Template talk:Chembox/Archive 10

Poor current citation precedent: I will state again, strongly...
what I did in an earlier, now archived discussion:

'The linking of the infobox as a whole to a list of citations is poor scholarly writing, and entirely un-encyclopedic. The fact that this list is poorly formatted, and a largely URL-only list—the Infobox references —further emphasizes this conclusion, ''but the key issue is with the lack of clear correspondence between facts and sources. '''

There will be—as long this manner of citation remains in place—no way for a reader or editor to easily verify content, or to follow up WP reading with deeper research into the sources of WP content.

The current manner in citing most infobox source material needs to fundamentally change, to a system where each and every infobox entry (field, datum, fact) is directly tied to its single source (in the case of a number), or small set of sources (in the case of more general facts). Alternatively, each infobox field could be listed at the page the reader is taken to by the Infobox references link, and alongside each field listed could be the single source form which the field is filled. If more than one sources are used to fill the field, the statement, "No single source; see article citations." would appear, and a citation at the article would then fully establish the fact's source.

If a reader cannot go straight from a datum/field to a source, we revert to "just trust us" writing, which—however much a current consensus may be raised against this truth—is contrary to WP:VERIFY and WP:ORIGINAL RESEARCH. (If we send interested experts off, repeatedly, on mare's nest / rabbit trail searches that, for practical reasons cannot, in any reasonable length of an editor's time, yield the verified information, then the cited fact is reasonably termed unverifiable, must be seen as the original knowledge/work of an editor, and therefore is a policy violation.)

Le Prof Leprof 7272 (talk) 20:46, 11 August 2015 (UTC)


 * First of all, stop shouting and lecturing. Second, be more specific. There are a large number of available links in this infobox, many of which are self documenting. Adding sources to support each and every link is redundant and not practical.  Wikipedia policy is to use common sense. Boghog (talk) 21:08, 11 August 2015 (UTC)
 * I now realize what you may be referring to is Chemical_infobox. I repeat, you need to be such more specific and concise in your statements. The post at the start of this thread is as clear as mud. I also note that the reference section in question is in project and not article namespace and hence the policies that you quote do not necessarily apply.  But I agree that the sources should be brought in-line. Boghog (talk) 21:20, 11 August 2015 (UTC)
 * Thank you for having the patience to come to understanding, and for stating your agreement that the "the sources should be brought in-line". Otherwise, I deny defenders of the status quo the "bye" that WP policies do not apply in project namespace—WP:V and WP:OR apply to article content, whether the violation occurs via a simple edit or through use of a project-generated template to install unverifiable text—but, but, will (absolutely) not argue this point further with you. Le Prof Leprof 7272 (talk) 04:34, 12 August 2015 (UTC)
 * Based on the link that you failed to provide but Dirk provided below it is now clear that you are not talking about the Chemical_infobox, but about physical constants of chemicals listed in chemboxes. Insisting that each and every value is sourced is not realistic. This data is normally obtained from tertiary sources as ChemSpider that are also linked in the infobox. I would strongly oppose mass deletion or tagging of such data. If this really bothers you, add the citations yourself. Boghog (talk) 11:42, 12 August 2015 (UTC)
 * No, you have arrived now at a wrong conclusion, and are boxing shadows. I am talking both about the status of the Chemical_infobox, and about the inability 100% of time, to be able to say, "Where did the value in the infobox come from?" Both questions are in play, see my response to DePiep, below. Vis-a-vis shadows, no one has proposed mass deletion or mass tagging. (You know I am not shy about this, and I have not tagged a one.) For a summary of what has been proposed —since you are not one to read carefully before replying—look for my short response to DePiep, below. Le Prof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * A reiteration of last time, differently worded: WP:VERIFY requires that data that is likely to be challenged to have references - we do not need to have a reference on everything. Most of the data in infoboxes is of a level that it can be found in tertiary sources - 'grass is green', 'the sky is blue', 'the melting point of water is 0°C' is information that is unlikely to be challenged, it does not need to be sourced, or it can be 'blanket sourced' (and the latter is what the infobox does).  Not every number in the infobox needs that specific a source.  That is not in violation of WP:VERIFY, and it has nothing to do with WP:ORIGINAL RESEARCH (which may be a problem with data on compounds for which there is only one primary source - the people who made the compound and measured a property).  Moreover, we require that information is verifiable, not that you, on the spot, are capable to see the original source (which may be behind paywalls, which may be in an obscure book in a Tibetan library, or where ever - that does not make a fact untrue).  Only if you can reasonably challenge the data (which would require that you come with a proper source showing that you are right in that the data is wrong) and it is not available from the blanket sources that we have, then you make a case for that the data is wrong (but since you have a proper source then, you have all reason to change the number and provide the source - that there was another number shows that the number may be challenged).


 * Hence, there is for a lot of that information no need to have that fundamentally changed. Some data are not that general, and for those it would be good to have them properly sourced, especially if for a specific data you can show that that data is not available from the blanket sources.
 * I can agree that the data in the chemical infoboxes that are not reasonably found in the CRC handbook, online chemical databases (many are linked through the identifiers), the common catalogues, or other tertiary source should have an inline reference, which can be easily added (some field have a specialised 'ref' field, or anyway a possibility to include the &lt;ref>-marks at the end of the data).


 * LeProf 7272 - this is a wiki, if you think that certain data needs a specific source over the 'blanket sources', you can add it yourself. There is no-one here who discourages that addition, nor has anyone ever discouraged that addition, nor has anyone ever removed said references for data that was there.  If that happens/happened, without proper reason, I will revert those edits and possibly block that editor (I have done that, especially if the editor is no).  I will also revert editors who, without proper reason, change data.  I will also revert editors who, without proper reason, remove data that is unsourced.  And if someone finds a reference stating that certain unreferenced data that we have in a field is wrong, I expect them to change the data and supply the proper source.  And I expect that if someone adds data that cannot be easily found in our blanket sources to add a reference where they found that number.  I strongly encourage you to do the same, and I will also not stop you, or anyone, if you add that information on other numbers that you verified (even if there is no requirement on having the reference if it is not likely to be challenged, there is also nothing wrong with having it on there either) - this is a collaborative project.  --Dirk Beetstra T  C 05:42, 12 August 2015 (UTC)
 * (due to ) - I deny your claim that the data is unverifiable if it does not have a reference - data with a reference is verifiable, data without reference is not necessarily unverifiable, and that is a misinterpretation of WP:V. --Dirk Beetstra T  C 05:42, 12 August 2015 (UTC)
 * Apologies, but what on earth is this logic? The reason we provide citations is to allow verifiability. What do you propose, that readers come to wikipedia, then take a fact or datum and go to google.com to guess where that fact or datum came from? You propose they search "oxygen 0.039" or some such to verify the natural abundance of O-17 (then ignore all the WP and mirror sites that rise to the top)? NO. The writer must state the secondary source of their fact or datum. That is the standard for science writing here (poorly followed though it is). By WP standards, data without reference is [INDEED] not verifiable. There is no sane alternative policy or understanding. That a source might exist, somewhere, to support an unsourced point does not make it verifiable, by the WP understanding. Le Prof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * No, one only needs to do that for data where one thinks that it is expected to be untrue. And yes, verifiability is about that, that the data can be verified, and I agree that it is the best that that data is directly verifiable with an attached reference, but it does not mean that data without reference is unverifiable, or that material with a non-linked reference or a 'generic'-type reference is unverifiable.  Something that is available from many tertiary sources online with a simple google search is more verifiable without an attached reference than something that is in a paper copy of a 2-of-a-kind printed book where one copy is in a remote Mongolian monestry, and the third one in the private possession of the Supreme leader of North Korea ..  Again, it is best that there is a linked and properly descibed reference, but it is by all means untrue that everything needs a reference, nor that data without reference is by definition unverifiable.  --Dirk Beetstra T  C 05:37, 25 August 2015 (UTC)


 * Up until this point in the talk flow, this is a reopening of a similar post by as was linked to. Being an editor who edits, I follow this closely. Given the earlier (linked) discussion, right now and here, IMO no new points were added. Even better, the responses here are more than adequate, wrt WP:V, WP:OR, WP:RS. I will not add. Below, I open an "arbitrary" new subsection to allow for an interesting follow-up discussion. -DePiep (talk) 21:15, 13 August 2015 (UTC)
 * I know that these infoboxes are your baby, sir, and that you feel defensive at any suggestion that they need change. But from a professional, formal, chemical perspective, they are a mess of (frequently) disattached individual data, for which the current (atrocious) list of sources is entirely unhelpful. I gave a clear option for change: A detailed reference list for the chemboxes that is (i) thorough, (ii) scholarly—i.e., good, complete, proper secondary sources, (iii) organization of the citations according to the FIELDS in the chembox, and (iv) a policy that if the source is other than the one assigned, an inline citation needs to appear in the box. Please, discuss this proposal. Otherwise, you are entitled to view the matter however you wish, just as others are entitled to ignore your attempt to cut-off discussion. Le Ptof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * - there is hardly anything to discuss - you can reference every single datapoint as you wish. Either through the 'detached' ref-fields, or through simple addition of &lt;ref>erences as you would in prose.  This is not a template-coding problem, this is a behavioral problem (and again, not chembox/drugbox specific, it is throughout chemistry, it is throughout Wikipedia.  (By the way, what would be possible is to have for every field that needs a reference an attached _ref field, and when that is not containing data, the item automagically shows a fact-tag ..).  That would stop the need for mass tagging 'our' 14.000 or so articles - that would be a policy-backed approach).  --Dirk Beetstra T  C 05:37, 25 August 2015 (UTC)

Abitrary break #1
You should have a look at Wikidata. Wikidata has sources for each statements, which should allow to generate proper linking to the source of informations if taken on Wikidata for the Infobox fields. TomT0m (talk) 10:34, 12 August 2015 (UTC)
 * And we could source that at some point .. --Dirk Beetstra T  C 11:06, 12 August 2015 (UTC)
 * Yep, this should be the way to do now ... everything is in place to be able to render references as Wikidata items with arbitrary items. For those who contribute on severals linguistic versions, note that you'll have to source only once in Wikidata and that the source will automitically added in all the projects, when the infoboxes will all be adapted. TomT0m (talk) 12:52, 12 August 2015 (UTC)
 * Wikidata does not have a reference for every statement (but a possibility do add it - like the chembox's). And a of the "references" (on Wikidata) are like "imported from: English wikipiedia". Christian75 (talk) 13:18, 12 August 2015 (UTC)
 * This is the current situation ... but it's best to focus on a project than splitting the efforts. Wikidata is best suitable for automated work when possible (data import from other databases for example, this become sourced by the database) so this can change very quickly, and it's a place for interlanguage cooperation, as it's language agnostic and that the information can be used by any Wikipedia. In short, when adding a new reference now you this will show up in this Wikipedia, and in others. When someone of another language will add a reference, this will show up as well. The imported from  statements are used to trace to the original wikipedia reference if there is one to copy it directly in Wikidata. So everybody wins by focusing on Wikidata. TomT0m (talk) 14:45, 12 August 2015 (UTC)
 * I would like to provide you with some kind of outside view: In de.wikipedia, the parameters that need an inline reference are defined (tick mark or ref in the last column). All articles transcluding the chembox do actually have inline references for all defined parameters. There was quite some work to have this completed some time ago. --Leyo 20:39, 12 August 2015 (UTC)
 * Please, everyone, take note of 's comment, which means both that there is a way forward to accomplish this, and that a sister wiki site has already done so. Thank you to a German colleague for attending to our page and sharing this. Please, those of you arguing the lack of necessity of this, or the absurdity, take note. A significant sister site in Europe, the birthplace of modern chemistry, has already taken this step. Le Prof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * @Beetstra. Just a comment about the fact that everyone can add proper reference to each data of the infobox. We had similar discussion in the past in WP:fr and from that experience don't expect that contributor will do more than what is done usually. People work by mimicry and if the trend is to refer to a general list of references people will do the same.
 * The main question is to know which is the choice of the project: do you want to increase the data quality by offering for each value the source or do you want to stay with the current system ? What is the objective for the future data ?
 * From personal point of view your system isn't the best one because you separate value and source and this is contradictory behaviour for a reference work using quite a lot of different sources. Snipre (talk) 07:51, 13 August 2015 (UTC)
 * , You have hit the very nail on the head, stating that to separate value and source... is contradictory behaviour for a reference work using quite a lot of different sources. I could not have said this better myself, but have often argued that the biggest determinant of the evolution of the quality of an article is its "entry point quality" when significant, active group of editors descend upon it. If they find it full of URL only sources, they follow that pattern. If they find it carefully and rigourously sourced, they understand that, rather, as the standard. Yes, yes, yes. We have never met, but I will offer you my first-borne son or daughter for betrothal (for clarity, and, here at en.WP, originality, of thought). Le Prof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * - I don't expect them to (realistically). I know that even with tagging there is no follow-up on it (and that goes in general).  I agree that an academic standard requires that one actually states upon writing a statement where that statement is from, but that is not how Wikipedia works (actually, that is more about attribution than verifiability).  Most editors however will not include a reference when adding a 'fact' (even when they copy it from another source!).  I know that a set of general references for the 'facts' in the chemboxes/drugboxes is not the best, but it is better than what most editors supply (no references at all, and that is not because the chembox has general references, they don't do it in prose either).  And as I said, verifiability is about the possibility to verify facts that are likely to be challenged, and about the ability to check a fact, not that one has to be able to read a fact and go to the source and double check it on the spot immediately etc.
 * I do think that good editing is 'hmm, I don't know whether is true, it does not have a reference, let me see what I can find' (in line with WP:AGF), and not 'hmm, I don't know whether is true, it does not have a reference, so I better delete this' (this is not a BLP, and even there the latter action is controversial - it took a lot of debate to get to that and still the former action should be preferred). The latter amounts to vandalism when it is applied consistently while the facts are generally true, and (albeit with (maybe significant) effort) verifiable.  --Dirk Beetstra T  C 03:50, 16 August 2015 (UTC)
 * Dirk, I will say again: to allow as the Wikipedia SOP that Most editors however will not include a reference when adding a 'fact' (even when they copy it from another source!).—to accept this as permissible, and to accept as the working WP standard, facts [that] are generally true, and (albeit with (maybe significant) effort) verifiable—to carry on discussions as if we have to abide by this lowest common denominator—is a statement that, while true to the reality of many editors is a choice to both
 * * (i) frustrate and drive away the best of editors, those that do add sources as they edit, and
 * * (ii) defy the principle and spirit of WP:VERIFY.
 * The consequence will be to relegate WP science article an eventual entirely poor state, because they will not improve over time with a "just trust us" policy of science fact presentation. Le Prof Leprof 7272 (talk) 22:39, 18 August 2015 (UTC)
 * I agree, and the facts should be referenced. But what you suggest is resulting in the point where you think that you can find consensus for me and my fellow editors to warn every editor who adds a 'fact' to the chembox without reference to not do that again without providing the reference, and for me and my fellow administrators to block every editor who will not abide (though it is a valid block reason ..).
 * Again, I agree that all should be referenced, and that we should improve on that. But it needs a behavioral change, not a change in policy or guideline - you have all the 'weapons'/'tools' you need in WP:V/WP:NOT/WP:RS etc., Wikipedia 'must' abide by those 'laws' (and not just our chemboxes).  Wikipedia is an ongoing work, and the ratio to 'referencing' to 'adding material' is an increasing number (we have a lot of data already, there is 'less' to be added, people go more towards gnoming).  --Dirk Beetstra T  C 05:37, 25 August 2015 (UTC)
 * Thank you for your comments, Dirk they are thoughtful, and apt. We simply disagree on a fundamental point — whether all factual information must be verifiable (Le Prof), versus whether facts that are disputable must be verifiable (attr. Beetstra). This is a fundamental difference in perspective, and for reasons of practicality, of doing honest scholarly work, and of perceived route to reliability of WP content, I stand by my demand for all. Le Prof Leprof 7272 (talk) 19:31, 10 September 2015 (UTC)

The conflict there is the difference in what we codify in our verifiability policy and a deeper lying problem, people taking values from the internet and not telling where they got it. Some of the data is historical, predating the more firm parts of policies that would preclude how to include it.

Let me be clear, it would be better if everything is properly sourced. But doing so takes a huge community effort, and not this discussion. The community has to do that by itself, setting up task forces for it is rather futile generally. Discussing that it needs to be done is similarly futile - we know it needs to be done and that is policy based, but actually doing it is a second thing. You could address the editors who are adding these numbers but who do not add proper attribution, but that is often resulting in bitey situations (editors rather walk away than doing the extra effort of adding the references ..).

Wikipedia contains a lot of information that needs proper attribution, not only the data in the chemboxes. I mentioned somewhere a number for how many articles contain unreferenced statements (which includes chembox data) which is huge. The real number of unreferenced statements is even larger (many articles contain multiple statements; many statements are untagged), and that needs a lot of research. The community is working on that, but that process is going to take several years, if it ever is going to be completed.

It is not that we are not trying to change what is written in our General disclaimer - but it will still be true for a long time to come. Until then, it is better that we have (albeit with effort) verifiable information that is not attributed (and if you take the effort to actually verify, please include that reference for the future - I will try to do that as well), than no data at all. --Dirk Beetstra T C 03:53, 13 September 2015 (UTC)

Off topic meta cherry picking subject

 * User:Beetstra: "me and my fellow administrators to block every editor who will ...". See. Why discuss at all? Someone & their admin friends will force it their way anyhow. Is Why I Left Wikipedia. -DePiep (talk) 01:01, 10 September 2015 (UTC)
 * Yet again cherry picking out of the same quote .. leaving out the part on the dots, the part that is policy based, trying to put me in bad light. and anywya, '... force it their way anyhow' is wrong, it is not 'their way', it is 'your way', as we all agree on the policies and guidelines that I am trying to keep up (and if we do not agree with the policy, we discuss that there).  --Dirk Beetstra T  C 03:53, 13 September 2015 (UTC)
 * Dirk: you wrote it. You put the cherry in. You use admin's power threats to make your point. Is why I leave. -DePiep (talk) 01:13, 12 October 2015 (UTC)
 * I understand. --Dirk Beetstra T  C 03:49, 12 October 2015 (UTC)

German drug law
I'm not sure if there's a "english speaking countries only" policy for the chembox template, but I thought asking won't hurt:

What about making "legal_GER = Anlage I" (and Anlage II and III) link to Drugs controlled by the German Betäubungsmittelgesetz? Aethyta (talk) 20:29, 1 July 2015 (UTC)
 * It's not a policy, but I'll draw the line for English speaking countries indeed. An improvement on this line could be welcome. But not all 200 countries this way. Maybe there is an alternative, say using iw or wd? -DePiep (talk) 21:08, 1 July 2015 (UTC)
 * Its English-speaking Point Of View. I propose to remove small countrys like New Zealand. Germany should definitely be included with its population of 80 million. (, do you still think it should be included?) Christian75 (talk) 17:21, 7 July 2015 (UTC)
 * Generally yes, although I randomly derped and meant to post this on the drugbox template as I prefer that one. Aethyta (talk) 17:25, 7 July 2015 (UTC)
 * Actually, Christian, it is not WP:EPOV (your link). As that section describes, EPOV denotes a preference for England and its culture as a POV. Maybe even your proposal to exclude NZ is an illustration. But the question is about which countries to include for their law and jurisdiction. This still being an English language WP, choosing English speaking countries is a quite systematic non-POV choice. But as said, there may be a better rule for inclusion/exclusion. Your suggestion to use country size would need working out to be made appicable. RE, sure Drugbox reaches many more drugs articles, but the two templates use the same input-options and data-setup ;-). -DePiep (talk) 22:43, 7 July 2015 (UTC)
 * quote first line from WP:EPOV: "Also be careful to avoid an English-speaking Point of View. Although country-specific and similar POVs are often easy to spot, this can be harder to spot." - choosing English-speaking-countries as a criteria is country-specific and POV. The text are using more time describing culture-specific because, as it says, its harder to spot. Christian75 (talk) 22:54, 7 July 2015 (UTC)
 * Read more that the first line. It is only about the English-spoken history view etc. OTOH, an English-language wikipedia is not country-specific. Anyway, choosing all English-speaking countries is not country specific. The English language in itself is not a POV. (Leaving out NZ is country specific). It is wiki-specific though. Now, what is your better criteria? -DePiep (talk) 23:26, 7 July 2015 (UTC)
 * Would this not be an issue for de.wikipedia? I feel I should point out that they don't use Drugbox in any event. --Project Osprey (talk) 15:25, 8 July 2015 (UTC)
 * The enwiki page provided by has no iw at all... Maybe we are supposed to serve the German-speaking countries too ;-). But seriously. There can be an argument to describe German law by drug (or, say, the law on cocaine in Afghanistan and Colombia). It's just: there are 200 countries. I am interested to learn by what criteria we could select countries for that list. -DePiep (talk) 18:00, 8 July 2015 (UTC)
 * Well, what about countries that have a list of their banned substances on Wikipedia? That would be extremely few. Aethyta (talk) 18:03, 8 July 2015 (UTC)
 * re Project Osprey: Found more. Infobox drug does iw to de:Wikipedia:Formatvorlage Arzneistoff. Looks like a drugbox to me ("Arzneistoffe", nice). Is this an answer? -18:12, 8 July 2015 (UTC)
 * Good suggestion by Aethyta. -DePiep (talk) 18:34, 8 July 2015 (UTC)
 * Who exactly makes such decisions, and when *could* it happen? - Aethyta (talk) 05:53, 27 August 2015 (UTC)
 * The English Wikipedia should have a world-wide perspective. However, right now the infobox covers a few of the 67 sovereign states where English is an official language, and I do not suggest to add all countries in the world. Maybe by size or something less random than what its now. Christian75 (talk) 23:23, 13 July 2015 (UTC)

The English Wikipedia has a funny problem here: it is used by many countries where English is not the primary language, just because the size of en.wikipedia (not necessarily true for de.wikipedia (they will cover most by themselves), but a small-language wikipedia might miss the article on obscure-drug X, whereas the English Wikipedia has. If that person is looking for info, they might turn to the English Wikipedia, and see what it says.  Now, should en.wikipedia have the legal data for a drug in that country (it is certainly encyclopedic what country Y does with drug X)?  --Dirk Beetstra T  C 03:31, 3 September 2015 (UTC)
 * There is generally a "Legal status" section within "Society and culture" that can contain this information for any number of countries. If this becomes excessive a sub-article is made. It is encyclopedic information but it doesn't all need to be in the infobox. We could determine the countries with the highest absolute populations of English speakers and use the top half-dozen or so. I don't know where to find such information. Conveniently we have this: List of countries by English-speaking population. Sizeofint (talk) 04:35, 3 September 2015 (UTC)
 * If we'd want to keep Australia under this criteria we'd have to allow the first fifteen or so. Sizeofint (talk) 04:41, 3 September 2015 (UTC)
 * That sums it up, it was what I meant to be the 'funny problem' here. Table-creep to the max if you include all 200+, 15 is already a lot.  --Dirk Beetstra T  C 13:56, 3 September 2015 (UTC)
 * If you sort by "As first language" you get a list remarkably similar to the one we now have. We'd only need South Africa and the Republic of Ireland to round it out. Alternatively, we could look at the number of visitors to the English Wikipedia by country. There are some statistics at . The top five are the United States, United Kingdom, India, Canada, and Australia. This is also remarkably similar to our current list. I would not be opposed to adding India and removing New Zealand. This latter approach allows us to target the countries with the highest usage of en.wikipedia and make it the most usable. Sizeofint (talk) 15:39, 3 September 2015 (UTC)
 * I still think that it makes more sense to only add new countries that have their drug laws listed on Wikipedia, for easy verification. Germany is #6 by traffic by the way, so not really a obscure little country by any means. Aethyta (talk) 03:21, 2 October 2015 (UTC)
 * You can use "|legal_status =" parameter to fill in German law. If enough articles list Germany it can be formally incorporated. Sizeofint (talk) 06:55, 2 October 2015 (UTC)
 * And then having to change them all again? That's not very efficient at all. If it was "formally incorporated" I'd add everything listed in BtMG. Aethyta (talk) 07:47, 2 October 2015 (UTC)
 * Hmm, looks like we have a chicken-egg problem then :p Sizeofint (talk) 08:05, 2 October 2015 (UTC)
 * I already add the German legal status for all new pages I create (in text, not the infobox). A few days ago I updated over 100 articles with their legal status in China. Seriously, if German was available in the infobox I would quickly update everything within a day! Aethyta (talk) 22:38, 2 October 2015 (UTC)
 * You could copy this into the infobox:  | legal_status = : Anlage I . If we ever implement legal_GER a bot would likely update the infobox. Sizeofint (talk) 20:49, 11 October 2015 (UTC)
 * Thanks, will do that! Aethyta (talk) 21:50, 11 October 2015 (UTC)
 * There are some 200 countries. We can add China and Germany. But why not Russia then? India? That is the point of this talk topic. -DePiep (talk) 21:22, 11 October 2015 (UTC)
 * I only proposed to add countries that have their drug laws listed here on Wikipedia. That does not include Russia, China or India. Aethyta (talk) 21:50, 11 October 2015 (UTC)

There is Drug prohibition law which has a list of countries for which we have drug law articles. It has around twenty countries and roughly half of those actually list the statuses of different substances. Sizeofint (talk) 18:56, 12 October 2015 (UTC)
 * I am not sure if a country having a list of substances by legal status is a good limiting principal however. For now it seems like we're just letting it grow organically based on common non-standard legal statuses. It isn't anything we have to deal with at the moment but in the future we may have to rethink this as more countries are added. Sizeofint (talk) 19:02, 12 October 2015 (UTC)

ChemBox / Drugbox on Mobile view
Dear all, I've not been involved much with the 'back end' of wikipedia. I also posted this over at Infobox_drug

The problem I've found (as others may have too), is that on mobile view there is sometimes up to 6 screen full to scroll through (chembox / drugbox) before getting to the intro of the page. I would have thought that the intro should be first?

Otherwise, if intro is not first for some reason, I wondered if there is a way to have the Chembox / drugbox collapse on mobile view? I searched the archives and found it's not collapsible but found no reason explaining why?

I'm not even sure this is the right place to post this? Thanks, Spannell (talk) 12:35, 4 October 2015 (UTC)
 * Yes you posted on the right place. But it takes a lot of discussion at wiki to improve it. -DePiep (talk) 22:54, 23 October 2015 (UTC)

Proposal: remove Jmol from Chembox
Articles that have Jmol in : (8948; Jmol uses SMILES input always). Technically, Jmol reads a SMILES input and can draw the structure from it (of course, because that is what a SMILES is).

Now Chembox has data row Jmol that like this, eg for Kojic acid:
 * input: SMILES
 * shows: Jmol-3D images =

My point is that this external link does not add info. If we want to show an image, we should add it by commons.

A minor point: Jmol is in section "Identifiers", but of course belongs in the top, "Images". However, today's section-structure does not allow SMILES input to be used in another section. (To show the Jmol link in the top images section, we'd have to repeat the SMILES input there). Once is in Lua, we could do that, but again why should we?

I propose keep concluded, see below to remove the Jmol data and its external link from. -DePiep (talk) 20:40, 1 November 2015 (UTC)


 * Strong keep Because structures can be rotated by the user (see mouse manual), Jmol depictions are far more useful than the static 3D structures that currently displayed in the Chembox and far less annoying than spinning gifs. Please note that the gifs only rotate about one axis where as Jmol allows full user control of rotation about three axis. Boghog (talk) 21:06, 1 November 2015 (UTC)
 * Strong keep per Boghog. -- Bduke   (Discussion)  21:57, 1 November 2015 (UTC)
 * Does it comply with WP:EL? -DePiep (talk) 17:43, 2 November 2015 (UTC)
 * Very easily complies with WP:EL. Per WP:ELPOINTS, the link is in an infobox, not the body of the article. Ideally JMol should be integrated within Wikipedia, but unfortunately it is not (see this discussion and WP:Using Jmol to display molecular models). And finally, JFW said it best, Jmol rocks! Boghog (talk) 20:05, 2 November 2015 (UTC)
 * Well, I'm with nearly three years now, but this is the first time I learned that I can use the mouse to rotate. Says something about our presentation (someone should have proposed to add that mouse-instruction to ). Wiki is not for the initiated.
 * I do think that the infobox is article-body. Maybe later more. In any case, WP:EL says that external links should be in the ==EL== section. I get your enthousiasm, but it breaks wiki-page consistency.
 * In general, of course we need a separate template in the ==EL== section for the dozens of external databases links. (think: near ). -DePiep (talk) 20:52, 2 November 2015 (UTC)
 * Infoboxes aren't article body; WP:EL says "External links should not normally be used in the body of an article. Instead, include appropriate external links in an "External links" section at the end of the article, and in the appropriate location within an infobox, if applicable." Otherwise it wouldnt say place external links in the infobox. Chembox has a lot of external links in the infobox; CAS, Chemspider, Pubchem, etc. Do you want them (re)moved too? Christian75 (talk) 23:02, 2 November 2015 (UTC)
 * #1 : Your quote says: "EL should be in section ==EL==". How do you read that being "in infobox is OK"?. #2 : Yes. As I said: other el's could be elsewhere in the page (what's your question?). Anyway, this is otherstuff. -DePiep (talk) 23:55, 2 November 2015 (UTC)
 * Reiterating what Christian75 wrote above, ... include appropriate external links ... in the appropriate location within an infobox ... – WP:ELPOINTS. Hence this guideline specifically allows external links in infoboxes. The other external links in this and other infoboxes is not "other stuff".  "Other stuff" is cherry picking (i.e., using unrepresentative/unusual examples as precedence). The widespread practice of including external links in infoboxes demonstrates it has community consensus. Boghog (talk) 03:35, 3 November 2015 (UTC)
 * Yes, "otherstuff" is cherry picking. So we better not use that at all. That is: the presence of the CAS link is not an argument for or against the presence of Jmol. -DePiep (talk) 10:27, 3 November 2015 (UTC)


 * Keep, as supplies a unique and useful resource that cannot be included in Wikipedia directly (unless we get Extension:Jmol) and that is in keeping with the EL guidelines that such extlinks are appropriate in infoboxes. If the problem is that users don't know how to use what they find when they go to an external link, then the external site could include some notes on the target page and/or we could include notes in the infobox entry. For example, we call it "Jmol-3D images", so our Jmol page could include a detail early that the rendering is manipulateable in 3D, not just static. DMacks (talk) 23:14, 2 November 2015 (UTC)
 * OK, keep. Now that advantage of interactivity is told (new to me, after I started this post), its presence is more relevant. External links can be in the infobox. Remains the point of being appropriate in there. And the difference with links like the CAS is that Jmol is not defining for the substance, it is illustrating. Hence it has lower relevance. We'll leave Jmol in there, maybe improve the text to note the interactive feature. -DePiep (talk) 10:27, 3 November 2015 (UTC)
 * I added a note to the first sentence of Jmol stating that structures can be manipulated vs just displayed (shocking to me that this wasn't emphasized at all!). Feel free to adjust further. DMacks (talk) 16:15, 3 November 2015 (UTC)

Jmol application in WP?

 * What happened to the idea to integrate Jmol? Was it just never fully implemented or did it face opposition? Sizeofint (talk) 05:48, 3 November 2015 (UTC)
 * Wikipedia would have to support the running of Java applets. As far as I know, this has never been implemented (see for example this discussion). Boghog (talk) 06:20, 3 November 2015 (UTC)
 * J S mol is pure JavaScript. I don't know how that technical/language distinction matters for security purposes (but it does increase portability for end-users). Or, given it would no longer require an applet, perhaps it could be implemented as a centralized or userland .js on wiki rather than having to be a formal WM extension? DMacks (talk) 06:25, 3 November 2015 (UTC)

I have not done anything with Jmol for a long time, but the last thing I did was to help modify a stack of web pages to use Jsmol rather than Jmol. Jsmol uses Javascript rather than Java. At the time this was being strongly recommended as being much safer on web servers than Jmol. As anyone suggested using Jsmol rather than Jmol here on wikipedia? -- Bduke   (Discussion)  16:40, 3 November 2015 (UTC)


 * According to WP:JS, Wikipedia user scripts are written in JavaScript. Your account's common.js file is here. Hence it appears that individuals could install JSmol capability immediately if there is an appropriate script. The disadvantage of this approach is that to use it, you must have a Wikipedia account and you must install the script. But it would be a great way to test it. If it works well, we would be in a strong position to lobby to deploy it more widely. Boghog (talk) 19:48, 3 November 2015 (UTC)

Chembox identifiers
Currently, the "Identifiers" subheading in the Chembox reads "IdentifiersCASnone=0". Can anyone find the issue and fix it? Thanks. ChemNerd (talk) 13:39, 21 December 2015 (UTC)
 * -DePiep (talk) 15:19, 21 December 2015 (UTC)

New GHS warning signs
The template currently lists old EU signs under the title "EU classification". I recommend e.g. adding the word old to it. (No need to remove it because the old chemical containers with the old signs stay with us for decades to come, so both are needed.) Thank you. Palosirkka (talk) 07:16, 18 October 2015 (UTC)
 * What about changing  to  ? --Leyo 15:42, 18 October 2015 (UTC)
 * I don't see a need for an admin here. --Neil N  talk to me 19:50, 21 October 2015 (UTC)
 * Are you looking for a template editor? Sizeofint (talk) 23:53, 21 October 2015 (UTC)
 * , can you specify which template (Drugbox or Chembox), and which subsection or parameter this is about? There are several EU-related data rows. Or maybe you have article examples? -DePiep (talk) 00:29, 22 October 2015 (UTC)
 * Hey DePiep et al. I'm looking for somebody with the powers to edit a template. Sorry I though I was on the talk page of Chembox, how did I ever end up here? Anyways if you look at e.g. article Sulphuric acid, you see the danger related pictograms, the old ones are currently titled "EU classification". Thank you. Palosirkka (talk) 04:32, 22 October 2015 (UTC)
 * So. Current Sulfuric acid has lefthand text linking to Dangerous Substances Directive (67/548/EEC), and saying "EU classification". Together: EU classification. What do you propose/request/want changed? I am not aware of the EU classification bein "old" as in outdated. Link? -DePiep (talk) 22:51, 23 October 2015 (UTC)
 * Its outdated, read Dangerous Substances Directive (67/548/EEC) and press a "relavant" external link, eg. Christian75 (talk) 12:05, 24 October 2015 (UTC)
 * , template talk:chembox redirects here Christian75 (talk) 12:05, 24 October 2015 (UTC)


 * ✅ I've applied Leyo's suggestion, it now says "EU classification (DSD)"; not other changes. If someone can elaborate a setup for the new CLP regulation, we can add that. -DePiep (talk) 13:34, 24 October 2015 (UTC)
 * Found: an overview of changes was described earlier by Leyo and Project Osprey: [|#What_to_do_about_R_and_S_phrases]. It includes considerations for transcribing old number - new number. -DePiep (talk) 06:50, 13 November 2015 (UTC)

Chembox edits (30 October 2015)
I have prepared multiple edits in the Chembox template set, in /sandbox pages. They are 'minor', as in: "won't change the reader's page negatively". Check me, and the testcase like Template:Chembox/testcases (#1--#11; m=mobile). -DePiep (talk) 20:59, 30 October 2015 (UTC)
 * In, every unknown parameter used will be mentioned exensively. But when saved, that message is not shown.
 * Parameter NIOSH_id produced ugly results. Not any more. (eg: DDT has 0174). Using new Chembox NIOSH (set)/formatPocketGuideLink
 * In the footer, set some words central.
 * I'm not seeming any issues in the demos shown (viewing in Firefox). Someone might want to try viewing by mobile as that's know to cause issues. --Project Osprey (talk) 21:23, 30 October 2015 (UTC)
 * Thank you. One can test, and check the . In today's test situation, the header background color is 'gold'. This test color will not be in mainspace (articles). -DePiep (talk) 21:45, 30 October 2015 (UTC)


 * Prepare going live. Remove testsettings. -DePiep (talk) 23:24, 30 October 2015 (UTC)
 * ✅. -DePiep (talk) 12:39, 31 October 2015 (UTC)
 * Resulting in 2500 . -DePiep (talk) 11:47, 1 November 2015 (UTC)
 * Reduced to 119 by now. These are mostly params that should be added, not removed (eg StdInChI1-indexing). Working on this, see below. -DePiep (talk) 06:52, 13 November 2015 (UTC)

LD50: too much detail
Recently we have added data row options like LD50 in the Hazards section. Example right (Note: it shows the sub-block 'lethal amounts' in 'Hazards').

parameter is the bot-maintened and relates to the markers).
 * Development: InChIKey1_Comment will not be added, because too cumbersome to use. Instead, InChI1_Comment will be used to specify the InChI and its key. Like:
 * R S
 * R: InChI=A1-Q-3-Z Key: ABC
 * S: InChI=P3-Y-3-Z Key: KLM
 * Background: this is most common usage of the _Comment parameter, and so is valid for both InChI and Key. See testpage for demo.
 * Note that the _Comment will be a prefix, not a suffix. Later more. -DePiep (talk) 10:31, 13 November 2015 (UTC)
 * Change of plan: no need to add indexes to standard InChI StdInChI2 . Standard InChI's can be entered in regular InChI parameter (like InChI2). No need to use different parameter for that (having the -prefix). This simplifies the  parameter set for the editor. Whether an InChI/Key is standard or not we can detect from the string, and at the moment we do not use that info. Existing parameters StdInChI, StdInChIKey are kept, unchanged. -DePiep (talk) 13:49, 16 November 2015 (UTC)


 * reference in a subheader: sometimes an editor wants to add a ref to the subheader, like "Hazards[1]". I'll add this option for each of the 9 optional subsections Sectionn. -DePiep (talk) 17:42, 1 November 2015 (UTC)
 * Prepared: added references for Chembox headers Chembox_ref, Name_ref in {the main Chembox, and by section names
 * Identifiers_ref, Properties_ref, Structure_ref, Thermochemistry_ref, Explosive_ref, Pharmachemistry_ref, Hazards_ref, Related_ref
 * See Template:Chembox/testcases3. &lt;ref> tags must be added: {{para|Hazard_ref|&lt;ref>{www.example.com&lt;/ref>}}. -DePiep (talk) 22:28, 7 November 2015 (UTC)


 * The CAS Registry Number label. Today the label (=the left-hand text) says "CAS Registry Number". That is the correct and formal name. However, we do not need to write the definition in there. I plan to change it to "CAS Number". That is not confusing at all, it is commonly known (recognized at 1st reading), it and has the advantage that it usually takes just one line of text, not two. -DePiep (talk) 18:03, 1 November 2015 (UTC)
 * Prepared. -DePiep (talk) 18:25, 10 November 2015 (UTC)


 * Drop (rely on the unk param cat). -DePiep (talk) 07:53, 9 November 2015 (UTC)
 * Prepared. DePiep (talk) 18:25, 10 November 2015 (UTC)


 * Improve layout for collapsible lists InChI and SMILES. Now width is used better (resulting in, on average, about 1 in 5 lines shorter).
 * Prepared. See testcases. -DePiep (talk) 18:25, 10 November 2015 (UTC)


 * "What is ✅?" in footer: improve logic (show more often, and only show 'verify' (version check) when version id is available. -DePiep (talk) 10:16, 15 November 2015 (UTC)


 * Preparing to go live. Testcases invalid now. -DePiep (talk) 15:22, 16 November 2015 (UTC)


 * ✅. Emptying the maintenance category (started with ~500 articles). DePiep (talk) 19:46, 16 November 2015 (UTC)

InChI: let's use as prefix, not suffix (specifier)
Parameter InChI has option InChI_Comment, that is added as suffix unedited (a space and the comment is written unedited right after the InChI string).

I propose to move this _Comment to the prefix place, before any InChI data. This because usually the _Comment is a specifier for variants within the article's substance (think R and S). Usually these variants are entered using indexes (like InChI1, InChI2 for R and S). This is a change of habit in InChI input, because you can not enter anytext freely anymore: it will be in a format (e.g., better not add brackets any more).

Demo:


 * InChI1 = 1S/C5H9NO2/c7-5(8)4-2-1-3-6-4/h4,6H,1-3H2,(H,7,8)
 * InChIKey1 = ONIBWKKTOPOVIA-UHFFFAOYSA-N
 * InChI1_Comment = (R)
 * InChI2 = 1/C5H9NO2/c7-5(8)4-2-1-3-6-4/h4,6H,1-3H2,(H,7,8)
 * InChIKey2 = AZIBWKKTOPOXYZ-OHFFFAOABC-A
 * InChI2_Comment = (S)

Side notes: -DePiep (talk) 19:20, 14 November 2015 (UTC)
 * The same change will be applied to SMILES_Comment: prefix not suffix.
 * The typographic indenting thing is apart: that is just a an improvement, not the topic here.
 * In the future, we can have those variants entered only once (not for every identifier).
 * In InChI and SMILES, I don't think free comment text (other than specifying the sub-substance) is useful or needed.
 * More demos in this tespage.
 * Some 8000 articles have an InChI, but I expect only some hundred s to be affected.
 * Preparing to go live. Testcases invalid now. -DePiep (talk) 15:22, 16 November 2015 (UTC)
 * ✅ -DePiep (talk) 19:46, 16 November 2015 (UTC)

Drugbank and ATC positioning
Today, Drugbank and ATC code can be entered in two sections: Identifiers and Pharmacology. I propose to allow for one place only, mainly for consistency (regular showing). That would be Drugbank in Identifiers (419 P), and ATC code in Pharmacology (435 P). btw, ATC is not an identifier but a class. -DePiep (talk) 12:03, 20 November 2015 (UTC)


 * Deprecated in section. . No changes for the Reader so far. Some 600 P at first counting. -DePiep (talk) 00:53, 30 November 2015 (UTC)
 * In the process, I order all sections in a single order (as numbered in template documentation). Eg, is in Section5 by standard. This makes  presentation more stable over pages. Actually, this is what almost all infoboxes do. -DePiep (talk) 15:25, 30 November 2015 (UTC)
 * ✅. Cleaned up. -DePiep (talk) 22:05, 30 November 2015 (UTC)

No CAS RN assigned
Discussion at Wikipedia_talk:WikiProject_Chemicals. -DePiep (talk) 14:02, 16 November 2015 (UTC)
 * ✅. Options now in : none, none, none.
 * -DePiep (talk) 01:14, 5 December 2015 (UTC)
 * -DePiep (talk) 01:14, 5 December 2015 (UTC)
 * -DePiep (talk) 01:14, 5 December 2015 (UTC)
 * -DePiep (talk) 01:14, 5 December 2015 (UTC)

Let's add "E number"
Propose to add data row E number. See anoxomer. -DePiep (talk) 23:06, 28 November 2015 (UTC)
 * Seems reasonable. Sizeofint (talk) 19:01, 30 November 2015 (UTC)
 * ✅. E_number and E_number_Comment. Do not confuse with EC_Number. -DePiep (talk) 01:05, 5 December 2015 (UTC)
 * And remember its often used as a synonym too (i.e. OtherNames). Christian75 (talk) 10:03, 5 December 2015 (UTC)


 * 304 articles:


 * Same in Drugbox. -DePiep (talk)

Maintenance categories regrouping
I have created:

They only contain subcategories, not articles.

This is to create more overview for these topics, especially in and. Now it is organized by topic (such as CAS RN), not by maintaining template like ), not by issue ('missing', 'not assigned'). Probably more identifiers to follow. -DePiep (talk) 16:20, 13 December 2015 (UTC)

Chembox changes (15 December 2015)
Preparing:
 * Moved InChI and SMILES are to be positioned at teh end of their Identifiers box. This is because they are non-human readible (machine only by intention), so they should not disturb the regular identifiers listing.
 * For InChI and SMILES (+ all their indexes & supports), the font-size and line-height are reduced. All in all they take less vertical space (minus 20% in situations).
 * Both changes are especially better in mobile view, which does not collapse their boxes (no show/hide). So for the reader this takes less vertical space.
 * PubChem external link will be updated into https and directly to the right target page (by request). -DePiep (talk) 18:52, 15 December 2015 (UTC)

Chembox changes, chemical formula (17 December 2015)
Preparing: improved presentation of the chemical formula (section Properties). See also Testcases9.
 * Add Formula_Charge to add charge. Will show after any number, not exactly above: O42+
 * Into empirical form (IUPAC Red Book 2005, section 4.2)
 * Order of elements is full Hill notation: when any C present -> CH in front, then other symbols by abc. When no C present: all elements by abc. (bug fix) (Red Book)
 * Not changed:
 * - Applied rule 'no space before a ref'. ie, no space between formula and ref note: O4[1]
 * - Molar mass calculation
 * - Showing input for Formula (will show unedited ie as a fixed text string)

-DePiep (talk) 13:36, 17 December 2015 (UTC)
 * ✅ Same subtemplate used in Drugbox, see Drugbox talk. -DePiep (talk) 14:59, 17 December 2015 (UTC)
 * I have no problem with the way its introduced in the chembox template (The Hill notation for formulas). But do not change the inorganic compounds articles to use the new system. There isnt consensus to use the Hill system for inorganic compounds. Christian75 (talk) 08:02, 18 December 2015 (UTC)
 * Hill notation explicitly covers inorganics (as in "when no C is present ..."). And 'no consensus' (in RL I understand) does not say there is any other preferred standard in this. ie does not answer the need. btw, Hill was applied before too, but crippled (some C & H logic misfired). -DePiep (talk) 09:38, 18 December 2015 (UTC)
 * The Red Book says (something like): "When no other rules are present, use Hill". Also: "it helps standardised searching and comparing". -DePiep (talk) 09:38, 18 December 2015 (UTC)
 * No; there is no consensus on Wikipedia to use Hill notation for inorganic compounds in the infoboxes. Christian75 (talk) 10:08, 18 December 2015 (UTC)
 * You mean to say that the no-consensus is at wp only, not in RL science? -DePiep (talk) 10:37, 18 December 2015 (UTC)
 * Do you have any sources which say the chemical formula should be written with the Hill notation? (In real life (I guess) or at Wikipedia) Christian75 (talk) 10:47, 18 December 2015 (UTC)
 * BTW. Does your proposal (at the top) say that you wanted to change the chemical formulas into empirical formulas? If so, it isnt very clear, and I oppose that. Christian75 (talk) 10:50, 18 December 2015 (UTC)
 * I don't get where you are heading. What is unclear in my OP? What is the answer to my latest question? -DePiep (talk) 13:57, 18 December 2015 (UTC)
 * Im saying that its fine you made the change, but the infoboxes in the articles for inorganic compounds (and some organic ionic compounds) shouldnt be changed too (like you did with a lot of drugs yesterday - which was fine - because it was for some big chunky organic compounds). An example: I think it would be a bad idea to replace |Formula = NaNO2 with |Na = 1 |N = 1|O=2 for sodium nitrite because the formula would be instead of  in the infobox. And at least it should be discussed first. And why do you want a reference? A reference for what? Christian75 (talk) 14:46, 18 December 2015 (UTC)
 * The only changes are: 1. fix bug in composed chemical formula, and 2. add option charge= to chembox. The rest is unchanged.
 * In writing empirical chemical formulae (those easiest ones), I still see no other standard than Hill aka IUPAC Red Book 2005 section 4.2 (RS). If people want to deviate from Hill (as your example does), we need source or ref for that. So far, no issue for me -- I only know Hill. Red Book 2005
 * -DePiep (talk) 19:35, 18 December 2015 (UTC)
 * Your reference doesnt use the Hill system. Section 4.2.1 (Empirical formulas) (which says the alphabetical way should only be used if structural structure is unknown) and says see section 4.4 (ordering principles) which state in 4.4.2.2 (alphanumerical order):
 * Quote from IUPAC, NOMENCLATURE OF INORGANIC CHEMISTRY "red book", page 58, html link:
 * "[...]The group NH4 is often treated as a single symbol and so listed after Na, for example.
 * To exemplify, the order of citation of some nitrogen- and sodium-containing entities is:
 * N(3-), NH2-, NH3, NO2-, NO3-, N2O2(3-), N3-, Na, NaCl, NH4Cl
 * Such ordering may be applied to entire formulae in indexes and registries etc., but may also be used for ordering parts of a given formula, sometimes in connection with the ordering principle of Section IR-4.4.2.1, as decribed below for various specific classes of compounds and ions"
 * end of quote. But remember this is for inorganic chemistry only. So I do not think I have objection agains the red book. Christian75 (talk) 22:14, 18 December 2015 (UTC)
 * What IR-4.2.1 describes is the same as Hill notation. Why do you say it is not?
 * You keep mentioning your own thoughts re in/organic. By what rule do you claim what exception for inorganics? (er, you only mention REd Book after I linked it. Does this mean you did not look for it before, after I mentioned it half a dozen times?) -DePiep (talk) 22:25, 18 December 2015 (UTC)
 * Rule 4.2 doesnt describe just the Hill system. It describes more than one way to do it. Rule 4.2.1 could (maybe) be used if the ordering criterion is "structural information" and not the alphabeth. But yes; Im against using empirical formulas for "simple" inorganic compounds in the infobox, and especially the alphabetic way, e.g. instead of . A lot of students etc. copy paste from the infobox and they would get "wrong" formulas. Consensus has been not to use the Hill notation, see  Christian75 (talk) 22:49, 18 December 2015 (UTC)
 * (And yes, I didnt look it up - I know the Hill system, and I know the red book. And I think we should use the Red book too - but not empirical formulas. But since you mentioned that book again and again I looked it up and realized that the paragraph (4.2.1 not 4.2) mentioned more than one way to write emipical formulas based on the alphabeth or structural infomation). Christian75 (talk) 22:55, 18 December 2015 (UTC)
 * "Your reference [Red Book] doesnt use the Hill system" you wrote above. Yes it does. It is the very first subsection after the "4.2" title. And the word "empirical" is right in the second sentence of my OP here. You have not pointed to a diff between the IUPAC description and Hill description. You are just trying to save yourself. A weird way of reasoning, and also abusing my patience.
 * Ok it also use the Hill notation in the absence of any other ordering criterion. Hill notation: If it contains C: then C, H and alphabetically. If no C's, just alphabetically. Agree?
 * Section 4.2.1 in the Red book: "For the order of citation of symbols in formulae, see Section IR-4.4, but, in the absence of any other ordering criterion (for example, if little structural information is available), the alphabetical order of atomic symbols should be used in an empirical formula, except that in carbon-containing compounds, C and H are usually cited first and second, respectively."
 * Not the same. (Look especially at in the absence of Christian75 (talk) 00:42, 19 December 2015 (UTC)
 * Meanwhile I have found your answer. As is unchanged, empirical formula can be used if no other rule is available. Well, when the structural chem formula is known, we can use that one. This is what you were pointing at. Of course we don't downgrade to the empirical form, and no one is proposing that. Has nothing to do with in/organic, also valid for Sodium bicarbonate. So I might have made some errors wrt this (turning structure into empirical). A small price for such a great improvement in correct chemical formula writing in 10k articles. -DePiep (talk) 00:15, 19 December 2015 (UTC)
 * Thanks. As I said "I have no problem with the way its introduced in the chembox template (The Hill notation for formulas). But do not change the inorganic compounds articles to use the new system.". I said it because you changed all the organic compounds (which I think is very nice) to use the |C=|H= notation. And I wanted to let you know before started a hard nights work changing the inorganic too. Christian75 (talk) 00:51, 19 December 2015 (UTC)
 * All clear then. What's left (message to myself): do not edit Chmembox/Drugbox articles downgrading a structural formula into empirical. Can do. -DePiep (talk) 09:00, 19 December 2015 (UTC)

Chembox changes (21 December 2015)

 * Added:
 * Removed from : (useful for  only)
 * Renamed: &rarr;


 * ✅ -DePiep (talk) 01:35, 21 December 2015 (UTC)

SpecRotation: chiral rotation vs. specific rotation
Is it normal to call specific rotation ([α]D) for chiral rotation? Is it a good idea to make it more general (temperature and not only the sodium line)? (Template:Chembox SpecRotation) Christian75 (talk) 18:49, 22 December 2015 (UTC)
 * Please elaborate. Editors would be helped if they do not need to enter formatted units, so giving a few standard-options would be welcome. -DePiep (talk) 07:54, 27 December 2015 (UTC)

add Triple point
asked (on my talkpage) to have added Triple point to.

It should be like T = −56.6 °C and p = 5.1 atm for CO2. Could be TriplePtC, TriplePtP then (into 3 temperatures; p=atm always?). Any comments? Placement in section Properties? -DePiep (talk) 17:28, 27 January 2016 (UTC)

Racemic mixture parameter
About Racemic mixtures (aka Racemate). I see a lot of and  articles are about such a mixture (example links below). The indication usually is scattered in data entries (like: S), and probably more often in the article text only.


 * Examples in chembox: Cysteine Alanine Glutamic acid Tartaric acid Lipoic acid
 * Examples in drugbox: Amphetamine Clenbuterol and this cat, from I (currently ~40)
 * Examples in article-text only: (pls request)

I consider/suggest we add a parameter Racemic to the two infoboxes to allow this description in a stable way, manner & place. First idea:
 * Improved proposal demo (22 Dec 2015)
 * 1:1 racemate RS -- (free text input)
 * racemate

&rarr;


 * {| class=wikitable


 * colspan=2 style="background:#eee;" |Properties/Chemistry
 * Formula || (as is)
 * Molecular mass || (as is)
 * Chirality
 * Racemate RS
 * }
 * Molecular mass || (as is)
 * Chirality
 * Racemate RS
 * }
 * }

Comments? (general, parameter name, lefthand label text, position, ...). -DePiep (talk) 09:25, 19 December 2015 (UTC)


 * I question if this is really necessary as it duplicates information that should already be specified in the chemical name [e.g., (R), (S), or (RS)]. Complicating matters is that the scope of many of these chemical articles is about both stereoisomers as well as the racemic mixture (for example, alanine) and the chembox contains information about all three. If such a parameter is included, it should probably be named enantiomeric_excess or enantiomeric_ratio as this one parameter could be used to specify that the substance is either racemic, optically enriched, or optically pure. Boghog (talk) 11:19, 19 December 2015 (UTC)
 * A rare example of an article about a optically enriched mixture is Adderall. Boghog (talk) 11:25, 19 December 2015 (UTC)
 * I'll have to study this... ;-) -DePiep (talk) 12:56, 19 December 2015 (UTC)
 * FYI :-) - Stereochemistry is a really really diffucult subject (for chemist too). Enantiomers (mirror images) have the same chemical and physical* properties unless they are in a chiral environment ("enantiopure enviornment"). R and S in the IUPAC name tells if the carbon atom is "right" or "left". If you have one compound e.g. D-glucose ((2R,3S,4R,5R)-2,3,4,5,6-pentahydroxyhexanal) and change all R and S, then you have the enantiomer: ((2S,3R,4S,5S)-2,3,4,5,6-pentahydroxyhexanal) (L-glusose). If you change some of them, you have a completely different compound (a diasteromer). (E.g. (2S,3S,4R,5R)-2,3,4,5,6-pentahydroxyhexanal is D-mannose). Thaladomide is a classic example of how dangers different enantiomers can be. (n.b.: If you change all R and S and get the same molecule, then you doesnt have a enantiomer but a meso compound. *The enantiomers rotate plane polarised light in opposite directions. Christian75 (talk) 14:33, 19 December 2015 (UTC)
 * User:Boghog, I think its more commons then rare. I thought all new drugs were chiral, but I found this which states 50% of all new drugs (in 2006) are chiral compounds (but normally only one of the enantiomers are notable). I found this short "funny" list enantiopure drug. Chiral proteins and saccharides often have their own artilce too, like glucose and L-glucose. I just checked a random amino acid, serine, and that article should be split (the text is a mixture of L and D-serine). Christian75 (talk) 14:33, 19 December 2015 (UTC)
 * Just to be clear, both racemic and enantiopure drugs are very common. What is unusual are enantioenriched ("pure, mostly" ;-) drugs.  EJ Ariëns  was a strong proponent in developing enantiopure drugs and he called the inactive enantiomer "isomeric ballast" that carried with it unnessary increased risk of side effects.  Nevertheless, some drugs racemize quickly in solution, so even if only one enantiomer has biological activity, there is no point in manufacturing the optically pure drug because it will racemize once the patient takes it. Finally all proteins, except a very few man made ones (see for example racemic crystallography), are optically pure. Boghog (talk) 15:22, 19 December 2015 (UTC)


 * OK, learned more than needed for this infobox question. Let me refine/rephrase the proposal.
 * I understand the topic is chirality and enantiomeric_excess (won't focus on correct wording for now). Racemic is a common 50/50 occurrance, 100%/0% exists too and rare are other ee percentages.
 * Chemicals with these property indeed have the indicating prefixes (R, S, RS D, L, +/-). That should do, as Boghog says. But, that is only enough for those who know what this is about. An (RS)-... name only says "chiral" for those who already know. I think that is not enough. I myself have been working with chemicals here a few years, but I never got that point. So I think it is good to add that info explicitly. To illustrate its importance: the opening paragraph of the given example Adderall is about this topic.
 * We can aim for standardised ee input (one or two percentages, racemic link) or free text.
 * Main question: can you all agree that adding this datarow is useful? Exact texts etc. TBD later. -DePiep (talk) 23:38, 21 December 2015 (UTC)
 * I am still ambivalent about this because the stereochemistry should already be defined in the chemical name (and in the SMILES string for that matter). I suppose it would not hurt to add a chiral parameters with allowed values: achiral, meso, racemic, enantioenriched, and enantiopure.  One also might consider a optical_activity parameter with two allow values: dextrorotatory (+) and levorotatory (−). Boghog (talk) 13:54, 22 December 2015 (UTC)
 * I see that there already is a SpecRotation (specific rotation) parameter, so that there is no need for a seperate optical_activity parameter. Boghog (talk) 14:08, 22 December 2015 (UTC)
 * Today some 25 articles have a SpecRotqation ({Chembox} only). Unless addressed here, I have no intention to change that data point. -DePiep (talk) 07:52, 27 December 2015 (UTC)
 * re :
 * - I take the "would do no harm" reply to go sandboxing with this ;-)
 * - Again, I think 'it is already in the name' is both correct and not enough. For example, the name can not specify the percentage. And mainly, of course the name does not mention 'racemic' nor link it. So even for the half-layman (like me), this does not give a clue. There are also the non-related name prefixes like trans-, so that adds to the invisiblility. A childish comparision point is (childish, I admit), that the chem formula and molar weight are already "present" in the structural image, so could be omitted too. Finally, editors quite often have tried to add this info to the infobox title (see Bromadoline, and ~40 todaty in cat, from I), which suggests is is important enough to add to the infobox. After all, there are two chemicals present.
 * - Related SpecRotation doesn't need edits (afaik). It has label Chiral rotation ([α]D)
 * - Label to be Chirality
 * - Input options provided here by Boghog best be recognised & wikilinked by the templates.
 * -DePiep (talk) 07:52, 27 December 2015 (UTC)


 * In Infobox drug: added chirality. Near specific_rotation. -DePiep (talk) 17:11, 29 January 2016 (UTC)
 * What should that contain; yes or no or? Christian75 (talk) 17:15, 29 January 2016 (UTC)
 * Can contain any input that relates to the topic of chirality. -DePiep (talk) 17:31, 29 January 2016 (UTC)

Vertical image placement
To consider: vertical alignment in L,R images to center? See Chromate and dichromate. -DePiep (talk) 12:31, 10 January 2016 (UTC)
 * No objections. However; if the images are different sizes would ImageName give text at two different heights? --Project Osprey (talk) 10:41, 6 February 2016 (UTC)
 * Already happens, see testcases7 ;-). Is a bug, must be fixed anyway. -DePiep (talk) 11:20, 6 February 2016 (UTC)

Add codons?
Add codons? -DePiep (talk) 22:40, 28 January 2016 (UTC)
 * There won't be over 50 uses, but Arginine sais it's quite big when it is used. (and add new section 'Biochemistry'?) -DePiep (talk) 11:25, 6 February 2016 (UTC)

Establishing standard element ordering for chemical formulas
Hi, I recently got a suggestion from an anonymous editor, (here), to change the chemical formula of Silicon Carbide from CSi to SiC. Some quick checks show that both these are in use, and further, there's no consistency in how ordering is handled in infoboxes.(see my talkpage) I propose we use Hill system ordering in all chemical templates, and conventional ordering in prose/reactions. If there's a style guide already in place, it should get a brief note on the chembox template parameters list I think. Forbes72 (talk) 23:08, 19 February 2016 (UTC)
 * For Chembox and Infobox drug, there is this:
 * By formula: When entered by parameter Formula/chemical_formula, the formula is shown as entered. This could be entered using Chem, like Formula
 * By elements: When entered by element like etc, Hill notation is applied.
 * When both are entered, the formula is shown not the Hill notation.


 * Background: Notations forms (including Hill) are described by IUPAC: IUPAC Red Book 2005, section 4.2 (pp.66/377 pdf). A discussion was in /Archive 10).
 * Note that Hill notation is a simple form only, called the empirical formula. However, a chemical formula can be more complicated having more information: 'molecular formula and structural formula. The extra information can be groups, brackets (square, curly), ion separation, repetition of groups by n, etc. The template can only automate Hill, for structural/molecular formulae the editor should work that her/himselves at input. Quite often, Chem is used. So far for template assistance.
 * It is common to add the extra info when relevant, and so editors are entering the more meaningful formulae. As IUPAC writes: "The choice of formula depends on the chemical context". In this question, I read all this as: do not impose Hill notation if a better formula is available. Once a formula is made, it better should be used all over the page (so better not changing in the infobox, as you propose). -DePiep (talk) 13:51, 21 February 2016 (UTC)
 * It may be more complicated. IUPAC Red Book writes that "structural formulae" can be 1-D form (single line, as is the topic here), but also 2D graphs are "structural formulae". OK. Next, there is section Structural_formula for these. -DePiep (talk) 08:23, 22 February 2016 (UTC)
 * Thanks for pointing to the prior discussion. So let me get this straight: the ordering is determined by on a case-by-case basis, according to what is the most informative format. Hill notation is not imposed as a standard, but serves as a fallback when other options aren't obvious. According to this, it seems best for me to standarize AxCy ordering on inorganic carbides. Is this correct? Forbes72 (talk) 00:48, 26 February 2016 (UTC)
 * All correct yes. And a good style form to use the same notation type over the whole article and over the whole set of carbide articles. (I assume writing AxCy and ZxCy is meaningful in the context of those inorganics; that is not my field of knowledge). -DePiep (talk) 10:57, 27 February 2016 (UTC)
 * An empirical formula isnt what you describe. The emprical formula is IUPAC Gold book, also Red book p. 54: "Formed by juxtaposition of the atomic symbols with their appropriate subscripts to give the simplest possible formula expressing the composition of a compound." i.e. the simplest whole-number ratio of atoms in a compound. E.g. the empirical formula of benzene: Chemical formula, but empirical formula is CH. Christian75 (talk) 16:25, 3 March 2016 (UTC)

For the chembox (and I believe for the drugbox), the preferred way to enter the molecular formula is by using the parameters 'C', 'H', 'N', 'Li', etc. That results in the Hill notation formula in the chembox (as well as other features). If editors or the situation require that another notation is used (nonstoichiometric compounds, separation of parts or whatever) then that can be overridden by using the Formula/Chemical Formula .. In other words, the only reason to use these fields is because there is a reason to circumvent the Hill notation. --Dirk Beetstra T C 12:53, 27 February 2016 (UTC)
 * Is what I tried to say, with one subtle difference: you say use Hill input unless otherwise required, while I say: adding exta info in the formula can be preferred (as editors choice). Should we clear our difference out, and into a guideline/MOS/doc thing? (Details: and yes, handles this the same; also, input by element numbering calculates & shows the molar mass straightforward - unless overwritten by molecular mass). -DePiep (talk) 14:06, 27 February 2016 (UTC)
 * I think it's worth building a consensus for a style guideline. Specifically, in what contexts Hill is "preferred" or just a "fallback option". These may be very similar in practice. I slightly prefer not enforcing Hill strictly, but either way would work, as long as the end result is consistent and clear. Forbes72 (talk) 17:48, 27 February 2016 (UTC)
 * My !vote would be that in the chem/drugbox Hill should be used where reasonable, for the rest of the text maybe the more common or more clear forms. I know that that may conflict (since we say 'silicon carbide' it is more clear to consider that that is SiC and not CSi), but for almost everything using Hill at least in one standardized part of the page (and the infobox is the best place then) results in Google being able to find when s.o. is looking for the chemical by Hill (C2H6OS finds Wikipedia, Google suggests the same when looking for C2H6SO; forcing Google to look for C2H6SO only does not find Wikipedia - for the rest of the text, Me2SO or similar is likely more clear to the reader).  So in that way, I would consider to enforce Hill in the infobox except for strange/extraordinary situations, and for the rest leave it to editor discretion.  --Dirk Beetstra T  C 12:47, 28 February 2016 (UTC)
 * Like the main line of this: we can do Hill in the infobox, and structures inline. However, I don't like the argument that we should downgrade to Hill because of search engines. Maybe we should hide such intended search-hits in the template (findable by google bots, but not visible). Isn't there a solution in this: we need a strong statement from RL inorganic chemistry: "In inorganic chemistry, we do not use Hill". In this, inorganic = without the pair C,H.
 * (Details: 10k Chemboxes + 6k Drugbox = 16k, 5k of these have formula entered; see maintenance cats. I need to research: I mentioned Red Book, but there also is Blue Book, about in/organic. And: InChI has the simplified formula. Is that Hill always? ). -DePiep (talk) 19:32, 28 February 2016 (UTC)


 * If changed to the Hill notation, I would prefer we change the "chemical formula" (in the infobox) to e.g. Hill formula. For most people (or chemist at least) ClNa, CH2O3, CHN and yes, CSi looks strange. In matter of fact I have to think before I can see its NaCl, H2CO3, HCN and SiC. Most of the organic compounds already use the Hill notation, but not small organic compounds like methanol, formic acid etc. My guess is that people students copy/paste from the infobox (I do), and "wrong" formulas are not good - the teacher will blame Wikipieda.... Christian75 (talk) 20:55, 28 February 2016 (UTC)
 * I thought that the C, H, N, .. input already did Hill for .. years. And I don't think that that is a 'downgrade' - it is a choice of system.  And to me, the chembox is for systematic display of data, not real prose.  And I do think that since that system is one of the ways that chemicals get standardized, it is a likely way that people can find chemicals (but invisible tags are a solution as well, displaying something else).  The RL statement 'in inorganic chemistry, we do not use Hill' is probably true (Sigma Aldrich talks about SiC, CB4, Si3N4, ...).  Hmmm ... but if we want to make the formula from the C, H, N system, then we need some order, and even complete alphabetic or by element number we get 'CSi', 'B4C', 'N4Si3')... or we have to abandon the whole C, H, N system, using a complex (LUA) module to extract the Hill formula, molecular weight etc. (so we can use Me2SO for DMSO, which is likely what most people recognise as DMSO).
 * That teacher should be fired, first of all the student is not wrong and secondly, the teacher should teach the student to think for themselves and not blindly copy from the internet/wikipedia. --Dirk Beetstra T  C 03:25, 29 February 2016 (UTC)
 * re per-symbol-input "thought ... did Hill for .. years" - so did I ;-). But recently I reprogrammed that piece of template, and it appeared that there was a hiccup (to do with ordering of H, IIRC). It now applies the Red Book rules, as linked, for empirical formula, ordered by Hill rule (ask me to tell more).
 * re "downgrade". As mentioned (token from the Red Book again), there is empirical, molecular, and structural formula inline notation (i.e. not graphical/2D). Molecular and structural add extra information (for example, writing a OH group separated). Removing that extra info is downgrading (in the example lumping all O's and H's together).
 * I am a bit reluctant to write the chemicals in such a way "people can find" them. First and foremost, we write them correct. Search help may be added, but only in that order (don't want to end up spelling like "C24H38O4"). Sure the infobox is not 'prose' (when I wrote 'inline', I meant to say: 'in the article text, sentences'). But it should be correct still.
 * "... or we have to abandon the whole C, H, N system" -- not that soon. First we better describe the mechanisms (rules) that are used in RL chemistry, possibly with the in/organic distinction. And when a molecular or structural formula is known, there is no need to fall back on empirical (be it by Hill or full alphabetical). Would take some research-time I currently do not have. I think the question we are looking to answer is: what is the rule in inorganics? -DePiep (talk) 08:27, 1 March 2016 (UTC)
 * I think that one of the points I am trying to make, is that there is also value in having the CxHyOz-based system (i.e., a plain count of each element in the system). In the prose that makes reading less clear, but in the infobox there is no harm - the more written out forms are/should be in the prose anyway (I would strongly object against using C2H6OS in the prose - you don't know if it is mercaptoethanol or DMSO .. C2H6OS is however nice for people who want to, say, calculate isotope distributions etc. as they do not have to add up the Cs and Hs to get to the result).  I also think that something systematic needs to be encoded (whether visible or hidden) that allows finding chemicals by the formula - we are writing here for the reader, but if the reader can't find what they are looking for it is not of much use.
 * I however guess that I don't really care in which order they are, but applying different systems depending on origin is going to be difficult there, you can't say 'it is inorganic, so the C comes after all metallic elements' or something similar (and there are always borderline inorganics/organics). What order does our dear IUPAC suggest for formulas which are stripped of all structural information (as that is what we are thinking of here).  I would suggest that we take one system and use that in the infobox for this, and whether that is Hill, alphabetic, or in order of the periodic table is all fine with me - but not 'this is inorganic so this method', 'this is organic so this method', 'this is organic but of interest for pharma so this method' ..  --Dirk Beetstra T  C 11:27, 3 March 2016 (UTC)
 * Right. Adding details, to consider: maybe in the infobox write both basic empirical formula and more complicated formula when available. (and search-only text tbd). The infoboxes can be automated for that (i.e., extract Hill from structural formula, in that direction). In article text, it is upon the editors to use the best one in context. Then there is the guideline for the editors to write/improve (as was the OP question here), to bring consistency in article text. -DePiep (talk) 12:21, 3 March 2016 (UTC)


 * I dont think that IUPAC inorganic (red book) are writing about formulas without structural information (except for empirical formulas - but I guess we arent interested in that), maybe because the cation and anion nearly always are known?. (DePieps inorganic nomenclature link from the top of this thread ) says there is two ways to "sequence of citation of symbols in formulae" (section 4.4): by electronegativity or alphanumerical order. Both are tricky because there is a lot of exceptions, but they have an example (section 4.4.2.2: "N(3-), NH2, NH3, NO2-, NO2(2-), NO3-, N2O2(2-), N3-, Na, NaCl, NH4Cl". They are mention the Hills system, but only for empirical formulas. It looks like they are ordering the cation and anion alphabetical? (they say that NH4 is an exception). There are lot of exceptions, e.g. in section 4.4.3.2 they write: "For carbaboranes, there has previously been some uncertainty over the order of B and C. The order 'B before C' recommended here conforms to both electronegativity and alphabetical order (i.e. it is an exception to the Hill order in section IR-4-2-1). In addition, carbon atoms that replace skeletal boron atoms are cited immediately after boron, regardless of what other elements are present (see also section IR-6.2.4.4)". Christian75 (talk) 14:19, 3 March 2016 (UTC)
 * I have striked it out because it was about how to order the compounds in lists, indexes etc. But at least the striked out formulas are the way IUPAC suggest it should be written in lists (for inorganic compounds). Christian75 (talk) 17:00, 3 March 2016 (UTC)
 * Re Google. Google are extracting information from our chemboxes (and cite Wikipedia). Search for Silicium carbide and Googles infobox says "Formula: CSi". Christian75 (talk) 17:26, 3 March 2016 (UTC)

The CAS RN linking

 * Note: after implementation, the examples do not show the previous (bad) situation any more

Today, the CAS Registry Number link in and  looks like this:
 * (for PVC)

Note: Infobox drug works similar, and so joins this talk+outcome.
 * My questions are:
 * Is it a source or an external link?
 * As a source, it should be a &lt;ref> reference and show like: 9002-86-2[1]
 * As an external link, it should show like: 9002-86-2


 * Then there is this: sometimes the target site link does not know the CAS number:
 * 60232-85-1 (PCHP)


 * Proposal:
 * First, make it show as an external link: CAS number: 9002-86-2
 * ✅ Now shows external link icon in both templates. Other issues contine below. -DePiep (talk) 22:21, 7 March 2016 (UTC)


 * Second, re blank targets - dunno. Any suggestion? Add option none, would show "60232-85-1" (unlinked)?
 * -DePiep (talk) 21:52, 5 March 2016 (UTC)


 * All external links in the chem/drugbox are formatted that way (and that is also how they should print). The links both need to be an external link (in principle a reader/user should be able to directly link out to the site and use the data in the external database (which, granted, for CAS is practically blanc; ping User:Walkerma) as well as a primary source (the link provides the evidence that the number is correct according to the issuer of the number).  As I said, the CAS online database is rather limited and is not very helpful - but that is their choice (I expect that people will turn to the more useful databases down the list).  With that, I would disagree to link it somewhere else (as CAS number is issued by CAS and not by any other more useful database).
 * Regarding the incompleteness of the external database - that is still beyond me, but it is a problem with the external database. For us, the numbers are generally correct, and should link to the 'correct' page, even if that page does not (yet?) exist.  Again, the disservice that CAS is providing there will result in readers not caring about following the link to their databases or services.  I do not think that blanking is a solution on our end, we should be providing correct CAS numbers.  --Dirk Beetstra T  C 04:01, 6 March 2016 (UTC)
 * To clarify: in the sample situation "CAS link page does not know CAS RN 60232-85-1", we could show but not link that CAS RN (not: hide that CAS RN). -DePiep (talk) 19:17, 6 March 2016 (UTC)
 * re Beetstra: In general, I can agree. Especially: the external link provides a useful link and the reference (source); we don't want a separate ref (superscripted). And I can live with the occasional zero-hit CAS link (like PCHP).
 * About the link formatting: this: we really, really should show it as an external link, with the el-icon: 9002-86-2 (not 9002-86-2 ). This is the guideline. The icon says "you are leaving the wiki site". Tellingly, it also shows in mobile view. (Sure other links in Chembox are showing with suppressed icon. But that is not a reason to skip that guideline. Instead, expect more change proposals). -DePiep (talk) 19:45, 6 March 2016 (UTC)
 * I think it would be fine to include the "external link" symbol, though as Beetstra says, it's mainly a primary source but slightly useful as a external link. Being able to link directly to a CAS page confirming a CAS number is invaluable, IMHO, since CAS nos. are so widely used as substance identifiers, even in government documents.  Regarding the no. of entries, the consensus at our Wikipedia gathering in Boston last August (at the ACS national meeting) was that we should ask to expand the no. of CAS numbers that are available on Common Chemistry.  Now that trust is well established, and both WP and CAS have expanded their no. of substances, it's about time to expand from the 7800 we can use currently (of which perhaps 5000 are used on WP).  Before we expand, though, I want to make sure that our validation system is working properly first.  Walkerma (talk) 02:10, 7 March 2016 (UTC)
 * I had a look around, and I think that indeed 'we' are the odd one out here, displaying these links without the 'tag'. I think we should bring that in line (it has been a couple of years ago that we decided to remove them, CCC) with the rest of Wikipedia.  In the end, it is not a big deal.  But please have that consensus for all of them, we don't need to discuss that separately for every one of them.
 * Regarding having the links there also if they go nowhere, I stand with that they need to stay, and need to go to the issuer of the identifier. If the external party decides to add a couple of hundreds of them, we should not need to edit hundreds of pages to remove the 'CASNo_link=none'.
 * I will see if I can find some time to look at the functioning of User:CheMoBot - it has not had my focus for months and months now, it seems to glitch (sometimes). --Dirk Beetstra T  C 09:08, 7 March 2016 (UTC)
 * ✅ external link icon now showing (isolated issue). Other topics still open. -DePiep (talk) 22:21, 7 March 2016 (UTC)
 * That should also be done for all other external links in the infoboxes. Now it is inconsistent.  --Dirk Beetstra T  C 03:19, 8 March 2016 (UTC)
 * Yes, and I already announced so. But this is an improvement anyway, even while other stuff exists. Also, given the discussion this one identifier caused (I'm not surprised), it would not be wise to open twelve simultaneously. -DePiep (talk) 10:59, 9 March 2016 (UTC)
 * I think most of the discussion was on the reference/external link question, and the blanks on outside sources, I think that if we would have decided to add the 'outside link icon' on all of them in one go (as I suggest above) that would have gone through fast. --Dirk Beetstra T  C 11:07, 9 March 2016 (UTC)
 * I explained more than once why I did what I did. And I told you that I knew the other options all along. You did not respond to any of these remarks. Instead, your commanding tone and hindsight-iknowbetter does not help anything. -DePiep (talk) 19:59, 9 March 2016 (UTC)

Expanding to other identifiers
I have prepared to deploy this ferature to most other external links: identifiers, ATC code. See:
 * /testcases (general examples), /testcases6#Indentifiers (test all parameters), /testcases8#ATC (ATC Code).

Will be added: external links have css-title (mousehover-text) showing the target website name. Drugbox already has these changes. -DePiep (talk) 18:37, 5 April 2016 (UTC)
 * To simplify matters & doc, ATCCode is abandoned. ATC is entered by ATCCode_prefix=, ATCCode_suffix=. One article needed a change. -DePiep (talk) 08:40, 6 April 2016 (UTC)
 * preparing. -DePiep (talk) 08:40, 6 April 2016 (UTC)
 * ✅ -DePiep (talk) 09:50, 6 April 2016 (UTC)
 * Really, I just looked for these edits' effects in big methanol and I enjoyed the result. Also note the textual changes re JMol. -DePiep (talk) 21:52, 7 April 2016 (UTC)

Solvent or Detergent chembox with HLB, CMC etc ?
Is there any derivative chembox for solvents or detergents with HLB, CMC etc parameter fields. If there isnt can someone please help make it ? --Dr.saptarshi (talk) 06:27, 6 February 2016 (UTC)
 * Moved here from subpage. ping -DePiep (talk) 07:33, 6 February 2016 (UTC)
 * If there is support for this, we can add those parameters. -DePiep (talk) 07:35, 6 February 2016 (UTC)
 * I would be in favor. Possible issues with CMC: it can be defined at either mol/L or g/L, we might want to define which units we're using now to prevent it becoming a mess (wikidata would thank us). Some of the most common surfactants are often produced as a mixture of chain lengths (e.g. Sodium dodecylbenzenesulfonate) in which case the CMC of a pure material may not be the same as the CMC of the 'normal' material - Do we want to allow for that? Maybe by allowing a range to be entered? --Project Osprey (talk) 10:36, 6 February 2016 (UTC)
 * I get this impression: these are rarely used parameters, but when used they are really relevant and helpful. I propose to add HLB, CMC to section . I can not see a pattern or fixed unit, so for now we should give them free text (input = shown unchanged). OK? -DePiep (talk) 20:23, 6 March 2016 (UTC)


 * See /testcases9#CMC_HLB. Any actual examples tot test?, . -DePiep (talk) 23:22, 7 March 2016 (UTC)

Test case using Tween20. Looks ok to me, did you hardcode any units for the fields? --Project Osprey (talk) 09:56, 8 March 2016 (UTC)


 * Not yet. Which unit would that be? I like the mentioning of 'Surface tension', cannot just show unknown codes. (tweaked the presentation). -DePiep (talk) 11:21, 8 March 2016 (UTC)
 * So. Shall we add CMC and HLB this way, or forget about it? ping, -DePiep (talk) 22:25, 28 March 2016 (UTC)
 * Sorry! I meant to get back to you but then forgot... For me this way is good. --Project Osprey (talk) 08:15, 29 March 2016 (UTC)

✅ CMC and HLB in section Properties. No precoded units, because they can vary. So it's anytext to enter. -DePiep (talk) 18:19, 29 March 2016 (UTC)
 * Still not used in Polysorbate 20. -DePiep (talk) 09:34, 9 April 2016 (UTC)
 * Working on it now.

Possible issue: CMC value can be a long text-string causing it to run over more than one line (see example below, were info from page has been moved to chembox without removing excess detail). The result is confusing, as the layout is disrupted. Is there a simple fix? (other than not adding 3 version of the same data) --Project Osprey (talk) 08:59, 11 April 2016 (UTC)


 * Will fix, later. -DePiep (talk) 11:19, 11 April 2016 (UTC)


 * ✅ FYI, instances are listed here. -DePiep (talk) 20:56, 11 April 2016 (UTC)
 * Many thanks. --Project Osprey (talk) 21:27, 11 April 2016 (UTC)

Pronunciation
We have added this to the drugbox as seen here metoprolol. I propose adding the same to this template. Doc James (talk · contribs · email) 18:24, 7 May 2016 (UTC)

Using the infobox for isomers
I found that the chemical infobox is used not only for a specific compound, but also for families of isomers (e.g. Sodium dodecylbenzenesulfonate). Of course, problems arise when the isomers have different properties. Nonetheless, the PubChem code can refer only to one specific isomer. I guess the infobox should be used specifically for a single isomer (maybe the most important of the family), am I correct? --Esponenziale (talk) 13:54, 20 May 2016 (UTC)


 * The solution I've adopted is to use one chemical infobox for the family of isomers, with only the properties and ID codes shared by the family, and separate infoboxes for the single isomers, with their particular properties and ID codes, see current Sodium dodecylbenzenesulfonate --Esponenziale (talk) 13:34, 23 May 2016 (UTC)

I see where you're coming from but we don't do it like that (you end up with too many infoboxes). Instead the chembox code can handle several Pubchem (and other) identifiers. For example, the code below:


 * PubChem = 23662403
 * PubChem_Comment = Sodium 2-dodecylbenzenesulfonate
 * PubChem1 = 23667983
 * PubChem1_Comment = Sodium 3-dodecylbenzenesulfonate
 * PubChem2 = 4289524
 * PubChem2_Comment = Sodium 4-dodecylbenzenesulfonate

Gives you the following.

--Project Osprey (talk) 15:04, 23 May 2016 (UTC)


 * Thank you, perfect. I've just corrected the page following your advice --Esponenziale (talk) 15:27, 23 May 2016 (UTC)

InChI & Main hazards
Hello,

Two questions.

First, is there any site where you can get InChI's? And I don't mean standard InChI's. If you go to ChemSpider or PubChem, they use the standard ones. If you draw a compound in ChemDraw and let it look for its InChI, it gives you the standard one. I don't know about other software out there, but do you really have to use the software from InChI Trust?

And second, what actually goes in the MainHazards field? Is it obligatory? What kind of reference should you use? I was editing pyridinium chlorochromate, and its MainHazards field is filled. But ammonia doesn't have it.

Please use Reply to. Thanks! Georginho (talk) 10:41, 11 June 2016 (UTC)


 * MainHazards is not obligatory. The opposite: it should only be used when the hazards are major. Listing minor also-hazards makes the infobox unncecessary long (the same for most other parameters). re pyridinium chlorochromate: don't know, but it looks to me that its hazards are not overwhelming. For example, none of these hazards are mentioned in the lede (compare DDT). Also, because the hazards are or should be present in the specific hazard data (eg as pictograms), the current MainHazards input may be redundant. -DePiep (talk) 11:04, 11 June 2016 (UTC)


 * personally I think that MainHazards should be deprecated in favor of GHSSignalWord as the Signal Word appears in modern SDS forms. --Project Osprey (talk) 19:53, 11 June 2016 (UTC)


 * Thanks! Yeah, I don't see anything major about PCC compared to other chemicals. I think I'll probably leave carcinogenicity and aquatic toxicity in the box per the SDS.
 * I also think that although GHSSignalWord appears in SDS's, its use shouldn't be obligatory either. Putting "Danger" in the box for PCC, for example, doesn't say much.
 * Cheers! Georginho (talk) 18:35, 12 June 2016 (UTC)
 * Great. Hope you find something about InChI too ;-). re Project Osprey: maybe in a few weeks or months time we should clean this up into modern SDS format then. -DePiep (talk) 18:41, 12 June 2016 (UTC)

Chemical formula: now accepts D= for deuterium
See Template_talk:Infobox_drug. Also used in molar mass calculation. -DePiep (talk) 20:57, 4 July 2016 (UTC)

Edit request
The current chembox does not show the footer (the footer starting with "Except where...") iff the article has a data page (e.g. ethanol does not have a footer because of Ethanol (data page) vs. Cobalt(II) chloride has a footer because of Cobalt(II) chloride (data page)). It is fixed by adding a new line before the template call of the footer template. It has been tested by adding chembox/sandbox for ethanol and cobalt(II) chloride in preview. The testcases doesn't show any problems either.

And the current chembox forget to call the Chembox Supplement template with the parameter e.g. the folowing dosn't work: gives

The sandbox version fix this:

The version should replace template:Chembox Christian75 (talk) 13:31, 20 July 2016 (UTC)
 * Yes check.svg Done GeoffreyT2000 (talk) 18:01, 20 July 2016 (UTC)

Plural IUPACName?
I always thought the purpose of IUPAC was to pick a single name, but allowing that the template doc allows the plural IUPACNames = as well as IUPACName =, that is fine.

But why does the page Luminol display as a plural when the source says "IUPACName = 5-Amino-2,3-dihydro- [html break] 1,4-phthalazinedione" Is the template trying to be smart and thinking there are two lines, there must be two names and changes to plural? Having the two lines, it is confusing to see the plural used on that box.73.81.157.243 (talk) 07:16, 24 August 2016 (UTC)
 * I resolved this on luminol by removing the HTML break. Chembox treats the HTML break (&lt;br>) as if the IUPACName= value is a list of several names, one per line, rather than a single name that is forced to wrap to multiple lines. Line-breaks were often inserted because chemical names can be so long that they overflow the width of the chembox. Given we have a plural field, the singular field problably should assume only a singular value is allowed. But that doesn't solve use of embedded linebreaks in individual entries in the plural field. Shouldn't we be using semantic list markup instead of a newline-delimited list (see Manual of Style/Lists and especially WP:UBLIST)? DMacks (talk) 07:37, 24 August 2016 (UTC)

FDA Approval Date Section for Medications
If possible, I would like to know if a section can be added to this template for future projects. For example, navboxes which may contain drugs to treat certain medical conditions may be sorted by date of approval. It is something I am interested in at this moment (for future projects). Any chance we can have an approval section by a governing agency/authority? Thanks! Twillisjr (talk) 17:08, 14 October 2016 (UTC)
 * this would be a Template:Infobox drug thing and more appropriate at WT:PHARM. Jytdog (talk) 18:43, 14 October 2016 (UTC)

Data page not showing at ethanol
The link to the data page should show automatically as long as Ethanol (data page) exists right? I'm not seeing it at the bottom of the infobox anymore. Sizeofint (talk) 04:08, 18 November 2016 (UTC)
 * Looking into this. -DePiep (talk) 12:47, 18 November 2016 (UTC)


 * #Edit_request (20 Jul 2016) did not work out well.
 * Expect to refill (~150 articles, iirc).
 * Thanks for pointing to this one.
 * -DePiep (talk) 16:04, 18 November 2016 (UTC)
 * -DePiep (talk) 16:04, 18 November 2016 (UTC)

Add option Jmol= to overwrite SMILES input
Today, the Jmol external link is fed with the SMILES input string (e.g. DDT). In a few situations this is undesired or plain wrong (Ferrocene IIRC). I propose to add option Jmol to overwrite that SMILES-input for the Jmol link. Also, the option none will suppress the Jmol data row (will not show at all). The SMILES output will remain untouched. See Testcases/demo. Documentation:
 * 1. By default, the Jmol external link is fed with the SMILES input. So Jmol does not need input.
 * 2. When None or none, the Jmol data row is suppressed (not shown).
 * 3. When &lt;some SMILES string>, Jmol links will show that string in 3D, no matter what SMILES is. (SMILES output itself is unchanged).

Additional notes:
 * Recently this same Jmol setup was introduced in Drugbox. (Before, did not have the Jmol external link at all). See talk.
 * Articles with any Jmol input are categorised in.
 * Labeltext is changed into "3D model (Jmol)" to open with the main point (3D), not jargon (Jmol). And so resorted to near top.
 * Jmoln is indexed: &lt;blank>, 1 to 5 corresponding with the SMILESn indexes. So Jmol3 will overwrite SMILES3 input.

Comments? pinged. -DePiep (talk) 13:06, 7 November 2016 (UTC)
 * Looks good. One question.  Are separate links for Jmoln created? I don't see an example of that in the sandbox. Boghog (talk) 15:13, 7 November 2016 (UTC)
 * Yes. I've added demo sections: #Jmoln indexes, #SMILESn Comment handling (needs some fixing)  . -DePiep (talk) 16:01, 7 November 2016 (UTC)
 * Added addTESTdata to headerbar, won't show in mainspace.


 * ✅ Asking to take a look at the ferrocene issue. -DePiep (talk) 09:44, 17 November 2016 (UTC)
 * Todo: add the promised tracking category. Fixed the ferrocene SMILES string for Jmol :-). -DePiep (talk) 16:10, 19 November 2016 (UTC)

E number from Wikidata
I have prepared the E number (food additive codes) data row to get the E number from Wikidata. Articles doing so will be tracked in. With this, old E number and E number Comment are deprecated; there is no option to enter local data. Go? -DePiep (talk) 22:19, 21 November 2016 (UTC)
 * ✅. See . -DePiep (talk) 17:14, 23 November 2016 (UTC)

Wikidata tracking categories
Through ECHA Card, has entered Wikidata world. I propose to add tracking categories for all our wikidata (wd) entries. When done structured, they can be helpful and even support maintenance (corrections). Inspiring example: Authority control categorisation.


 * Simple form: ECHA has automated wd data only (no local parameter input). All Chemboxes with such wd ECHA input are to be listed in.


 * About the category name pattern:
 * 1. Use the wd property name ('ECHA Infocard ID') for the data point.
 * 2. Write pattern "[wd name] from wd wikidata" ('ECHA Infocard ID from Wikidata'). 1. name in front, 2. short (it is a tracking category, not a readers category).

Comments? -DePiep (talk) 21:13, 11 November 2016 (UTC)
 * Having a tracking category is probably a good idea. I would, however, propose to write “Wikidata” instead of just “wd”. --Leyo 13:07, 17 November 2016 (UTC)
 * Well, expect dozens of categories (CAS, InChI, ...) in one article. Tracking categories they are. I boldly dare to say: these are editor-only cats, not Reader space cat names! Description can be in a cat lede (=on the cat page). Shortened cat names, with dozens of tracking cats, the editor will understand. (maybe write "WD" not "wd"?).
 * Example of bad: see here. Not WP:IAR ie not All, but let's improve that rule for tracking cats. -DePiep (talk) 23:51, 17 November 2016 (UTC)
 * “WD” is surely better than “wd”. --Leyo 22:03, 18 November 2016 (UTC)
 * Thanks. I am even podering like 'cat:ECHA Infocard ID from wikidata'. But definitely not "cat:Articles using Chembox and having ECHA Infocard ID fetched from wikidata". -DePiep (talk) 22:54, 18 November 2016 (UTC)
 * looks good. -DePiep (talk) 23:29, 18 November 2016 (UTC)
 * Wikidata has a capital “W”. --Leyo 23:37, 18 November 2016 (UTC)
 * Not in this proposal. DePiep -23:41, 18 November 2016 (UTC)
 * OK, "W" it is. -DePiep (talk) 12:53, 19 November 2016 (UTC)


 * Proposal now: -DePiep (talk) 21:00, 19 November 2016 (UTC)
 * About ready to go live. -DePiep (talk) 22:13, 21 November 2016 (UTC)


 * ✅. See . -DePiep (talk) 17:16, 23 November 2016 (UTC)

Jmol live in WP
I have mentioned this suggestion in the Village pump (tech). If you know more, please add there. I am very low in the J/JS area. -DePiep (talk) 23:34, 17 November 2016 (UTC)
 * Archived here for future reference. Sizeofint (talk) 09:41, 25 November 2016 (UTC)

Chemical articles without CAS Registry Number
Not sure that this goes here but can someone please undelete Category:Chemical_articles_without_CAS_Registry_Number? --Project Osprey (talk) 23:55, 27 November 2016 (UTC)


 * See Category:Chemical_articles_without_CAS_registry_number. The category has been renamed to follow Wikidata writing (note the lowercase r, n). The overview is in.
 * Also will change: &rarr;
 * Also will change: &rarr;
 * Before we can make and  a fullblown Wikidata machine, I needed to clean up some details. Call me if something is wrong. -DePiep (talk) 00:20, 28 November 2016 (UTC)


 * Will do. Many thanks --Project Osprey (talk) 00:52, 28 November 2016 (UTC)
 * Expect a Wikidata pilot shortly (using CAS registry number). Dozens of details to be researched (eg, what with ?). We want multiple tracking categories for a single data point, especially "Wikidata value =/= local value". Follow this page. -DePiep (talk) 01:13, 28 November 2016 (UTC)
 * "CAS registry number" is spelled "CAS Registry Number". So should the category. Christian75 (talk) 05:30, 28 November 2016 (UTC)
 * As explained, Wikidata wrote in lowercase (at the time of decision). -DePiep (talk) 13:29, 28 November 2016 (UTC)

Pronunciation
We have pronunciation in the Template:Infobox_drug. Would be useful to have in this one aswell. Doc James (talk · contribs · email) 14:59, 1 September 2016 (UTC)
 * Agree. there was a brief prior discussion on this Wikipedia_talk:Chemical_infobox/Archive_9 Sizeofint (talk) 17:30, 1 September 2016 (UTC)
 * I can build something. Please specify the location for that datarow, e.g. by these demos. Would be like "right under the Names subheader", "under the top infobox name", etc.


 * Note that the pron should be unambiguously close to the name it says. There should be no confusion, while there are multiple names and name-types in . (Note: in, I have proposed these order changes. When accepted, the pron datarow will be the top text row (under the images), near the infobox title.) -DePiep (talk) 14:45, 11 November 2016 (UTC)
 * I would put it under the "names" heading. Doc James  (talk · contribs · email) 18:15, 11 November 2016 (UTC)


 * ✅ (any text) will show in section Names. -DePiep (talk) 17:37, 19 November 2016 (UTC)
 * Not working at glucose User:DePiep Doc James  (talk · contribs · email) 17:44, 29 November 2016 (UTC)
 * Fixed. Good point. I said, in the editsummary: /* top */ fix: |pronounce= should be outside of any {{Chembox|SectionN=} } (yes I know. yes I know. But hey, this template now does handle 575+ parameters :-) )). -DePiep (talk) 18:48, 29 November 2016 (UTC)

Major proposal: add Wikidata *external link* to the Identifiers
I propose to add data row "Wikidata: [wikidata item external link]" to and, section Identifiers. By default, the link should be provided automatically. Example would be: Carbon monoxide&rarr; Q2025

In practice, nearly every {Chem/Drug infobox} will show this link. Articles will be tracked (categorized) usefully. Later on, we will use that Wikidata link for properties, like PubChem numbers. -DePiep (talk) 21:55, 1 December 2016 (UTC)
 * I would propose not to show, i.e. Q2025. --Leyo 23:34, 1 December 2016 (UTC)
 * Good. If this is the only note ... ;-). (Just think of my elation this morning when I woke up thinking not to use WD for say property CAS number, but as an external link itself! Made my day today). -DePiep (talk) 23:58, 1 December 2016 (UTC)

Proposal: add label for indexed identifiers
Indexed identifiers (index can be:

&lt;blank>, 1, 2, 3, 4, 5)

CASNo

ChEBI

ChEMBL

ChemSpiderID

DrugBank

InChI, InChIKey

IUPHAR_ligand

KEGG

PubChem

SMILES, Jmol

UNII In Chembox, we have some 13 indexed identifiers (see box right). Using these, the can handle up to six different compounds (by entering values for CASNo=, CASNo1=, CASNo2= etc., see example linalool below). At the moment, over 500 es use more that one CAS number.

Currently, the different numbers are specified by an extra comment=-input for each id.

I propose to add these six parameters to the : This will happen: when used, the same label will be added before each of the values with that index. (so: index2_label will precede CASNo2, and PubChem2, and ...2, values). This simplifies this sub-identification, and it stimulates that the editor can align the indexed values (i.e, make sure that same-index == same-substance. Or: index2_label == CASNo2 == PubChem2 etc.).
 * index_label=
 * index1_label=
 * index2_label=
 * index3_label=
 * index4_label=
 * index5_label=

Example:


 * (R)
 * (S)


 * (R)
 * (R)
 * (S)
 * (S)

Minor notes: using these labels is also more simple, compared to having to add a separate comment for each data row. The comment input options may be less needed. And above all: this aligning the indexed identifiers is a good preparation for the Wikidata changes to come. It allows systematic data loading from Wikidata -- later more. More demos are in /testcases5.

Any comments, or support right away? -DePiep (talk) 21:38, 30 November 2016 (UTC)
 * Funny. I am also testing Wikidata things, so the demo looks strange. For now and this: just look at the (R) and (S) texts plese. -DePiep (talk) 21:03, 2 December 2016 (UTC)

ECHA InfoCard ID
What about adding ECHA's InfoCard ID to the chembox? They were introduced in January and are available for 120.000 chemicals. According to ECHA, an InfoCard serves as a high-level summary for a broad public, consisting of information that is most relevant to an audience of consumers, downstream users and professionals active in the chemical industry.

The InfoCard ID of DEHP is  and the full URL is http://echa.europa.eu/substance-information/-/substanceinfo/100.003.829.

As a sidenote, the same ID also works for other ECHA databases, e.g.: BTW: On Wikidata, I made a bot request to import the data to chemicals' items using the newly created property. Once, this will be done, the IDs may be obtained from there. --Leyo 00:15, 1 March 2016 (UTC)
 * http://echa.europa.eu/brief-profile/-/briefprofile/100.003.829
 * http://echa.europa.eu/information-on-chemicals/cl-inventory-database/-/discli/substance/external/100.003.829
 * http://echa.europa.eu/information-on-chemicals/ec-inventory/-/dislist/substance/100.003.829
 * http://echa.europa.eu/information-on-chemicals/registered-substances/-/disreg/substance/100.003.829
 * http://echa.europa.eu/candidate-list-table/-/dislist/substance/external/100.003.829
 * http://echa.europa.eu/addressing-chemicals-of-concern/authorisation/recommendation-for-inclusion-in-the-authorisation-list/authorisation-list/-/dislist/substance/external/100.003.829
 * http://echa.europa.eu/addressing-chemicals-of-concern/restrictions/substances-restricted-under-reach/-/dislist/substance/external/100.003.829
 * I've put a not note at d:Property_talk:P2566 on why the id is not the same as the EC number (/list number). -DePiep (talk) 07:49, 1 March 2016 (UTC)
 * I replied there. --Leyo 08:51, 1 March 2016 (UTC)
 * Concerning your comment “Job for wikidata then, useless route to add this locally.”: Yes, the IDs should be added to the Wikidata items. What I am talking about here is to obtain these IDs from Wikidata in order to show them (incl. the link to the ECHA InfoCard) in the chembox. --Leyo 15:17, 2 March 2016 (UTC)
 * We understand. It looks like this datapoint should be retrieved from wd only, no need for local parameter option.
 * Back to the prime question: add this external link?: yes, looks like a very useful and helpful link. (In the longer future, this should go into an "External links" box. Example: Gout). -DePiep (talk) 19:28, 5 March 2016 (UTC)

Update
The Wikidata property is now contained in a few thousand items. By adding the following code to the chembox, the ECHA InfoCard ID is being shown in each corresponding article.

We may either add it to Chembox Identifiers or to Chembox Hazards (InfoCards contain relevant regulatory information). --Leyo 08:36, 3 October 2016 (UTC)


 * I would just source this from WD, no need for first making this local while we are already discussing for some information to make the move (and, depending on how 'comparable' WD and local data really is, I am in favour of sourcing over local data). Support adding the code to Chembox Identifiers and/or Chembox Hazards (there is different information on different urls).  --Dirk Beetstra T  C 09:02, 3 October 2016 (UTC)
 * I am not sure if I got your first sentence right. The suggestion is to transclude the ID from Wikidata, not to copy it. --Leyo 09:31, 3 October 2016 (UTC)
 * That is what I meant as well: transclusion, Leyo. --Dirk Beetstra T  C 10:31, 3 October 2016 (UTC)
 * Since the InfoCard (example) contains a summary and links to other interesting pages, it is probably enough to link only this one for now. --Leyo 20:35, 3 October 2016 (UTC)

I have boldly implemented it in the Chembox Identifiers, see diff. Please revert if it does break somewhere, it is rather difficult to test this in the testcases. --Dirk Beetstra T C 05:14, 4 October 2016 (UTC)
 * Thank you. I haven't spotted any issues so far. --Leyo 08:22, 4 October 2016 (UTC)

Add option for local input?
Today, in the 10.600 articles some 6000 are showing this d:ECHA InfoCard ID value. Should we add a local (enwiki) parameter ECHA InfoCard ID to allow local overwriting/adding? These situations may occur: 1. the WD value may be incorrect for the article, or need detailing. This requires a local source to be used (ECHA InfoCard ID ref), and checking these articles is a maintenance task (using a tracking category). 2. The en:article title may not be the exact WD item title, so the WD check fails and returns a blank. Useful? -DePiep (talk) 08:22, 17 November 2016 (UTC)
 * If an InfoCard ID is missing at Wikidata, it should get added there. Or could you provide an example where is wouldn't work? --Leyo 13:04, 17 November 2016 (UTC)
 * I don't have an example. I thought of the situation where the article title differs from the WD-item. This category lists ~500 with multiple chemicals (by using CASNon-indexes). Sure these indexed chemicals could have their own ECHA (which could be indexed too or could use the expensive non-natural item fetch, if I'm correct). -DePiep (talk) 13:14, 17 November 2016 (UTC)

Also in {Drugbox} then?
Today, Infobox drug does not have some parameter ECHA Infocard. Worth adding, by Wikidata automated? -DePiep (talk) 21:26, 19 November 2016 (UTC)
 * Well, drugs are not so much in the scope of REACH and CLP the thus of the ECHA. However, I assume that there are still many compounds with drugbox that have an ECHA InfoCard. --Leyo 18:40, 20 November 2016 (UTC)
 * You might be able to verify that using the lists of articles with a Wikidata item containing that property (irrespective of the infobox used in the articles). --Leyo 23:26, 23 November 2016 (UTC)


 * To learn what? If we (enwiki editors) decide that we should add the ECHA data points to the drugbox, then we´ll do it (from Wikidata or by local input does not matter). It is not the right route to say "Wikidata has it, so let's add it". -DePiep (talk) 10:41, 24 November 2016 (UTC)
 * It's about the availability of InfoCards at ECHA. The InfoCard IDs (if existing) were added by a bot to all Wikidata items with a CAS RN. --Leyo 22:40, 24 November 2016 (UTC)
 * Ergh, no. If we, editors, decide ENCHA data link should be in : it will be in there. Either by Wikidata, or by local input. BUT. Today ENCHA is not an accepted input option, so ECHA is not releveant input. -DePiep (talk) 22:49, 24 November 2016 (UTC)
 * What's ENCHA? --Leyo 22:58, 24 November 2016 (UTC)
 * "ENCHA" is a common, generic name for "any data thing you want it to mean". By context, that should be clear already. If you need extra individual advice by example, just read: "ECHA". -DePiep (talk) 23:10, 24 November 2016 (UTC)
 * Well, if in the lists mentioned above there were very few articles with drugbox, it would probably not be worth adding such a parameter to drugbox. This is what is the potential use of these lists. --Leyo 23:17, 24 November 2016 (UTC)
 * That's my question in point: how is it "worth" inclusion/exclusion by this reasoning? If few drugs have an ECHA, these must be remarkable exceptions. If many have, it's quite common to have an ECHA. Both ways: good to add. Be sure, this can be said about dozens of data points (today already: E number food additive). A good reason not to add ECHA (either local or from Wikidata) is: "ECHA is not relevant (enough) for a drug" (I do not know this myself, to be clear. It is what I'm asking here). -DePiep (talk) 06:59, 25 November 2016 (UTC)

There are currently [//petscan.wmflabs.org/?language=en&project=wikipedia&templates_yes=Infobox%20drug&outlinks_any=User%3ALeyo%2FInfoCard%2F0%E2%80%93K%0D%0AUser%3ALeyo%2FInfoCard%2FL%E2%80%93Z&interface_language=en&active_tab=&doit= 2579] out of 6529 articles (40%) with drugbox that have an InfoCard ID on Wikidata. --Leyo 08:08, 30 November 2016 (UTC)
 * . I've formally put the question at Template_talk:Infobox_drug. (I'd rather not spend time advocating, I'll just see what happens. I want to spend my wikitime on introducing Wikidata big time in these templates). -DePiep (talk) 16:23, 1 December 2016 (UTC)
 * ✅, see talklink. Note: before this edit, the category listed 6084 Articles (that's from the ~10200 Chembox's only). -DePiep (talk) 10:48, 8 December 2016 (UTC)

Tracking category

 * Added: . -DePiep (talk) 17:17, 23 November 2016 (UTC)

Article ECHA InfoCard needed
We need article ECHA InfoCard. Today, 6000 articles link to it. Link to ECHA is not enough. -DePiep (talk) 19:40, 27 November 2016 (UTC)

Searching for duplicated pages
With ~16k compounds listed in chembox and drugbox I've sometimes wondered if there are duplicated pages. Obscure compounds could be named all sorts of things, particularly things like Category:Substituted_amphetamines and other designer drugs, where many people are trying to access every possible simple analogue but where indexing is understandably poor. I presume the best way to do this would be to scan all of the identifiers (CAS, Pubchem, etc) for duplications. There's not previously been a tool for doing that, and building one seems like a pretty poor use of time as it might not find anything. Now that these values have been moved to wikidata would searching for duplications be any easier? --Project Osprey (talk) 23:38, 5 December 2016 (UTC)
 * I'm working to (first) get the Wikidata link in the infoboxes, and the CAS number from WD. PubChem is next (because higly present in WD). Drugbox in tandem. Need some smart categorisation (cat:no Wikidata item/value, cat:enwiki local input differs from Wikidata value, etc.). These categories are a maintenance job (why no item?, why different CAS values?, PubChem values?). Numbers of pages in those cat's, I can not guess. I also expect more systematic issues: Widata truly wants: one compound=one item, nicotine seems to have two CAS numbers. These might need Wikidata involvement. Also, we have 500+ articles with multiple compounds (indexed CAS numbers), to handle & categorise re Wikidata.


 * But that does not address your question. For now, I cannot even think of a smart query setup. May I suggest: within days/weeks, there will be enough articles listed to walk through to check re Wikidata, by these other categories. Duid you have any plans for the holydays? -DePiep (talk) 23:58, 5 December 2016 (UTC)
 * A brute-force approach is use a database dump. One can pre-select a certain category if there were to be one that tracked all pages using chembox, but even the raw whole WP would be usable. If we have actual values in the WP chembox templates, it's easy to parse out a list of [pagetitle]->[chembox_casno] pairs and then look for dups in the [chembox_casno] values. If the values are imported from WD, then either it needs to be done on a WD export (and I don't know how to parse that!) and also cross-checked against any that are still hardcoded in WP. But it might be possible to dump the rendered pages or something like that, with the WD values pre-imported to WP. DMacks (talk) 00:06, 6 December 2016 (UTC)
 * That does sound like a truly awful way to spend Christmas... If a convenient tool doesn't exist now I'm sure it will later. Any large database needs a way of detecting duplicated entries. --Project Osprey (talk) 00:12, 6 December 2016 (UTC)


 * (ec) Still, wouldn't it be better to have a primary cleanup first: remove/fix all erroneous CAS numbers (categorised as: conflicting local:CASNo versus d:CASnumber)? -DePiep (talk) 00:16, 6 December 2016 (UTC)

There are Parnaparin sodium vs. Bemiparin sodium or Tinzaparin sodium vs. Semuloparin sodium vs. Dalteparin sodium sharing the same CAS RN. --Leyo 08:52, 6 December 2016 (UTC)


 * You could make a container category Category:Chemical duplicates by CAS-number, categorise all by Category:, and have a bot fill all the Category: pages with 'hidden category Category:Chemical duplicates by CAS-number'. One could then just browse through the pages of Category:Chemical duplicates by CAS-number, see all the subcategories, and each subcategory would tell how many pages are in there.  All that have more than one are suspect.  It is however going to be an awful tree with no further use.  The only other thing I can think about is to have a script parse out all the CAS-numbers, toss them in a .csv with the pagename, load the .csv into a spreadsheet program and sort them by CAS-number.  Add a column with "=if(A2=A1,'duplicate',' ')" in each cell, and see which ones show up.  The former solution is easy to implement before Christmas .. the latter is indeed an awful way to spend Christmas ..  --Dirk Beetstra T  C
 * The latter solution should by the way be done on wikidata-data, making a listing like I said. That could then also nicely include all pages on other wikis, so one also kills the other bird: Chemicals which exist elsewhere but not here locally (but could be ported).  --Dirk Beetstra T  C 10:32, 6 December 2016 (UTC)
 * I won't spend any time on this, I promise, provided that this excercise does not interfere with the developments I am working om. Especially the sandbox stacks and CASNo param handling are off-limits. Deal? -DePiep (talk) 11:54, 6 December 2016 (UTC)
 * We only might ask you to categorise chem/drugboxes by CAS-number by adding a little piece of code (well, it is already there, just the other half of the 'IF there is no CASNo THEN categorise as a compound without CASNo') .. I will not touch it myself either (without consultation). --Dirk Beetstra T  C 13:04, 6 December 2016 (UTC)
 * Already today there is . (Today there is no Wikidata check at all).
 * And as I wrote a few posts earlier, within days/weeks the first WD + CASNo tracking cats will go live. Any expansion of the logic is a headache and bad development practice. I see no need to speed up this particular dups-check when the primary check (align d:CASnumber with local CASnumber) is about to be available. -DePiep (talk) 13:23, 6 December 2016 (UTC)


 * Meanwhile, over at Wikidata: this. -DePiep (talk) 15:18, 8 December 2016 (UTC)
 * I was going to suggest such, but the constraint vios cannot be filtered to en.wp-specific content ([un]fortunately). --Izno (talk) 15:20, 8 December 2016 (UTC)

Tracking category "mass overwritten" dropped
Tracking has been removed from the populating infoboxes {Chembox}, {Drugbox}. Empty now, and will not be populated again. The original intention was: compare entered mass (MolarMass) with calculated mass. Well, that did not work. -DePiep (talk) 22:26, 10 December 2016 (UTC)

Added parameter | Drug_class to pharmacology section
Similar to Infobox drug, I have addded Drug_class to section. It shows in top, right above the ATC code. Also, I have moved the Legal_status data to the bottom of that section, outside of medical (clinical etc. ) data. -DePiep (talk) 11:13, 17 December 2016 (UTC)