would you show that a property is used in only one in a particular version of the template? What about modules, where the reference is not immediately obvious? --Izno (talk) 20:04, 15 April 2015 (UTC)
    - This proposal requires nothing of the developers (but of course they can make use of it as they see fit). As noted above (twice!), "Values should have qualifiers (e.g. applies in English Wikipedia (Q328))". The names of parameters are irrelevant; whether something is called, say, "date of birth", "birth date", "DoB" or some other variant has no effect. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:01, 15 April 2015 (UTC)
      - Then I read this as the same request as the now-recently rejected Wikidata:Property_proposal/Archive/29#property_used_in_this_template. Are you sure that's what this property is going to be? Amir's confusion is my own, and I want to be clear about what it is you are proposing here because I don't want to oppose data which might be useful and which wouldn't end up being subsumed by developer investment into automated tracking of properties in templates or modules. --Izno (talk) 18:08, 16 April 2015 (UTC)
        That was not rejected, it was withdrawn, apparently in the light of this proposal. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:33, 25 April 2015 (UTC)
Comment I'm also not sure that I understand. Let's say that Template:Infobox architect has a "date of birth" parameter in the English Wikipedia and a corresponding "nascut" property in the Catalan Wikipedia. Where will the names of the parameters be stored? Or am I completely misunderstanding the purpose? --Amir E. Aharoni (talk) 21:15, 15 April 2015 (UTC)
- If needed, that could be a monolingual-text qualifier of the property. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:33, 25 April 2015 (UTC)
  - Except that that would be different for every language, and Wikidata items for templates are multilanguage... --Yair rand (talk) 12:52, 28 April 2015 (UTC)
Oppose hackish, short-sighted and Wikimedia-centric workaround. --Ricordi samoa 05:20, 30 April 2015 (UTC)
- We have lots of "Wikimedia-centric" properties; see Wikidata:Property proposal/Property metadata. There's nothing wrong with having properties that help us to keep the Wikidata and its sister projects working, or to document how they work. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:23, 27 May 2015 (UTC)
Oppose per Ricordisamoa. Snipre (talk) 12:30, 22 May 2015 (UTC)
@Pigsonthewing, jheald, Jura1, multichill, Filceolaire, Izno:

Not done due to lack of consensus.

Archived Josh Baumgartner (talk) 21:05, 6 November 2015 (UTC)

uses

Done: uses (P2283) (Talk and documentation)

Description	subject makes use of object
Data type	Item
Domain	actions?
Allowed values	objects?
Example	punch (Q1973927) => fist (Q16743864), walking (Q6537379) => leg (Q133105), defenestration (Q220277) => window (Q35473)?
Proposed by	Popcorndude (talk)

Discussion

I was trying to define punch (Q1973927), ("A hit or strike with one's fist." Wiktionary) so I added subclass of (P279) => strike (Q3242199), but couldn't find any way of expressing the "with one's fist" part, and I imagine such a property would be useful for defining other actions. (If the third example above is included, this be better termed requires rather than uses.) Popcorndude (talk) 21:22, 7 February 2015 (UTC)

has use (P366) is almost an inverse property here. --Izno (talk) 02:38, 28 March 2015 (UTC)
Question Could we broaden item operated (P121) to "item used" to cover this? Josh Baumgartner (talk) 16:09, 11 June 2015 (UTC)

Well, it does already have a constraint violation rate above 90%, having been used to list tanks and guns used by Angolan Army (Q17597972), among other things. Though this seems like it would end up a little too broad. "Equipment used/required by an organization, group, or action." I would say broaden item operated (P121), but still have this be separate. Popcorndude (talk) 21:01, 11 June 2015 (UTC)

@Popcorndude: Excellent point, item operated (P121) should probably have its constraints reviewed to reflect current use, but still remain limited to equipment used by organizations. I can get behind this proposal, but would like it better labeled and described to reflect its scope (perhaps: 'executed with', 'physical object required to complete the subject action'). Josh Baumgartner (talk) 17:05, 12 June 2015 (UTC)

How about "action requires"? Popcorndude (talk) 18:20, 12 June 2015 (UTC)

No problem. Josh Baumgartner (talk) 23:44, 18 June 2015 (UTC)

Done Archived Josh Baumgartner (talk) 21:23, 6 November 2015 (UTC)

pattern

Done: format as a regular expression (P1793) (Talk and documentation)

Description	Property representing the pattern statement from current constraint usage (see above for detailed explanation).
Data type	String
Allowed values	Regexes
Example	pattern ==> (tt\|nm\|ch\|co\|ev)\d{7}
Robot and gadget jobs	Migrate constraints from property talk pages.
Proposed by	Ivan A. Krestinin, Jonas.keutel (talk)

Please support or comment, but do not implement until we are ready to migrate. Jonas.keutel (talk) 12:59, 7 March 2015 (UTC)

Discussion

Comment in the meantime, there is format as a regular expression (P1793). --- Jura 06:19, 23 March 2015 (UTC)
Archiving --- Jura 02:09, 7 November 2015 (UTC)