Wikidata:Property proposal/display size

display size edit

Originally proposed at Wikidata:Property proposal/Generic

Not done

Description	diagonal lenght of the display in an electronic device.
Represents	display size (Q861066)
Data type	Number (not available yet)
Template parameter	display
Domain	computer (Q68)
Allowed values	number
Allowed units	inch, centimeter
Example	Nexus 4 (Q49047) → 4.7 inches
Robot and gadget jobs	yes
See also	Q12538706

Motivation

I believe it is obvious why, where and how this property is needed and will be used Ogoorcs (talk) 01:54, 29 January 2017 (UTC)[reply]

Discussion

Support ChristianKl (talk) 10:40, 30 January 2017 (UTC)[reply]
Oppose I suggest using has part(s) (P527) display device (Q327065) (or a subclass of (P279)) with qualifying properties of either width (P2049) and height (P2048) in combination, or length (P2043), determination method (P459) (measurement method) and aspect ratio (W:H) (P2061) in combination. If an item such as "diagonal screen measurement" doesn't exist already, this would need to be created for use with determination method (P459). This suggested approach would far more accurately model displays on computers, televisions, mobile phones, etc and would also be far more flexible in allowing multiple methods of measurement to be used. There would however still be an assumption about how width/height/aspect ratio are measured for each device (i.e. how is the device rotated and which side of a rectangle is considered "width" versus "height"). Dhx1 (talk) 14:24, 8 February 2017 (UTC)[reply]

I see your point here, but I can't see why new properties proposals that can already be expressed trough a combination of "another property + qualifiers" couldn't have their specific equivalent property, particularly considering the benefits in user experience that simple properties provide (also against damages caused by newcomers trying to fill the gaps improvising properties failing to check it in similar entities).

I think that if your priority here is to preserve some kind of uniformity and to reduce complexity, opposing simple and catchy properties and so forcing users to think as machines really doesn't help this project; on the contrary, I think you should advocate for an equivalent to property meta-property: a way such that newcomers can shorcut an absurdely long "consist of: screen -> {determination method: diagonal measurement, lenght: value}" with "diagonal length: value" (equivalence classes for properties are needed for queries too: a simple query such as "dutch rappers" need a UNION operator for elements having properties "genre: rapping" and "instance of: rapper").

Anyway to me your argument is totally invalid, since you can apply it to proof that almost any property, for example birthplace, can be made useless having the right amount of information:

Julius Caesar -> birthplace -> Rome

is equivalent to

Julius Caesar -> partecipant of -> birth -> { of -> Aurelia Cotta

date: 100 BCE

place: Rome }

Ogoorcs (talk) 05:09, 12 February 2017 (UTC)[reply]

Oppose per Dhx1. ArthurPSmith (talk) 16:56, 8 February 2017 (UTC)[reply]
Oppose per Dhx1. Not buying the Julius Cesar slippery slope argument. An alternative, maybe even better, way to express the statement with qualifiers would be: Nexus 4 (Q49047)length (P2043)4.7 inches with applies to part (P518) = display device (Q327065) and criterion used (P1013) = diagonal (Q189791). --Swpb (talk) 17:06, 15 February 2017 (UTC)[reply]
Oppose per Dhx1. Andy Mabbett (ArthurPSmith); Talk to Andy; Andy's edits 21:49, 18 March 2017 (UTC)[reply]
@Ogoorcs, Dhx1, ArthurPSmith, ArthurPSmith: Not done, follow the process layed out by Swpb. ChristianKl (talk) 07:44, 29 March 2017 (UTC)[reply]