Wikidata:Property proposal/salinity

salinity

Originally proposed at Wikidata:Property proposal/Place

   Done: salinity (P4350) (Talk and documentation)

Description	dissolved salt content of a body of water
Represents	salinity (Q179615)
Data type	Quantity
Domain	body of water
Allowed values	0-1
Allowed units	No
Example	Sea of Marmara (Q35367) → 0.0022
Source	Various
Robot and gadget jobs	Unclear (Mathematica have this kind of information, but it's not free)

Motivation

GZWDer (talk) 16:17, 25 June 2017 (UTC)

Discussion

Are you using the definition at the top of en:Salinity? If so then I understand your lack of units (although I would prefer users to be able to specify this in different units if they wish), but then your range of 0-1 is off. That definition has a potential theoretical range of 0-1000, and the value for en:Marmara should be either 22 or 38 depending if you mean the surface or the average at depth. --99of9 (talk) 02:09, 26 June 2017 (UTC)

•  Support ChristianKl (talk) 22:47, 26 June 2017 (UTC)
• I just saw the comment about the lack of units. I don't think that's a good idea. ChristianKl (talk) 16:47, 27 June 2017 (UTC)
•  Oppose. See also Wikidata:Property_proposal/Archive/49#salinity. Use has part(s) (P527) table salt (Q11254) qualifier proportion (P1107). --Yair rand (talk) 20:13, 6 July 2017 (UTC)
  • @Yair rand: Currently, table salt (Q11254) is not about pure NaCl but about table salt that's mostly NaCl. I think it would be bad to use it in this context. Given that this naive solution is problematic, there's additional reason for having a specialized property. Having the number in a specialized property makes it easier to query the data (both in terms of writing code and in terms of database retrival times). ChristianKl (talk) 09:59, 23 August 2017 (UTC)
    • The label of that item was changed since I wrote the comment. It used to be just "salt". Since the scope has been changed, a different item should be used (probably sodium chloride (Q2314)). --Yair rand (talk) 06:27, 24 August 2017 (UTC)
•  Oppose per Yair rand & previous discussion Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:42, 8 July 2017 (UTC)
•  Support Matěj Suchánek (talk) 18:26, 8 August 2017 (UTC)
•  Support seems overly complicated to me to use "has part" for such a commonly reported property --Ita140188 (talk) 08:22, 14 August 2017 (UTC)

  I would think that creating a whole new property would be the complicated way, instead of just using a generic "this thing is made of that other thing, this much" property, which would allow more general comparisons of physical makeup without requiring loads of extra code everywhere. --Yair rand (talk) 06:27, 24 August 2017 (UTC)

•  Support Salinity may change over time, it makes more sense to have separate property for that. Yarl 💭  17:52, 23 August 2017 (UTC)

  Time qualifiers can be used for P527 just as well as for a new property. I don't see how salinity changing is an argument in favor of creating a new property. --Yair rand (talk) 06:27, 24 August 2017 (UTC)

•  Support "has part" should be in this case be reserved for parts of seas and subregions of lakes, not for salt. There is a zillion things that can be "part of" sea and that would just lead to chaos. Salinity is a well known property in hydrology. --Vojtěch Dostál (talk) 17:46, 12 September 2017 (UTC)
• @Thryduulf, Tobias1984, Snipre: you all commented on the previous proposal, but not here - any further thoughts? @GZWDer: there is an unanswered question above from 99of9, can you clarify the definition here? ArthurPSmith (talk) 19:34, 13 September 2017 (UTC)
•  Comment Salinity is a common measurement and takes qualifiers of its own (measurement depth, coordinates, measurment equipment, ...). So naturally a dedicated property seems fitting. But it is difficult to say where to draw the line. All elements would require around 100 properties. For chemicals the number of properties are infinite. Then again we can't rely on casual users to understand our property-qualifier grammar to the fullest. Wouldn't it be great if we could create shadow-properties. So if someone adds salinity into the UI it automatically converts it to the has-part + qualifiers representation. The UI could even show both. Can't really make up my mind about this property. The complexity of the topic in my opinion goes way beyond a typical property proposal. --Tobias1984 (talk) 17:32, 14 September 2017 (UTC)
•  Support I was not convinced by the arguments in opposition in the last proposal, and I'm still not convinced now. I would prefer a property that allowed units though. Thryduulf (talk) 21:06, 17 September 2017 (UTC)
•  Comment For made from material (P186) I experienced the need for a qualifier property which tells about the percentage of concentration of a solution. Could not this be applicable for this special case too? --Scoid d (talk) 15:06, 18 September 2017 (UTC)
•  Support I've marked this as ready. Thierry Caro (talk) 16:21, 11 October 2017 (UTC)
• @GZWDer, Thierry Caro, Thryduulf, Tobias1984, Snipre, ArthurPSmith:  Done Created as salinity (P4350). ChristianKl (talk) 11:24, 12 October 2017 (UTC)
• @99of9, Yair rand, Pigsonthewing, Vojtěch Dostál, Scoid d: some of you were missed in ChristianKl's ping. I also wish we had settled the units issue here - is salinity really a dimensionless quantity? Is it well defined? Do we need usage instructions to clarify? ArthurPSmith (talk) 13:34, 12 October 2017 (UTC)
  • I'm surprised to see ChristianKl create a disputed property, when he had already not only expressed support, but argued against those disagreeing with him. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:27, 12 October 2017 (UTC)