Nonoxynol-9 should probably remain separate as it has a distinct medical use but the others could easily be rolled into alkyl phenoxy polyethoxy ethers (or similar). The pages are generally of a poor quality being either unreferenced or primarily web-referenced. Much of the information on synthesis and uses can be combined. As for specific properties, I was thinking of a table where 'alkyl tail' and 'PEG length' were compared again the CMC and HLB. Anyone have any issues with this before I begin? --Project Osprey (talk) 12:32, 17 May 2016 (UTC)
I support the idea. Images of the structures would also be nice. I took a look at the IUPAC names. For the title of the merged article, I think it should be "Alkylphenoxypolyethoxyethanol" because all of them end with "-ethanol". Thoughts? Georginho (talk) 14:16, 17 May 2016 (UTC)
So is Triton X-114, which has the same general structure but a slightly lower level of exothylation (7.5 vs 9.5). This sort of represents my issue with these pages; none of them is actually describing a discreet chemical. They're all tradenames for mixtures of things with the same general formula. With ethoxylation that is somewhat inevitable but I still feel like these pages are largely duplicates of one another. --Project Osprey (talk) 21:34, 17 May 2016 (UTC)
Latest comment: 8 years ago5 comments2 people in discussion
Hi... Would someone please check and if necessary correct the File:DDQ-aromatization.png? It appears below. The relevant reference abstract (doi:10.1021/ja060050p) describes the reaction as involving a double methyl migration but as far as I can see one methyl has simply vanished. EdChem (talk) 04:18, 18 May 2016 (UTC)
I think the ref is doi:10.1039/J39710002566? I don't have electronic access to it (J.Chem.Soc. C from 1971), but think I can locate a print copy; will check next time I'm near the right library. DMacks (talk) 04:48, 18 May 2016 (UTC)
Okay, found the reaction in an index...should be a methyl rather than OH in the D ring. Good eye, EdChem! Under acidic conditions, that tertiary alcohol is lost as water to give a carbocation, then a 1,2-methyl shift. I uploaded File:DDQ aromatization rearrangement.png as a corrected replacement. Given how long that file has existed, didn't want to overwrite it at the same location, with anyone using it not realizing it needed to be replaced. DMacks (talk) 05:08, 18 May 2016 (UTC)
Latest comment: 8 years ago5 comments4 people in discussion
The article on bis(pinacolato)diboron claims that the "B-B bond adds across alkenes and alkynes to give the 1,2-diborylated alkane and alkenes. In the presence of organorhodium complexes, it can be installed onto saturated hydrocarbons. Citing doi:10.1055/s-2003-43344, it offers the following remarkable equation:
Would someone please figure out what this should be? It loses a B atom (maybe a coefficient error, though the text suggests a diborylated alkane) and the H36 appears as H<sub>3</sub><sub>6</sub> suggesting something might be missing. Thanks. EdChem (talk) 14:30, 21 May 2016 (UTC)
Probably my sloppy editing, and another editor and I are trying to fix it. Usually these B-B containing reagents release the borane and form B-C bonds. Thank you for catching this.--Smokefoot (talk) 14:59, 21 May 2016 (UTC)
(edit conflict):I have changed the reaction so its like the one in the "article". its a "spotlight" (very short review):
That is, consistent representation of each chemical unit rather than "turning around" the borane when it becomes attached to the alkyl. DMacks (talk) 19:10, 21 May 2016 (UTC)
Oxibendazole
Latest comment: 8 years ago3 comments2 people in discussion
Help! I wanted to disambiguate a link to Urethane in the article Oxibendazole, but I am totally at sea with what is presented. Chemistry is not my strong suit! The whole "Synthesis" section seems to be missing steps 1 and 5, and I wouldn't even know where to start to make it work. Could some kind soul check it out and at least fix the dab link? TIA! (Please {{ping}} me.) — Gorthian (talk) 04:28, 30 May 2016 (UTC)
@Gorthian: I've made some changes. FYI, the numbers refer to the compounds not the steps, so step one converts substance(1) into (2), etc. The urethane dab link definitely meant the carbamate functional group. Have a look, see what you think. :) EdChem (talk) 05:17, 30 May 2016 (UTC)
EdChem, thanks, much better! It's actually readable now. Seems I just can't let an article go till I've tried to fix it somehow... — Gorthian (talk) 05:58, 30 May 2016 (UTC)
Article-title help
Latest comment: 8 years ago3 comments2 people in discussion
I've been working on the cubane article, which has material about chemicals having this same cubic molecular form but no other chemical relationship. I've been adding material about other chemical analogs that have this shape but aren't necessarily strictly "derivatives" in a synthetic sense (I'll probably fork all of that off into their own page if I get enough to content). Is "cubic molecular geometry" a good term for this general structural class, a place to offload the general idea away from the specific chemical? DMacks (talk) 19:04, 6 June 2016 (UTC)
Latest comment: 8 years ago6 comments4 people in discussion
This strange thing appeared: Chapter 2.2: NH3 MOs with the implication that future "chapters" will be contributed. Possibly something more appropriate for Wikitextbooks or such. These are the outcome from an unsupervised graduate class out of UCLA.--Smokefoot (talk) 15:38, 12 June 2016 (UTC)
I've been trying to figure out what's going on there - they had wanted to work in WikiBooks, but we couldn't support that at the time. I suspect that there was either some miscommunication between the instructor and the students, or the instructor and us. I plan to contact the instructor on Monday. Ian (Wiki Ed) (talk) 16:18, 12 June 2016 (UTC)
I moved that one back to userspace. Should the Molecular point group (which started out as 1.2 Point Group) also go back into mainspace? I trimmed that back before I figured out why they were coming up with such odd names; I have my doubts about it as a stand-alone article even once it's trimmed, but it's possible that it's worth leaving as a stub. Ian (Wiki Ed) (talk) 16:27, 12 June 2016 (UTC)
I just extended the assignment date to catch anyone else who works beyond the end date of the class but the class is done so I'm not expecting anything more from them. Ian (Wiki Ed) (talk) 16:30, 12 June 2016 (UTC)
Of those two alternatives I would definitely choose the first, with two hyphens, but in prose I’d rather write Polymers derived from mineral oil or something, to avoid the piling-up of modifier-nouns (along with the hyphens).—Odysseus147922:36, 30 January 2016 (UTC)
Thanks for your reply. I agree, but the problem arises when the bracket is included. Currently, it says
It's "mineral-oil-based polymers" because we write "oil-based polymers". "Oil-based" is an adjective modifying "polymers".
"Mineral oil-based polymers" is not correct because it implies "mineral polymers", while we're actually talking about "mineral oil".
To prevent the hyphenation and bracket issue, I think "Polymers that are based on mineral oil and called 'plastics' are virtually non-biodegradable" is a better edit.
I would prefer "mineral oil-based polymer". The hyphen connects "mineral oil" to "based". This may be counterintuitive, but this usage is recommended in Fowler's Modern English Usage, which gives "kippered herring-merchant" as an example. Maproom (talk) 07:31, 15 June 2016 (UTC)
Latest comment: 8 years ago9 comments4 people in discussion
What is the point of this template? It appears to be a random selection of some articles about chemical compounds containing a nitrogen atom, along with really weird red links. (Pentafluoropropionitrile? WTF?) There is no organization, there is no attempt at listing significant classes or examples within that class, there is no apparent purpose, etc. It was created by a user that has been indefinitely blocked for disruptive editing and seems to have little to no chemistry knowledge. Unless someone wants to completely redo this template, I suggest removing it from all articles where it is used. ChemNerd (talk) 17:33, 14 June 2016 (UTC)
Agree, a pointless template, a listing of binary nitrogen compounds would have merit but not his one. V8rik (talk) 19:08, 14 June 2016 (UTC)
Regarding the complete set: this sort of thing makes sense for {{Uranium compounds}} (for example), but for an element like N it is ridiculous. It desperately needs splitting (although I suppose it would make sense to keep the current page with links to subtemplates, or else it is inevitably going to get recreated). Double sharp (talk) 03:46, 15 June 2016 (UTC)
I agree that most of the templates in Category:Chemistry compounds templates are reasonable, but that the Hydrogen and Oxygen ones suffer the same problem as Nitrogen. I suggest removing those three templates from all articles they are currently placed on then marking them with {{Deprecated template}} to prevent anyone discovering them in the future and thinking they should be used. ChemNerd (talk) 20:38, 15 June 2016 (UTC)
While we're on the subject of these element templates, I would also like to point out that the same disruptive editor involved in the problematic templates is also making a mess of categorization by trying to use the templates to categorize articles (see the history of Template:Fluorine compounds, for example). An explanation of the problems caused by this are outlined at WP:TEMPLATECAT. Can we agree that the element templates shouldn't be used to force categories upon the articles the templates are used on? ChemNerd (talk) 12:22, 17 June 2016 (UTC)
It would be best not to have the category there transincluded in the template, as subcategories would likely be a possibility. And some uses of transclusion, such as in talk pages or drafts should certainly not categorise them. Graeme Bartlett (talk) 22:40, 17 June 2016 (UTC)
Request Image
Latest comment: 8 years ago3 comments2 people in discussion
I have added a laboratory synthesis section to asparagusic acid. It's a fairly straight-forward synthesis from a diethyl malonate derivative. I would appreciate an image for this synthesis. Thanks. EdChem (talk) 13:25, 19 June 2016 (UTC)
Latest comment: 8 years ago7 comments5 people in discussion
Hi everyone, short questions: The introduction was changed, could anyone check it? Checking it might be worth anyway as English is not my first language. Regards --Minihaa (talk) 13:11, 15 June 2016 (UTC)
I've improved the English in the lead, and removed the first mention of "CH-acidic methylene group", as I don't see its relevance to the subject. A better chemist should figure out how it applies. Maproom (talk) 13:28, 15 June 2016 (UTC)
A welcome article. It would be useful to see what DMacks and I did. Liquids are clear, so dont say that. Rarely are compounds both colorless and yellow - the good stuff is usually colorless, bad samples are yellow. It is not toxic. Promoting that something is non-natural is misguided. The highly acidic CH2 is "everything" about this species. etc--Smokefoot (talk) 13:48, 15 June 2016 (UTC)
Bromine is clear, Guinness is not a pure substance. Clear does not mean colourless. Copper(II) sulfate solution, for example, is blue and clear. EdChem (talk) 14:57, 15 June 2016 (UTC)
Latest comment: 8 years ago1 comment1 person in discussion
Which chemist wants the honor of being the first in the history of chemistry to explain the following exciting new discovery concerning one of the most important chemical reactions in more detail so that normal readers can understand? To most people it's completely counterintuitive that conversion of weak bonds into strong bonds releases energy instead of requiring more:
Fire is hot because conversion of the weak double bond in molecular oxygen, O2, to the stronger bonds in the combustion products carbon dioxide and water releases energy (418 kJ per 32 g of O2); the bond energies of the fuel play only a minor role here.
Latest comment: 8 years ago6 comments3 people in discussion
Under the [2,3] shift and the 2,3 Wittig rearrangement an extra carbon atom is erroneously added in the products molecular structures of the rearrangement. The attention of the writer or the editor should be called to this to confirm. I may be wrong.
@Kayodekoleoso:@EdChem: The article is part of Mevans86's work on behalf of (?) Organic Reactions, a reference book from Wiley. It also publishes the chapters in their Wiki site, including this rearrangement here. The image there was already corrected. What is the policy of asking them to allow us using their image?
@Georginho: I see no reason you couldn't just take the corrected image and upload it here as a new version. I have no idea what copyright claim they might make, but I think it is {{PD-Chem}} and so uncopyrightable. EdChem (talk) 12:17, 25 June 2016 (UTC)
@Kayodekoleoso:@EdChem: Cheers. I copied the image and uploaded it as a new version.
Latest comment: 8 years ago3 comments3 people in discussion
So I submitted a draft article for the dark hydrogen discovery. It has been rejected. Feel free to do whatever you'd like to it, if you want to. I'm not going to improve it any longer. -- 70.51.200.20 (talk) 05:44, 30 June 2016 (UTC)
Latest comment: 8 years ago3 comments2 people in discussion
I'd appreciate if someone could check (and improve on, if desired) my recent edit to Bromocresol purple. I created the article a decade ago, but I am neither a chemist nor a microbiologist so I'm not confident in how accurately I'm reading these journal articles. Thanks, Pi.1415926535 (talk) 20:39, 8 July 2016 (UTC)
Latest comment: 8 years ago2 comments2 people in discussion
In articles like space-filling model, there are electrostatic potential surfaces included (in this case, for sulfur dioxide). The term "electrostatic potential surface" wikilinks to electrostatic potential which in turn redirects to electric potential. The electric potential article contains no example of electrostatic potential surfaces nor any but a mathematical discussion. I wonder if there is a better article for our readers to understand what an electrostatic potential surface of a molecule illustrates and how to interpret it? If not, should one be written? EdChem (talk) 14:26, 8 July 2016 (UTC)
There is a picture used in electronegativity that shows a brief synopsis. Electrostatic potential map of a water molecule, where the oxygen atom has a more negative charge (red) than the positive (blue) hydrogen atoms One challenge is that electrostatic potential surface means different things to physicists and chemists. In physics, an electrostatic potential surface is a surface of constant electrostatic potential. In chemistry, it seems like it is a surface of constant electron density upon which the electrostatic potential is calculated and color-coded. A chemistry-specific article or section may need to be written. --Mark viking (talk) 19:20, 9 July 2016 (UTC)
Latest comment: 8 years ago42 comments13 people in discussion
I'm introducing <ce> tag in the new features of Mediawiki. It uses the TeX's mhchem package that formats chemical equations automatically. See Help:Math#Chemistry. It's superior, easier and more maintainable than HTML tags. -- Cedar101 (talk) 16:54, 18 May 2016 (UTC)
More specifically, the discussion should focus on whether this capability should be used, and to what degree. There have been disagreements on how it's been applied, including uneven use within an article. Tarl.Neustaedter (talk) 16:58, 18 May 2016 (UTC)
Here are examples of the old and ce formats: old: BaSO4 "ce" . If there is a way to render the "ce" format sans serif, I am ok with it. Otherwise, chemists typically publish graphics in Helvetica or Arial fonts. --Smokefoot (talk) 17:30, 18 May 2016 (UTC)
Is there some documentation on this new package? In particular, I worry about "automatic". How does it handle formulae like magnesium sulfate heptahydrate? Tarl.Neustaedter (talk) 18:27, 18 May 2016 (UTC)
FWIW the above example doesn’t work properly in my obsolescent browser (Safari v5), probably for the same reason I had to change my math preference to use pre-rendered PNGs—the TeX version is much too small to read (in this case the entire formula takes up about the same space as a letter “m”, but floating high above the baseline). So I would oppose unless a similar workaround is made available.—Odysseus147923:12, 18 May 2016 (UTC)
When I try to copy and paste the example from firefox, it does not highlight, and after copy it has this plain text: "BaSO 4 {\displaystyle {\ce {BaSO4}}}". So the style sheet needs work, and there looks to be some other strange issues. Trying the suggested MgSO4·7H2O Failed to parse (syntax error): {\displaystyle \ce{MgSO4·7H2O}}
results in an error message for me. Trying an ion H2+ (looks OK for me) and an endohedral Ne@C60: . Graeme Bartlett (talk) 00:00, 19 May 2016 (UTC)
For addition compounds, simply use period(.). But It doesn't support at sign(@) that is de facto but non-IUPAC standard. -- Cedar101 (talk) 01:21, 19 May 2016 (UTC)
Reaction arrows all look terrible. It's as if the various parts of (each arrow-head, each shaft) are separate graphical elements that are mis-aligned with each other. Like a low-res .gif scan of a printed copy that isn't quite squared. And default font that is "yet another different standard" from other ways of doing things makes overall not better result. And both the font and the arrow suffer from lack of sharpness/focus, as if they were rasters that got rescaled. DMacks (talk) 02:34, 19 May 2016 (UTC)
Some are good, and some have thickened bits: eg . Trying our current practice A → B Failed to parse (syntax error): {\displaystyle \ce{A → B}}
. Graeme Bartlett (talk) 07:52, 19 May 2016 (UTC)
One good feature is the charge signs appearing above the subscripts. eg SO4^2- but it cannot handle unicode minus − which is our current practice. SO4^2− Failed to parse (syntax error): {\displaystyle \ce{SO4^2−}}
. So it would be good to add support for ← → − · and • if we are ever going to convert existing formulae. Graeme Bartlett (talk) 07:47, 19 May 2016 (UTC)
Renders [ and ] for complexes differently: and and hapticity? Failed to parse (Conversion error. Server ("https://wikimedia.org/api/rest_") reported: "Cannot get mml. TeX parse error: Misplaced &"): {\displaystyle {\ce {[Fe(η5-C5H5)2]}}}
and Failed to parse (syntax error): {\displaystyle \ce{[Fe(η5-C5H5)2]}}
and Failed to parse (Conversion error. Server ("https://wikimedia.org/api/rest_") reported: "Cannot get mml. TeX parse error: Misplaced &"): {\displaystyle {\ce {[Fe(η^5-C5H5)2]}}}
and Failed to parse (syntax error): {\displaystyle \ce{[Fe(η^5-C5H5)2]}}
and even Failed to parse (SVG (MathML can be enabled via browser plugin): Invalid response ("Math extension cannot connect to Restbase.") from server "http://localhost:6011/en.wikipedia.org/v1/":): {\displaystyle \ce{[Fe(η^5C5H5)2]}}
. Still some issues, separate from the look which Smokefoot has described well. EdChem (talk) 08:04, 19 May 2016 (UTC)
I vehemently oppose using this feature for inline text. If you write a paragraph and use this feature for the chemical formulas in that paragraph, it will be unreadable. The formulas will have different fonts and sizes than the normal Wikipedia font.
Outside inline text, i.e., new lines for chemical reactions, I would be okay with it.
On the contrary, many ions with high oxidation numbers, such as H2O2, MnO4−, CrO3, Cr2O72−, and OsO4 can gain one or two extra electrons and are strong oxidizing agents.
with this:
On the contrary, many ions with high oxidation numbers, such as , , , , and can gain one or two extra electrons and are strong oxidizing agents.
Just an outside observer, but I can't quite follow what problem this is supposed to solve. It isn't obvious to me how it's "easier and more maintainable". It might be helpful to explain how this works, Cedar101 (especially to people who know nothing about TeX). Ian (Wiki Ed) (talk) 19:22, 19 May 2016 (UTC)
I updated the comprehensive comparisons of <ce>...</ce> and equivalent HTML codes in Help:Math#Chemistry. You can compare them. In my opinion, HTML codes are too verbose to write and read. -- Cedar101 (talk) 05:27, 20 May 2016 (UTC)
@Cedar101:, I think you are going overboard with the math help page. You do not need to fill the page with examples for every feature, and you certainly do not need to list every difference between <ce>...</ce> and HTML. In fact, HTML should not even be there. We have {{chem}} for inline chemical formula, so I don't really see any point to this. -- [[User:Edokter]] {{talk}}06:51, 20 May 2016 (UTC)
That is not what I ment; HTML examples should not be there at all... I told you we have {{chem}} for this, so that should be used instead. -- [[User:Edokter]] {{talk}}08:27, 20 May 2016 (UTC)
I would like to ask something about this {{chem}} template. What is (again) the consensus? Compare H 2O with H2O. As you can see, the number 2 is bigger if you use {{chem}}. This enlarges the space between the lines. I think it's also a reason why we don't use it in our Manual of Style.
The difference was unintentional (will be fixed soon). The template's purpose is to save typing, surely no consensus is needed for that. It also offers additional formatting like SO2− 4. -- [[User:Edokter]] {{talk}}11:17, 20 May 2016 (UTC)
Consensus?
Given what appears to be pretty negative views or skepticism above, is it agreed that the TeX formatting not be implemented? or implemented for the following situations, or awaiting repair of the following glitches. But somehow, it seems, we need a proposition to seek consensus about.--Smokefoot (talk) 16:52, 19 May 2016 (UTC)
I'm against 1) automatic implementation and 2) inline implementation.
I would keep the status quo. But I would leave it to any editor to decide, on case-to-case bases, if it's best to use the feature. Like Graeme Bartlett said, it does put the charge directly above the subscript. And I have no example directly in my head right now, but I think the ability to put texts above and below an arrow could be useful.
Failed to parse (syntax error): {\displaystyle \ce{A → B}}
Failed to parse (syntax error): {\displaystyle \ce{SO4^2−}}
Failed to parse (syntax error): {\displaystyle \ce{[Fe(η5-C5H5)2]}}
Failed to parse (syntax error): {\displaystyle \ce{[Fe(η5-C5H5)2]}}
and so on. I'm running the absolute latest MacOS (10.11.5) and latest Safari (9.1.1), so this isn't an issue of obsolescence, but assumptions being made by the package which aren't working. Tarl.Neustaedter (talk) 19:02, 19 May 2016 (UTC)
The issue is that simply viewing this talk page shows a dozen errors. Having Wiki pages which appear to users with errors is simply unacceptable. This isn't ready for general use. Tarl.Neustaedter (talk) 14:06, 20 May 2016 (UTC)
I think that those that know what they are doing with it can use it in reactions that stand alone on their own line. We should discourage its use inline (unless there is an option to match the font of the text in the paragraph). Graeme Bartlett (talk) 06:14, 20 May 2016 (UTC)
I think one of the problem with ce versus chem is that you cant copy/paste the line (without getting a lot of TeX stuff too). E.g. copy/paste this H 2O gives "H2O", while gives "H 2 O {\displaystyle {\ce {H2O}}}". One example from the top of this thread: "On the contrary, many ions with high oxidation numbers, such as H 2 O 2 {\displaystyle {\ce {\mathsf {H2O2}}}} , MnO 4 − {\displaystyle {\ce {\mathsf {MnO4^{-}}}}} , CrO 3 {\displaystyle {\ce {\mathsf {CrO3}}}} , Cr 2 O 7 2 − {\displaystyle {\ce {\mathsf {Cr2O7^{2-}}}}} , and OsO 4 {\displaystyle {\ce {\mathsf {OsO4}}}} can gain one or two extra electrons and are strong oxidizing agents." Christian75 (talk) 08:29, 21 May 2016 (UTC)
Side-notes about pre-consensus editing -- cancelled in GF understanding
The following discussion has been closed. Please do not modify it.
Aside from whether this is a useful tool or not, User:Cedar101 is going around implementing it before we've even made a decision (for example, 1, 2, 3). This is really annoying. I think User:Cedar101 is on a mission to make this happen come what may. That's not how we roll here. --Ben (talk) 19:39, 20 May 2016 (UTC)
Cedar's implementation also includes replacing images (see this edit to isomer regarding the isocyanate ion). Using this tool for ion structures is raises the possibility of non-linear structures being presented as linear - imagine a similar representation of ozone. I agree with Ben, implementation is premature, especially given that the in-line changes look awful and the concerns of wiki-chemists deserve discussion. EdChem (talk) 01:41, 21 May 2016 (UTC)
It would appear our discussion here has served no purpose; I notice Cedar101 is still busy modifying articles to use his new formatting template. So what if some users get errors when they try to read articles? Tarl.Neustaedter (talk) 01:56, 23 May 2016 (UTC)
I gave him a level-3 warning for disruptive editing. If I weren't involved, I would have blocked him myself. If he continues, please alert us here and/or at the appropriate admin notice-board. DMacks (talk) 03:05, 23 May 2016 (UTC)
Okay, I'm sorry for my bold editing. I will not edit the HTML notations and text-like({{Should be text}}) images until the consencious is reached. I'll edit only the existing <math>\mathrm</math> equations. -- Cedar101 (talk) 04:25, 23 May 2016 (UTC)
This is the author of mhchem. While I am happy to see mhchem begin integrated into MediaWiki and Wikipedia, I also see the need for improvements. Most importantly, the current implementation modifies what a user inputs. In some cases, this leads to wrong output. I also think, the alternative representation (see copy&paste) could be improved. Maybe we even find a solution for the 'inline' problem. Many of the parsing issues (unicode minus, etc.) are already implemented on my side, but they need to be packed for shipping and go through the various update lines. Mhchem (talk) 20:11, 23 June 2016 (UTC)
Latest comment: 8 years ago4 comments3 people in discussion
The pages Iothalmate and Iotalamic acid are both spelt wrongly (should be Iothalamate and Iothalamic acid respectively) and I think they should be merged, probably as Iothalamic acid. As far as I know merging two pages into a new third page requires admin rights. I'd be grateful if someone could sort that out. --Project Osprey (talk) 21:25, 22 July 2016 (UTC)
According to drugs.com, "iothalamic acid" is the USAN and "iotalamic acid" is the recommended INN. Since WP:PHARMOS prefers using INNs as titles, I think the content at iothalmate can be merged into the article at iotalamic acid and iothalamic acid and other synonyms/spellings can become redirects. No admin rights should be necessary for this, but I'd be happy to help out if needed. -- Ed (Edgar181) 21:37, 22 July 2016 (UTC)
Latest comment: 8 years ago4 comments3 people in discussion
There's a thread starting on quora as to why our article on Sodium bisulfate has an infobox that says it decomposes in alcohol. Would someone who understands chemistry care to run an eye over the article? Thanks ϢereSpielChequers14:25, 22 July 2016 (UTC)
The statement is vague and unreferenced, and doesn't make sense chemically to me, so I have removed it. -- Ed (Edgar181) 15:23, 22 July 2016 (UTC)
Latest comment: 8 years ago48 comments9 people in discussion
Hello everyone! We have put substantial effort into improving data quality and compound representation in Wikidata (data coming directly from Drugbank, Pubchem), especially for FDA approved drugs, but more to come. As Wikidata is currently evolving into a generalized, structured data store for all Wikipedias, my suggestions would be to modify the chemical compound and drug infoboxes so they can pull as much data as possible from Wikidata. A good start would certainly be to get the chemical identifiers like CAS, UNII, Drugbank ID, Inchi key, Inchi, Smiles, etc from Wikidata. As we have shown for the Gene Wiki infobox, infoboxes can pull their complete content from Wikidata. In order to be able to do that for chemical compounds, some data in Wikidata is still missing (e.g. legal status, pregnancy category), but adding those should be fairly easy.
Main motivation:
High data quality due to data import from open primary data sources like Drugbank, Pubchem, NDF-RT, ChEMBL
All data with rich references to original data source.
Regular, comprehensive updates from the original source by (our) bots
Many more chemical compounds in Wikidata than in English Wikipedia, if a new article on a compound should be created, the structured data is most likely already in Wikidata. Also allows easy creation of compound stub articles.
What would be your view of this proposal? I will create a sandbox demo and post it here as soon as I have it. Sebotic (talk) 09:22, 27 June 2016 (UTC)
I would be in favour of this change. I see the benefits that you describe (I've been doing some of those things here locally, and it is a pain parsing the chemboxes and not flooding en.wikipedia). User:Sebotic, as most people will not be watching wikidata-changes (even though that can be done through a setting; Special:Preferences#mw-prefsection-watchlist, select option "Show Wikidata edits in your watchlist"), can you inform us a bit about anti-vandalism measures on WikiData regarding this, mainly, numerical data that most people do not know by heart (we know that if s.o. is changing the boiling point of water to 52°C it is wrong, but for more obscure liquid chemicals 99+% of the people seeing the change in numerical value would need to look it up to see whether the change was for the good or the bad). --Dirk BeetstraTC10:45, 27 June 2016 (UTC)
Sure. One question is, should d: overwrite any local input, or keep local input as prio?
And I'd like to hear Beetstra on this wrt the CheMoBot. Using wikidata for say CAS RN would make the bot verify/validate process (and the tick/cross markings) superfluous in this, right? -DePiep (talk) 10:53, 27 June 2016 (UTC)
I would !vote for overwriting all local data, and then remove all local data after full migration.
Regarding CheMoBot - I'd consider to abandon it on en.wikipedia, and if/when time allows consider a version on wikidata. In the way, it is related to the vandalism question I posed to Sebotic (though it also poses a question of 'how correct is the data on WikiData'). --Dirk BeetstraTC10:56, 27 June 2016 (UTC)
We could check for "local input differs from wikidata value", and categorise them. (After a cleanup that cat can go). -DePiep (talk) 10:58, 27 June 2016 (UTC)
That should be easy to program (now) into the chembox: 'if property on WikiData != local value, put it into category 'Wikipedia pages where property X has a different value than on WikiData'. For the true numerical data (identifiers) that should be a good comparison - for the rest it is likely going to be more hellish (depending how the local and WikiData value are stored - which unit, or whether units are given on both sides). --Dirk BeetstraTC11:16, 27 June 2016 (UTC)
We could use a tracking cat for each main data point apart (CAS RN, PubChem, ..): "Category:CAS Number locally different from wikidata". -DePiep (talk) 11:40, 27 June 2016 (UTC)
That is the type of thing that might be giving us insight and a feeling whether switching to WikiData is a go - see what percentage of articles with a CAS number has a different CAS number locally in comparison with WikiData, and then from those, do a random check to see which ones are more correct on average.
Wikidata might be better for identifiers, because (I assume) bots can relatively easily get other identifiers once one is given (judging by the more number of identifiers on wikidata than here), but not for other data items. GalobtterTalk to me!16:29, 2 July 2016 (UTC)
@Beetstra:@Christian75: I agree that there need to be simple ways to track changes to the Wikidata item. Fortunately, User:H4stings on the French Wikipedia recently created a JavaScript extension to display the history of the Wikidata item which is linked to a certain Wikipedia page. That seems to work quite well, and as his version was only working on French Wikipedia, I modified his script so it would work with English Wikipedia. I expect that there will be some kind of official extension soon to handle that.
Regarding vandalism in Wikidata in general: It has been really low and most errors are rather mistakes by users than intentional vandalism, but certainly, the more users, the more vandalism is to be expected. Therefore, I think integrating the revision histories of Wikipedia and Wikidata item is a good way to go. In addition, the Wikidata SPARQL endpoint allows for extensive consistency checks of data and our scheduled bot runs will also make sure that data stays clean. Sebotic (talk) 08:48, 28 June 2016 (UTC)
It should be possible to add {{citation needed}} (its not possible in the gene infobox). And for non-trivial values it should be referenced too (at enwiki; it may be possible to retrieve the refererence from wikidata too?). Christian75 (talk) 11:55, 27 June 2016 (UTC)
How much vandalism has Wikidata? Personnaly, I check vandalism/changes in chemical articles from the chemicals article history (and not my watchlist) Christian75 (talk) 11:55, 27 June 2016 (UTC)
We don't have to source reference each and every point, do we? And anyway, the source is in wikidata, not to be added locally. -DePiep (talk) 12:09, 27 June 2016 (UTC)
The data should be referenced on WikiData - there it should be clear where the number is coming from. If you don't believe a datapoint, it should be tagged on WikiData. If there are references in WikiData, it should be trivial to transclude these as well (though I don't know if that is to be desired, a ref for each datapoint). It would also be great if the data on WikiData is in need of a citation, that that is transcluded as well. --Dirk BeetstraTC12:38, 27 June 2016 (UTC)
I agree that references in Wikidata only should be fine. If a users really wants to know the origin of the data, the Wikidata item is just one click away, to make that easier, a dedicated Wikidata link could be added to the infobox. (technically, it would not be a problem to pull the reference data from Wikidata and add e.g. a tooltip to each identifier in an infobox) Sebotic (talk) 08:48, 28 June 2016 (UTC)
I think that in principle this could be a good idea. BUT.... There is a one big problem - there is currently no agreed definition as to what the subject of a WikiData item is (ok that is a bit of a problem on wikipedia pages too). My point is; take a look at the wikidata entry for Norvaline. There are 2 SMILES strings - one for the L-isomer the other with no defined stereochemistry - there are similarly 2 different UNIIs for the two isomers - looking at the associated Wikipedia articles the EN, FR, JA articles seem to be considering the racemate - the Italian one lacks any detail about the structure and the DE, RU and PT wikis seem to be about the L-isomer. Of course it is possible to create separate wikidata items the real question is: What are the criteria for knowing what chemical species a wikidata item is about? For organic compounds the InChI/InChI key could be a reasonable (but admittedly imperfect) mechanism for 'registering' a compound. Another case in point is the entry for Sparteine, the image used is for (-)-sparteine, there are two SMILES strings one with stereo one without, the InChI is for the (+)-isomer and there are 2 InChIkeys one for the (+)- and the other for the (-)- isomer. Depending on what field of the item you are looking at would affect what data might be added or changed in the item. Therefore, it is my contention that there is a need for a field that identifies the chemical species that relates to the wikidata item and to which all other data can be compared/validated. And this needs to happen before we try and map ChemBoxes to WikiData items --The chemistds (talk) 14:23, 27 June 2016 (UTC)
DePiep - I'm saying that at the moment it isn't clear to me how the three Norvaline substances are defined in the the WikiData model. Specifically, what attribute should used to define what chemical species a WikiData item refers to. Ideally, such an attribute would only have one value per WikiData item and would be unique to that WikiData item - and would be the fact that one could use to validate all of the other data fields against. Names are not specific enough as they often have one-to-many mappings to structures, depending on context and level of specificity. --The chemistds (talk) 15:42, 1 July 2016 (UTC)
Stereochemistry is a complicated beast, I agree. So the general consensus in Wikidata is that everything which has it's own Inchi key gets its separate item. So for the D-, L- and racemate form, a separate item should be created. These items can be liked to each other in Wikidata, e.g. the DL- racemate item 'has part' the D- item and the L- item. With Lua code, that info can be retrieved from Wikidata an represented in a chembox. Sebotic (talk) 08:48, 28 June 2016 (UTC)
Sebotic - could you point me to where the consensus about everything getting InChIkeys is as I haven't seen that discussion and need to read up on it. --The chemistds (talk)
@The chemistds: Here is the discussion, it does actually talk about CAS numbers, but as CAS numbers are ambiguous sometimes as well, imho, the conclusion is that InChI keys should be used to solve the stereochemistry issues. What also speaks in favour of having one Wikidata item per stereoisomer is the importance of the configuration in biochemistry. The great thing about InChI keys in this context is that one can immediately recognize two stereoisomers when looking at the InChI key (At some point, each of these items should have all relevant identifiers and notations anyway, smiles, inchi, inchi_key, etc). But I think it would be a great time for you to join the discussion on Wikidata as well, getting things right at this point is important because they will be substantially harder to change later on. Sebotic (talk) 20:43, 1 July 2016 (UTC)
I have no familiarity with wikidata so I feel I should ask: if I find data that it wrong, how do I change it? Will the chembox still be there? If I change data in the chembox will it just be changed back? If I fill blank values in the chembox will they be changed back? In sort, how will this work for a normal editor? 09:25, 28 June 2016 Project Osprey (sign added)
The {{Chembox}} will be there. But it will be setup to read & show wikidata for main properties (it will read CAS RN directly from wikidata). Edits for that data should be made in wikidata, which is more difficult for us regular editors. (wikidata is supposed to have high-quality data for example by bots scraping the defining databases like PubChem). We could set it up so that the locally entered value takes prevalence (will be shown). Also, we could add tracking category ("Chemicals having CAS RN fromm Wikidata"). -DePiep (talk) 14:17, 28 June 2016 (UTC)
Thanks. A lot of the discussion so far has been about verifying existing data but I'm also interested in how adding new data will work. If I create a new page with a new chembox, or fill-in a blank field in an existing chembox, will the new data be moved automatically to wikidata or will I need to make edits in both wikipedia and wikidata?--Project Osprey (talk) 14:34, 28 June 2016 (UTC)
I don't know. wd is a black box to me, let alone editing wd. Also unhelpful is that nowhere (nowhere) is explained what the data model is. (This far I got: what we call "pagetitle" (article) in wd is "item", a "Q.." number). -DePiep (talk) 22:41, 28 June 2016 (UTC)
... but of course en:Wikipedia:Wikidata helps. It does, for the 2013 interwiki link change (iw link was the easiest, earliest, most bottable wd feature. Actually, the only advertisement. Note the word "collaborative" in the lede). I did not check HELP:Wikidata. -DePiep (talk) 22:47, 28 June 2016 (UTC)
I would assume that you would have to edit the wikidata page and the chembox would update with the data. If it worked otherwise it would seem very weird and a pain to manage. GalobtterTalk to me!16:17, 2 July 2016 (UTC)
So I was fiddling with editing wikidata and it seems reasonably easy to edit, but data quality (significant figures and references) seems lower than in the corresponding chembox, atleast for the pages I checked: water and ethanol. So I'm not sure about automatically overwriting local data. Also is data in chembox going to be automatically transferred to wikidata if there isn't the corresponding piece of data on wikidata? (wherever possible that is) GalobtterTalk to me!16:12, 2 July 2016 (UTC)
Sorry to arrive so late to the discussion, but I'd like to add my support to this change. I pretty much agree with everything mentioned above, and I'd like to see us perhaps do some tests before we dump all our Chemboxes. In particular, I'd like us first to save the information stored currently in en:wp chemboxes into a database - with citations - if someone can write a bot to do that. I had planned to spend much of my free time for the next month checking entries in Wikipedia:WikiProject_Chemicals/Index, but if CheMoBot becomes obsolete that's a waste of my time. How can I help on Wikidata and it's implementation on WP:CHEM, Sebotic? I'm happy to help out for the next few weeks if I can. Walkerma (talk) 23:23, 12 July 2016 (UTC)
Overwritten local data: it is depending on your lua module (see module:Wikidata) importing data from WD. But the consensus is that local data prevents the import and the display of WD data. This allows to create a local consensus when the topic is disputed and no "international" consensus can be found.
Anti-vandalism tools: Not so many until now. Along the traditional watchlist, we are using constraint analysis which allow us to detect some non compliance to some rules (like the format, the uniqueness,...). See for example the constraint reports about Pubchem CID here: this identifier should have a specific format (only numbers), to be unique among all Pubchem CID used in WD and only one Pubchem CID has to be present in one item. As you can see in the report, we have 6 items with several Pubchem CID and we have to clean these items. If it is correct to have several identifiers for one chemical, then we can add it in an exception list in order to remove it form the constraint report. Constraint reports are updated every day. This doesn't help if someone wants to introduce wrong data but in order to do it, he has to follow some rules: 1) to respect the format of the data (for example, CAS numbers have to have the following form XXXXXX-XX-X) and 2) to be different from all other CAS numbers existing in WD. These kind of rules prevent stupid vandalism.
Then there is a general tool to fight vandalism, ORES, which is working in WD.
Reference. Display of reference is possible according to the local lua module importing data from WD (module:Wikidata and perhaps other modules like Module:Infobox). This should be solved in WP:en by programming the import of the reference data and the insertion of the reference in the chembox.
Isomers/stereoisomer: From definition point of view each isomer/stereoisomer and each mixture of isomers/stereoisomers has to have an item in WD. Relation between stereoisomers and mixture is done by "instance of" relation. For example d:1-butanol is considered as an instance of d:butanol. Butanol is considered as a class (a collection of instances) of different type of chemicals. Same for d:(S)-2-pentanol and d:2-pentanol. This classification is based on ontology scheme but need to be developed in WD in order to create a logical relations network between items. A lot of work is still needed to distinguish what is the mixture and what is the isomers but this is in progress (see here for a list of chemicals needing an analysis based on UNII identifier).
Depending on the rules you have in WP:en, you have to change the wrong values in WD in order to change the display in WP:en articles. Or to write the correct value in the wikicode of the article in order to avoid the data import from WD (if local values overwrote WD data).
My personal advices:
import data from WD only if the data have a reference in WD (reference in WD can be detected by the used of property "stated in" in the reference section of the statement (this allows to neglect data imported from other WPs recognized by the use of the property "imported from")). This can be done by defining a constraint before data import from WD in the wikicode of the chembox. See your lua modules for details.
perform once a comparison check between your data in your chembox and the data in WD. This is the best to decide if the data set in WD has a similar quality as the data already present in WP:en.
your local data should have the priority over WD data. Perform the deletion of your data only after checking that the data in WD is the same. If there is a difference, first check if your article is connected to the right item, if yes try to solve the data contradiction before any deletion.
Thank you for your comments Snipre with respect to the points I raised a lot of people have seen this as an issue of stereochemistry which it isn't really - it was just in the examples I chose that stereochemistry was the problem. My point is about the identity of a chemical compound and I don't think that the use of an ontology really addresses that. The ontology defines some relationship between two concepts (in this case Wikidata items) - I'm concerned about how a chemical compound is defined as a concept - ie what feature of wikidata item can be used to determine if new data maps to an existing wikidata item or a new one - and if it is a new wikidata item what ontological relationships that item has with other wikidata items. --The chemistds (talk) 15:27, 18 July 2016 (UTC)
@The chemistds: This is not yet clearly defined: as the chemistry project is not well represented, no rule was defined so any good proposition can be presented. My first idea was to used the InChI as parameter: if you can define an InChI which is different from the others, you can create an item. But then I found the problem of the conformers and tautomers. Here I don't know if we have to go into that level of details. So my proposition is the next one: any chemical which have an unique InChI and can be physically isolated in a pure form to measure some properties. But again we are still free to define the rule if we want to have one. here we reach an ontological problem: we should first define the purpose for the data we want to include in the WD database. And I don't think we should copy the Zinc database which collects all possible chemicals. We should aim for chemicals with more than a chemical formula and a drawing of the chemical structure. Snipre (talk) 16:46, 18 July 2016 (UTC)
My angle, in subtopics:
1. For now: in {{Chembox}} and {{Infobox drug}} : show local input value over d: value. (~17K articles on enwiki)
2. Use tracking categories, smartly.
3. For high-end topics like CAS Number: use track cats: "CAS from local", "CAS from d:", "CAS d: and local differ", "CAS locally set to 'none'", "No CAS at al", etc.
4. Interested editors can check & improve situations (by tracking cat, edit d:). Say, one year from now, we can switch (because enwiki-chem editors trust d: & have fixed diffs): d: overwrites local input for CAS Nr.
5. Later on: what about multiple-value articles (multiple CAS Nrs?). But hey, rule #1 does not disrupt this.
It should be relatively trivial to rewrite one of our subboxes to compare local data to WikiData data (I would suggest that the first subbox to apply this to would be the identifiers box), and then categorise those into hidden maintenance categories (WD = Local; WD yes, Local no; WD No, Local yes; WD != Local, WD referenced; WD != Local, unreferenced - and that for each identifier). We wait a couple of days for the categories to populate properly, and then see what the contents is in each of them (total numbers of . That should give us a feel for the situation without 'disturbing' anything we have here. For now we are just blind in saying anything. For all we know the data on WikiData is all the same as what we have here, or just better than what we have here. When we know some numbers (maybe doing some maintenance where needed), we can then decide on the first subbox to move to WikiData and 'wipe' the data locally where appropriate (I like the idea of being able to overwrite where e.g. stereoisomers between local and WD don't match, but there then still should be the aim to solve those situations - that however has a lower priority than improving locally incorrect data). At the time of move, I would then suggest to use a Y/N/? system (like CheMoBot) to mark differences (but at that time we abandon CheMoBot, for now it may be a way of getting some data updated on WD which is there incorrect).
I agree with all the identifiers. One big problem with the rest of the data is the state (temperature/pressure) where we assume its standard state (at 25 °C, 100 kPa) but Wikidata doesnt. (btw. the info is broken for some infoboxes, eg. ethanol, ethane but not CoCl2). Christian75 (talk) 07:15, 19 July 2016 (UTC)
Just a comment: don't modify your current chembox when activating data import from WD but create a new infobox. Why ? To avoid any display problem when restoring an old version of the article.
@Christian75: WD requires pressure and temperature when necessary so when the value is different between WD and WP, just add an additional check to verif if the pressure and the temperature in WD are the ones of the standard state. Snipre (talk) 13:28, 19 July 2016 (UTC)
@Snipre: that would mean about 10.000 edits to replace the old infobox with something with a new name that is not as easy. For the chembox we might then go to 'infobox chemical', but for infobox drug that is not really an option. --Dirk BeetstraTC13:20, 19 July 2016 (UTC)
Dirk Beetstra This is just a comment, feel free to do as you want. But most of the time, adapting the current infoboxes to include lua programming can lead to deep modifications which can lead to strange output when applying old wikitext to modified templates. I don't know your chembox and how complex it is, so perhaps my comment is out of scope. Snipre (talk) 13:28, 19 July 2016 (UTC)
@Snipre: it is very complicated, so I do see your point. I guess that the editor(s) who is/are going to try to implement this should try this on versions of a couple of years old to see if it generally doesn't break. Some of these problems do exist already, since some parameters have 'moved on' (either are not supported anymore, or have moved between sub-boxes, or have been repurposed). --Dirk BeetstraTC13:42, 19 July 2016 (UTC)
Dirk Beetstra We had that kind of discussion in WP:fr especially because one of the famous sentences of WP is "Feel free to contribute" based on the fact that wrong edits or bad modifications can be reverted to a previous version without any lost. But in reality and especially concerning templates we saw that some modifications in templates don't allow to restore old versions. This problem is more a breach in the Wikipedia concept than anything else and perhaps a stricter rule concerning modifications of templates should be discussed at community level. Snipre (talk) 15:27, 19 July 2016 (UTC)
If I understand this subthread right, I do not agree. I don't sdee the requirement having to accept old page versions. Template changes are improvements, and we should not be tied to old versions. That said, editors might want to restore old data (not page version). That of course is covered. -DePiep (talk) 16:45, 25 July 2016 (UTC)
@Snipre: I'm new to Wikidata. I checked ethanol[4] and saw some entries without temperature, eg. crystal system (and related) (the temperature was in the reference), and pKa. Is it correct I can't add some kind of citation needed. I have two choices, ignoring/fixing it, or removing it (without the posibility to give a reason in edit summary?). Christian75 (talk) 14:48, 19 July 2016 (UTC)
@Christian75: Because for these two properties no requirement about temperature was defined. Thanks, I will add these requirements as mandatory constraints. For your information, constraints for properties are defined in talk's page of the property and missing data can be tracked using the constraint report. See for example density property where the pressure, the temperature and the phase have to be specified and then the corresponding constraint report to see how people respect these requirements. Snipre (talk) 15:27, 19 July 2016 (UTC)
@Beetstra: I am not so sure about your point regarding data in WD being the same as in the English WP chem infoboxes, my impression (based on the imported from citations) is that the data are a mashup of data imports from different Wikipedias, primarily from the English, German, Russian, Spanish, Italian. But the next step will be a data quality analysis and correction in WD anyway. Furthermore, I am currently working on an infobox displaying and comparing infobox content from English Wikipedia and Wikidata. See the Infobox drug sandbox. Sebotic (talk) 23:48, 19 July 2016 (UTC)
@Sebotic: I think you're making the same point as me, in a different way. We have some clue about the 'correctness' (or incorrectness, if you like) of many values in the chembox. What we don't know is how well WD matches up with en.wiki. I presume that actually WD does pretty good, and we could safely switch (i.e. if the data on WD is the same as on en.wiki, whether correct or wrong on both ends). But then, it is rather easy (through some clever template engineering) to compare (especially for the identifiers) and then get some numbers. Whether that is tested on the drugs or on the rest of the chemicals is equal to me, as long as we have a reasonable subset - I expect that the relative numbers are going to be the same (for the identifiers) between chembox and drugbox (in the time that I was actively working on the identifiers I have made no discrimination between drugboxes and chemboxes). --Dirk BeetstraTC11:17, 20 July 2016 (UTC)
@Sebotic:, I saw your infobox drug/sandbox. Can I build the {{Chembox}} sandbox with a slightly different approach? I'm aiming at the maintenance categories Beetstra mentioned above. Indeed they should be helpful for the maintenance tasks (comparing WD-value vs local value). -DePiep (talk) 16:56, 25 July 2016 (UTC)
@DePiep: Sure, I was only trying to display values from either the infobox parameters or Wikidata, depending where data exists and if it matches. Seems to be doable, but I agree that maintenance categories are a better way to get it done. I will explorer if it makes sense to offload some of the functionality for pulling chem data from WD into dedicated Lua scripts. Thanks! Sebotic (talk) 17:13, 27 July 2016 (UTC)
Thx. Yes a module would be useful. Some major ideas I want to push, eh propose:
Better not use visible notices (like Y/N or color-marked text). Because the Reader should not be bothered with quality-markings. Tracking categories should do the trick.
Both {{Infobox drug}} and {{Chembox}} should use the same setup (cat scheme, logic, etc). As I have done for current tracking categories, both infoboxes can & should populate the same, say, CAS RN category ({drugbox} and {Chembox} articles together in there).
There is a third major issue. That is: {{Chembox}} today is build in wikitable format (for example, see the simple datarow/subtemplate {{Chembox AdminRoutes}}). And it has ~555 parameters.
Q: So why not make it use basic {{infobox}}, and maybe even use Lua for parameter treatment?
A: Because it works great, so far (as in: very very far). Current structure supported very many (isolated) improvements over the years.
Detailing: I have edited the last three years both {drugbox} and {chembox}. All data-row changes (that is: individual data input requests) could be handled nicely in the templates, based on content talk on talkpages. Not one request has been denied for template-technical reason. Meanwhile, same-data got the same treatment & presentation in both {drugbox} and {chembox} (eg CAS Number).
Q: Then when will {chembox} go into Lua and {infobxox}?
A: before or after {chembox} goes using wikidata as talked here. But one can not develop {chembox} into Lua/{infobox} when the core is changing simultanuously wrt WD.
My opinion: for both {drugbox} and {chembox}: let's go wikidata for identifiers & classes first. That's best for the Reader. -DePiep (talk) 21:04, 27 July 2016 (UTC)
Latest comment: 8 years ago4 comments3 people in discussion
I can't decide is this is a COI page or not. Its essentially been made a single editor and cites 10 papers, all by the same author... which is often a bad sign. I also find it had to believe, that in the 100 or so years we've had the Fischer esterification, that no one else has done this - it feels like someone has slapped their name on something that's been around for a while. On the flip side; the chemistry seems legitimate and I have found a few papers (not by Shiina) that refer to a Shiina esterification. See also the Shiina macrolactonization reaction; the page has been made by the same editor but the reaction is more prominently mentioned in the literature. --Project Osprey (talk) 16:33, 27 July 2016 (UTC)
Without secondary refs for this specific variant and name applied to it, it does not meet WP:GNG for a stand-alone article. The more general idea of activating a carboxylic acid as a mixed-anhydride, using either an acyl halide or anhydride, for purposes of making esters, amides, etc. is a useful trick that could probably be an article. But it's much older and general than just using these conditions with aryl anhydride reagents. Compare to Yamaguchi esterification, which has secondary refs, an RSC ontology number, etc. DMacks (talk) 17:12, 27 July 2016 (UTC)
The two articles Shiina esterification and Shiina macrolactonization discuss essentially the same reaction, just intermolecular vs intramolecular. I'd be hesitant to just delete them - they are nicely written articles about a technique that is indeed used in organic synthesis. If brought to AFD, the articles would be unlikely to end up being deleted anyway because there is significant precedence for maintaining articles about relatively non-notable reactions on Wikipedia. See the nearly 500 articles in Category:Name reactions, for example. But as DMacks notes, Shiina's reactions are just variations of the older, more well known Yamaguchi esterification and maybe its worth consolidating multiple articles about esterifications/macrolactonizations (or maybe consolidating just the two Shiina articles). Based on my experience, there is almost certainly a COI issue here and the articles/talk pages can be tagged with {{COI}} or {{connected contributor}}. The only real way to fix the potential problem with the COI, unfortunately, is someone taking the time to do a thorough survey of the associated literature and bringing in references to other researcher's contributions. -- Ed (Edgar181) 12:25, 29 July 2016 (UTC)
Latest comment: 8 years ago10 comments5 people in discussion
Although only a few have commented, it seems likely that editors would agree that the obscurely titled Dipolar bond should be renamed Coordinate covalent bond, which is currently a redirect. The practical problem is that a lot of housekeeping would be required to fix several dozen links and minimize double redirects. So my question is, iif there is agreement to do the move, would others help with this housekeeping? --Smokefoot (talk) 14:31, 24 July 2016 (UTC)
Support rename (reverse redirect). I don't think anything else has to be done. Links to the name of the current article would continue to work because it would be left as a redirect to the new location. If it bothers anyone that articles are still displaying the old name, easy enough for AWB or some other script to fix it. And Wikipedia:Double redirects are usually fixed by bots pretty promptly, so likewise don't need any immediate attention. If admin help is needed with the actual move/redirect process, let me know. DMacks (talk) 18:13, 24 July 2016 (UTC)
WikiFactMine: Proposed Wiki Grant for chemical Wikidata
Latest comment: 8 years ago9 comments4 people in discussion
We are proposing a Wikimedia Project Grant
(WikiFactMine) which will automatically scan the daily peer-reviewed scientific literature (up to 10000 articles /day) and extract scientific entities (definable in Wikidata). These, with their citations, are then offered to Wikidata editors for potential inclusion/update. These can also alert Wikipedia editors to new citations. The project includes a Wikimedian in residence in the University of Cambridge. The initial emphasis will be on articles which contain chemistry in a biological or medical context (natural products, drugs, etc.)
We'd be grateful for comments, endorsement and offers of help.Petermr (talk) 10:37, 2 August 2016 (UTC)
Editing and data are good components when introduced well but can be awful when contributed by determined do-gooders without a lot of expertise.
Worrying components in the above message: Who is the royal "we"? Why is it relevant to link University_of_Cambridge - seeking validation, are we?
The relevance of Cambridge is that that we have legal access to extract data from the closed access scholarly literature.Petermr (talk) 14:45, 2 August 2016 (UTC)
Wikipedians in Residence have not proven particularly useful here, IMHO. The chemistry project is just too techy, and the quality of our article portfolio is strong already. If there are particular problems/opportunities that "we" have identified, please speak up specifically. The current community of editors can try to address such gaps without fanfare. We kinda do that already, but informally.--Smokefoot (talk) 12:56, 2 August 2016 (UTC)
The relevance of the Wikimedian In Residence is to work with the library to have access to the closed scholarly literature.Petermr (talk) 14:51, 2 August 2016 (UTC)
Paywalls pose a serious barrier for editors and readers alike. So a Wikipedian in residence (WiR) for that function would be indeed welcome. There are however several sources that editors here have grown to lean on, such as chemical catalogues, ChemSpider, NIST, and such. The greater problem is access to broad information, such as Annual Reviews of ... type things, but in those cases, the WiR could probably not legally pass on pdf's to other editors. Here are some projects for a WiR:
all articles have a general reference.
the top 10,000 compounds (CAS ranking?) are described, and if not the WiR generates a list of which are missing
photographs of compounds.
the spectroscopy articles are often deficient in representative spectra.
About the missing topics, perhaps the keywords or titles could be used to work out where Wikipedia is deficient in a popular substance or process. But there are already plenty of deficiencies, and no shortage of possible things to write about. When I look through title lists of chemical journals I see quite a few topics that we have missing, but the actuall content of the article would have little to contribute, being primary references. New reviews and textbooks would be of much more interest to determine what is notable.
With images we would need freely licensed images, so only a few journals have such free licenses that could have images lifted for inclusion here. However it can be possible to ask authors to supply a different freely licensed image. So if our WIR can ask authors that are likely to have an image of something that we are missing here to supply a picture that would be great. Graeme Bartlett (talk) 23:43, 2 August 2016 (UTC)
What you're suggesting is ambitious to say the least. If I'm reading it correctly, this would almost automate the writing of Wikipedia - the program will collate facts on a concept or compound and all we'd need to do is copyedit them into a new page. It goes without saying that there is lots to go wrong there (I've worked with ontology-data-mining in the past, I could write you a long list). Personally I'd like to see a proof of concept on a much more limited scale. Many of our articles list melting and boiling points but fail to provide a reference. I think this might be a nice place to start; the ontology should be simple (it's pretty much always flagged mpt, bpt) and there is a way of weighting the data (normally the higher the mpt the purer the sample). --Project Osprey (talk) 00:02, 4 August 2016 (UTC)
Category move for discussion
Latest comment: 8 years ago1 comment1 person in discussion
Latest comment: 8 years ago6 comments5 people in discussion
@Cedar101: has been slowly rolling these changes out. In some cases the results are baffling. I don't know what's happening with at Wurtz–Fittig reaction or Bicyclobutane - I think perhaps they're trying to display reaction schemes as text but I think perhaps its not rendering properly for me because I just see lots of seemingly random characters.--Project Osprey (talk) 21:46, 22 August 2016 (UTC)
I take screenshots of the pages to show how it looks like for me: [6] and [7]. It indeed looks kind of weird, and it looks weirder if you don't have the rendering support.
Same problem on a whole bunch of aryl-addition reactions. Perhaps the intent is to have it is alt-text instead of caption, since it is completely redundant to the adjacent image? But ⏣ is not a suitable benzene ring for inline reactions...it would need to be about 5–10x normal size of text in order to be chemically correct. Likewise with other unicode cyclic structures. And they do not align properly when trying to make them bonded to other things. Examples include this cyclopentane, this chlorobenzene that I just undid, and this bicyclobutane that EdChem undid. The fact that it uses uncommon Unicode means it is possibly not suitable for widespread deployment yet (especially as what appears to be an alternative for well-drawn clean images). These edits are making things worse not better at this point. @Cedar101: please stop. DMacks (talk) 15:46, 23 August 2016 (UTC)
Latest comment: 8 years ago3 comments3 people in discussion
Someone has run a script for us to generate a list of name reactions where the ReactionBox is missing one or more of the chemists. It was created for today's Edit-a-thon at ACS since we're working on chemist articles. I thought it would be useful here as a list of chemist articles we (probably) need to create. Walkerma (talk) 13:25, 24 August 2016 (UTC)
Some time ago we had a Wikipedian in Residence and all we got out of it were articles on chemists, chemical societies and medals. Not much progress was made on actual chemistry topics. Now I see the same topics listed here. We have the attention of the ACS community and what do we get? Alternative proposal: read 10 chemistry related articles and correct any mistakes / errors (with citations) V8rik (talk) 19:44, 24 August 2016 (UTC)
I noted that for the first two articles in this list the 'missing' person is not wikilinked. I think that these warrant to be linked, to promote creation of the article (WP:REDLINK: Red links for subjects that should have articles but do not, are not only acceptable, but needed in the articles. They serve as a clear indication of which articles are in need of creation, and encourage it. Do not remove red links unless you are certain that Wikipedia should not have an article on that subject.). --Dirk BeetstraTC10:10, 25 August 2016 (UTC)
![{\displaystyle {\ce {A->[{\text{text above}}][{\text{text below}}]B}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/1c4298e79e43789e1e847e4aba3885e82120b264)
![{\displaystyle {\ce {[RhCl(CO)2]2}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/326332b206858279502d23879597e5aaefc1f14e)
![{\displaystyle {\ce {K3[Fe(CN)6]}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/bb08d7cf071fc731a06221ff3303ff278b0eacd3)
![{\displaystyle {\ce {[Fe(\eta^5-C5H5)2]}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/223bb3e6afc787238e9493f63b468360da2884bf)
/
/
system (like CheMoBot) to mark differences (but at that time we abandon CheMoBot, for now it may be a way of getting some data updated on WD which is there incorrect).
/
![[6]](http://i65.tinypic.com/r242ur.png){kind=link}
![[7]](http://i66.tinypic.com/2dwf4g2.png){kind=link}
