Wikidata talk:WikiProject Mineralogy/Properties/Archive/2014/05

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

Mohs hardness Property:P1088

For many minerals is reported as Mohs hardness a range, for example mindat for ferro-actinolite says 5-6. I think it is expressed in this manner because hardness is half way between 5 and 6 (maybe 5½). Mohs hardness is empirical, only comparative. How can we express it in wikidata property Property:P1088? Using two values is wrong; using the lower value is correct but we lost some information (sorry for my english). Suggestions? --Sbisolo (talk) 14:16, 2 May 2014 (UTC)

Minerals are natural products, their properties are within a range. Higher than 5 and lower than 6, would be right. I didn't know that we have this problem. --Chris.urs-o (talk) 14:59, 2 May 2014 (UTC)

Copy n paste from Wikidata:Project chat: "You can enter the value 5.5±0.5". Snipre (talk) 11:24, 5 May 2014 (UTC)

Unique value for Nickel-Strunz 9th edition (updated 2009) (P712) and Nickel-Strunz '10th ed', review of (9th ed/ 2009 update) (P713)

In my opinion unique value constraint for Nickel-Strunz 9th edition (updated 2009) (P712) and Nickel-Strunz '10th ed', review of (9th ed/ 2009 update) (P713) should be removed because is not an error condition. Maybe can be added a format constraint. --Sbisolo (talk) 08:47, 7 May 2014 (UTC)

You are probably right, tourmalines, spinels (oxides), garnets have the same ids, solid solution series and homologous series as well. --Chris.urs-o (talk) 11:46, 7 May 2014 (UTC)

Still on Strunz classification: for Strunz 8th edition (series ID, updated) (P711) mineralogical class is written in some case with roman numerals, in other using arabic numerals. I don't know which is the correct one but I think we have to establish to use only one (maybe adding a constraint). Where I can request these modifications to constraints? --Sbisolo (talk) 07:47, 9 May 2014 (UTC)

de.wikipedia uses roman numerals, MinDat uses arabic numerals. I think that roman numerals is correct. User:Akkakk and User:Ivan A. Krestinin understand the language, Wikidata:Bot requests might help you. --Chris.urs-o (talk) 09:24, 19 May 2014 (UTC)

Note: Ra'ike (de.wikipedia) says that the original work (8 ed) doesn't have the last two numbers too. So Strunz 8 ed wouldn't have unique values too. I should check it at a library. --Chris.urs-o (talk) 13:04, 23 May 2014 (UTC)

color (P462) and streak color (P534)

I'm not sure how using color (P462) and streak color (P534), for example for fluoro-cannilloite (Q3746853) in mindat color is "grey-green" and streak is "grey-white" and in A new hyper-calcic amphibole with Ca at the A site: Fluor-cannilloite from Pargas, Finland (Q16915730) color is "grayish green" and streak is "grayish white". How to map them to existing colors? --Sbisolo (talk) 10:36, 28 May 2014 (UTC)

How about: item grey and qualifier adjective & green color and white streak? Is it possible? Items are all nouns, right? Adding two colors is easier (MinDat). --Chris.urs-o (talk) 17:37, 28 May 2014 (UTC)

Another case: sometime color is expressed as a range (for example "emerald green to pale green"). --Sbisolo (talk) 08:55, 29 May 2014 (UTC)

Natural products have a range. Wikidata doesn't express ranges very well (yet?). Colors: emerald (Q2465494) (green), green (Q3133) and gray (Q42519). --Chris.urs-o (talk) 13:14, 29 May 2014 (UTC)