Wikidata talk:WikiProject Mineralogy/Properties/Archive/2013/04
|
This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion. |
Strunz classification
In my humble opinion, could be useful adding Strunz classification. The problem is that it depends from edition used. Maybe we can add simply "Strunz classification edition 8", "Nickel-Strunz classification edition 9" and so on or there's a better way? (sorry for my English) --Sbisolo (talk) 15:36, 16 April 2013 (UTC)
- German Wiki uses Strunz_8 and Strunz_9 as separate fields. I think that is probably a good thing to copy.--Tobias1984 (talk) 18:10, 16 April 2013 (UTC)
- Nickel–Strunz Identity doesn't work very well.
- We can use Strunz 8 ed (1982), Nickel–Strunz 9 ed (2001) and Nickel–Strunz updated by mindat.org ("10 ed, pending publication").
- We need mineral groups by Fleischer’s Glossary of Mineral Species (10 ed, 2008) after rruff.info/ima/.
- We need structural groups by rruff.info/ima/ (Univ. of Arizona), that would be helpful too.
- We can exclude Strunz 8 ed, we still have 4 infobox lines (separate fields) already. Sometimes 6 lines (the Glossary and rruff.info have a mineral systematic)
- Note, we have: tschermakite group → Calcium amphibole subgroup → w(OH, F, Cl)-dominant amphibole group → amphiboles (amphibole supergroup) → inosilicates
- Have a look at en:Wikipedia:WikiProject Rocks and minerals/Worklist, en:Template:Strunz and Commons:Template:Mineralnavigation
- Tobias1984 and I are able to read German in an emergency.
- Cheers --Chris.urs-o (talk) 18:28, 16 April 2013 (UTC)
- BTW; use the chemical formula on rruff.info/ima/ and CNMNC Master list 2012. Thx --Chris.urs-o (talk) 18:30, 16 April 2013 (UTC)
- Thank you for you input Chris. I think we can gather the information from Strunz-8 anyway. Some Wikis use it in the infobox and those that don't, don't have to pull that data from Wikidata. What is you opinion about numerical values like density or melting point. Is there a reference that has a consistent data set? If it is a range should we aquire it in two fields? What about standard deviation for certain measurement like the melting point? --Tobias1984 (talk) 20:51, 16 April 2013 (UTC)
- The minerals are impure by nature. A data range with few significant numbers is better than a number and its standard deviation. Mineralienatlas, mindat.org and Handbook of Mineralogy are the references; so a data with its reference. I might be wrong, but I think that we have the melting point only for some pure metals and oxides. --Chris.urs-o (talk) 04:23, 17 April 2013 (UTC)
- I found melting points also for some organic compound. There's an "official" grouping defined by IMA-CNMNC: for example, just published garnet supergroup definition (see http://pubsites.uws.edu.au/ima-cnmnc/) --Sbisolo (talk)
- Univ. Arizona - rruff.info/ima/ follows IMA --Chris.urs-o (talk) 09:21, 17 April 2013 (UTC)
- I found melting points also for some organic compound. There's an "official" grouping defined by IMA-CNMNC: for example, just published garnet supergroup definition (see http://pubsites.uws.edu.au/ima-cnmnc/) --Sbisolo (talk)
- The minerals are impure by nature. A data range with few significant numbers is better than a number and its standard deviation. Mineralienatlas, mindat.org and Handbook of Mineralogy are the references; so a data with its reference. I might be wrong, but I think that we have the melting point only for some pure metals and oxides. --Chris.urs-o (talk) 04:23, 17 April 2013 (UTC)
- Thank you for you input Chris. I think we can gather the information from Strunz-8 anyway. Some Wikis use it in the infobox and those that don't, don't have to pull that data from Wikidata. What is you opinion about numerical values like density or melting point. Is there a reference that has a consistent data set? If it is a range should we aquire it in two fields? What about standard deviation for certain measurement like the melting point? --Tobias1984 (talk) 20:51, 16 April 2013 (UTC)
I think the way we need to add the information to increase the query flexibility is this way: --Tobias1984 (talk) 09:08, 17 April 2013 (UTC)
| Property: Nickel_Strunz_9 | Carbonates and nitrates, halides and borates | Carbonates | Anhydrous carbonates | Calcite |
|---|---|---|---|---|
| 1 (String) | 05 | 05 | 05 | 05 |
| 2 (String) | - | A–E | AB | AB |
| 3 (String) | - | - | - | 05 |
| Output (Query) | 05 | 05.A-E | 05.AB | 05.AB.05 |
Discovered by
Is not so common finding who discovered a new mineral species, specifically for newer species. Maybe we can use as discoverers authors of paper that reports new species? --Sbisolo (talk) 13:26, 17 April 2013 (UTC)
- For really long-known minerals (quarz, feldspar) there might not even be a discoverer known. I filled in for Quartz that the discoverer is unknown (little gear symbol). But in all other cases I think your suggestion is excellent. --Tobias1984 (talk) 14:01, 17 April 2013 (UTC)
- Difficult, names need at least first name, first letter of middle name, family name, birth year and death year to be specific. --Chris.urs-o (talk) 05:04, 18 April 2013 (UTC)
- I did some searching and Perovskite is a good example. Both the discoverer and the namesake have Wikipedia articles (therefore items on Wikidata). So we don't have to worry about giving more information because it is included in the respective items. For less noteworthy people the field is eventually going to accept strings too, but that is still under discussion. I can also not add the proper references yet, because those, too, are still on discussion on how they should be handled. --Tobias1984 (talk) 06:28, 18 April 2013 (UTC)
- Difficult, names need at least first name, first letter of middle name, family name, birth year and death year to be specific. --Chris.urs-o (talk) 05:04, 18 April 2013 (UTC)
Request for comments
One of the properties we need is currently being proposed. Please leave your comments at Proposal:Term:Color (item). A property needs about 5 supporters and one week of review time until it can be created. --Tobias1984 (talk) 07:50, 18 April 2013 (UTC)
- I've been having a discussion on the German MinProject about the color property, where people were skeptical of the applicability of this property. It looks like we have to brain about that one some more. --Tobias1984 (talk) 20:13, 18 April 2013 (UTC)
- See also: Wikidata:Property_proposal/Term#Mineralogy 23PowerZ (talk) 20:15, 18 April 2013 (UTC)
- Hi 23PowerZ! Thank you for helping. Could you maybe give a few examples how you would use streakcolor as an item? Would it work like this?
- What would you do about streak colors that don't have an item like "yellowish brown" (as a bad example). --Tobias1984 (talk) 20:26, 18 April 2013 (UTC)
- Seems I haven't thought that out. Could be changed to string instead, but here's not the place to discuss it. 23PowerZ (talk) 20:30, 18 April 2013 (UTC)
- See also: Wikidata:Property_proposal/Term#Mineralogy 23PowerZ (talk) 20:15, 18 April 2013 (UTC)
Varieties
What are we going to do with the varieties (emerald, aquamarine, ruby and others)? --Chris.urs-o (talk) 18:41, 18 April 2013 (UTC)
- That's also a hard question. We could make a property (e.g. ruby: variety of = corundum). --Tobias1984 (talk) 06:03, 19 April 2013 (UTC)
- Q43513 (emerald) shouldn't be a subclass of Q7946 (mineral) but a subclass of Q103480 (beryl). --Chris.urs-o (talk) 07:35, 19 April 2013 (UTC)
- Your right. Best thing is to "subclass of" all varieties to their respective minerals. --Tobias1984 (talk) 08:23, 19 April 2013 (UTC)
- I saw that u made Q104692 (actinolite) an "instance of" Q7946 (mineral) and a "subclass of" Q17159 (amphibole). Q50769 (pyrite) as well. Q17159 (amphibole) and Q192880 (pyroxene) a "subclass of" Q279440 (inosilicates), a "subclass of" Q178977 (silicate minerals), a "subclass of" Q7946 (mineral) is tempting. What now? --Chris.urs-o (talk) 12:02, 19 April 2013 (UTC)
- I was experimenting a bit on how we could work with "subclass of". I din't make the changes systematically. The "instance of" is my mistake. It is depreciated because it should only be used for individual specimen (e.g. a really big diamond with a name). I think it is up to us to find an appropriate structure for the "subclass of" property. We could of course add both to each mineral (1) subclass of = mineral (2) and subclass of = silicate nomenclature, that you suggested (e.g. Amphibole). --Tobias1984 (talk) 12:50, 19 April 2013 (UTC)
- If amphibole is a subclasss of inosilicates that is a subclass of silicates that is a subclass of mineral, amphybole is automatically a subclass of mineral. I trying to setup a bot that will do these jobs in steps: in 1st step every mineral species will be a subclass of mineral, in next steps will refine this classification based on IMA classification. What about it? --Sbisolo (talk) 13:09, 19 April 2013 (UTC)
- Sounds good ;) --Tobias1984 (talk) 13:14, 19 April 2013 (UTC)
- @Sbisolo: I think that it doesn't work. Amphiboles and pyroxenes are items, most mineral groups and structural groups aren't one on Wikipedia. --Chris.urs-o (talk) 15:23, 19 April 2013 (UTC)
- Items don't necessarily have to be an article on Wikipedia. If it is relevant to us we can create hundreds of them and all of them can hold their own statements. Which will be practical to store general information about those mineral groups. --Tobias1984 (talk) 15:33, 19 April 2013 (UTC)
- @Chris.urs-o: most mineral groups and structural groups are on it.wikipedia. --Sbisolo (talk) 15:50, 19 April 2013 (UTC)
- @Sbisolo: it.wikipedia goes after the Fleischer's Glossary mineral groups. There are the rruff.info/ima/ structural groups as well. I'm in doubt. --Chris.urs-o (talk) 08:40, 20 April 2013 (UTC)
- You are right, mineral grouping in it.wikipedia is a bit messy. There are Fleischer groups and also IMA groups taken from published articles (for example "Nomenclature tunings in the hollandite supergroup" or "Nomenclature of the garnet supergroup"). We have to choose a grouping to follow. --Sbisolo (talk) 09:28, 22 April 2013 (UTC)
- @Sbisolo: it.wikipedia goes after the Fleischer's Glossary mineral groups. There are the rruff.info/ima/ structural groups as well. I'm in doubt. --Chris.urs-o (talk) 08:40, 20 April 2013 (UTC)
- @Sbisolo: I think that it doesn't work. Amphiboles and pyroxenes are items, most mineral groups and structural groups aren't one on Wikipedia. --Chris.urs-o (talk) 15:23, 19 April 2013 (UTC)
- Sounds good ;) --Tobias1984 (talk) 13:14, 19 April 2013 (UTC)
- If amphibole is a subclasss of inosilicates that is a subclass of silicates that is a subclass of mineral, amphybole is automatically a subclass of mineral. I trying to setup a bot that will do these jobs in steps: in 1st step every mineral species will be a subclass of mineral, in next steps will refine this classification based on IMA classification. What about it? --Sbisolo (talk) 13:09, 19 April 2013 (UTC)
- I was experimenting a bit on how we could work with "subclass of". I din't make the changes systematically. The "instance of" is my mistake. It is depreciated because it should only be used for individual specimen (e.g. a really big diamond with a name). I think it is up to us to find an appropriate structure for the "subclass of" property. We could of course add both to each mineral (1) subclass of = mineral (2) and subclass of = silicate nomenclature, that you suggested (e.g. Amphibole). --Tobias1984 (talk) 12:50, 19 April 2013 (UTC)
- I saw that u made Q104692 (actinolite) an "instance of" Q7946 (mineral) and a "subclass of" Q17159 (amphibole). Q50769 (pyrite) as well. Q17159 (amphibole) and Q192880 (pyroxene) a "subclass of" Q279440 (inosilicates), a "subclass of" Q178977 (silicate minerals), a "subclass of" Q7946 (mineral) is tempting. What now? --Chris.urs-o (talk) 12:02, 19 April 2013 (UTC)
- Your right. Best thing is to "subclass of" all varieties to their respective minerals. --Tobias1984 (talk) 08:23, 19 April 2013 (UTC)
- Q43513 (emerald) shouldn't be a subclass of Q7946 (mineral) but a subclass of Q103480 (beryl). --Chris.urs-o (talk) 07:35, 19 April 2013 (UTC)
Silicates
Currently we are linking different structures of silicates (neso, soro, cyclo, ino, phyllo, tecto) to silicate. This has reminded me that we should give a reason and a source for this. This is a pretty nice intro to the Libau silicate classification link. I was thinking that we could link for example Cyclosilicate with "subclass of" to silicate, set the qualifier to "Libau" and the source as "Friedrich Liebau: Structural Chemistry of Silicates: Structure, Bonding and Classification. Springer-Verlag Berlin, Heidelberg, New York, Tokyo 1985. 347 Pages, ISBN 3-540-13747-5". We are also propably missing some items (e.g. double ring silicates) Suggestions? --Tobias1984 (talk) 14:21, 20 April 2013 (UTC)
- I also have no idea how we should add the complicated strings of the libau classification. This is for example a single chained silicate: {uB,11∞}[pSiO2O2/2]2- ;) --Tobias1984 (talk) 14:27, 20 April 2013 (UTC)
- Stick to Dana and Nickel–Strunz, it is good enough, the items are from de.wikipedia and redirect from en.wikipedia, no reason needed. Libau is older. Prefixes neso, soro, cyclo, ino, phyllo, tecto are used by the CNMNC. --Chris.urs-o (talk) 19:26, 20 April 2013 (UTC)
Mohs hardness
Mohs hardness generally can't be expressed as a single value because often it is reported as an interval (example: olivine: 6½-7) because is a simple comparison between analyzed mineral and 10 mineral references in terms of "can be scratched", "can scratch". --Sbisolo (talk) 13:23, 23 April 2013 (UTC)
- How about this?: --Tobias1984 (talk) 14:38, 23 April 2013 (UTC)
| Mineral | Property | Value | Qualifiers |
|---|---|---|---|
| Mineral A | Mohs hardness | 5 | min |
| 6 | max | ||
| Mineral Kyanite | Mohs hardness | 4.5 | min parallel c-axis |
| 5 | max parallel c-axis | ||
| 6.5 | min parallel a-axis parallel b-axis | ||
| 7 | max parallel a-axis parallel b-axis |
- Sounds good :) --Sbisolo (talk) 15:30, 23 April 2013 (UTC)
- Thanks. We can still think about the details until the Datatype "Value" becomes available. --Tobias1984 (talk) 16:48, 23 April 2013 (UTC)
Color
Hey guys! The property for color became available (Property:P462). I tried it out on Fluorite and I could actually fill in all the colors listed on mindat with the exception of golden-yellow. It looks to me like we don't have to create that many new items for oddball colors. --Tobias1984 (talk) 18:22, 25 April 2013 (UTC)
- golden-yellow is equivalent to gold --Sbisolo (talk) 07:57, 26 April 2013 (UTC)
Luster
I have a doubt about luster: in page en:Lustre (mineralogy) lists adamantine, dull, greasy, metallic, pearly, resinous, silky, submetallic, vitrous, waxy. All of these but pearly and silky are related to refraction index. Is possible to order them in this way (from less to more refractive): dull, greasy, vitreous, resinous, submetallic, metallic. Pearly is commonly used for micas, silky is commonly used for some fibrous mineral. I think using the complete list is better because it reflects the common use of this property. --Sbisolo (talk) 08:27, 26 April 2013 (UTC)
- I think we can also switch the datatype to item. There are so few descriptions that datatype "string" would be not so ideal. As an alternative we could also call the property "luster resembles" and then just link to "metal", "wax", "pearl", "grease", ... --Tobias1984 (talk) 08:36, 26 April 2013 (UTC)