User talk:Snipre

Welcome to Wikidata, Snipre!

Wikidata is a free knowledge base that you can edit! It can be read and edited by humans and machines alike, and you can help. Go to any item page now and add to this ever-growing database!
Need some help getting started? Here are some pages you can familarise yourself with:

• Wikidata:About - About the project.
• Wikidata:Introduction - Another introduction to the project.
• Help:Contents - The main help portal for editing and using the site.
• Wikidata:Project chat - Discussions about the project.

If you have any questions, please ask me on my talk page. If you want to try out editing, you can use the sandbox to try. Once again, welcome, and I hope you quickly feel comfortable here, and become an active editor for Wikidata.

Regards, --Jitrixis (talk | support my candidacy) 13:27, 1 November 2012 (UTC)

Coup de main ?

Salut !

J'essaie de travailler sur les communes de Suisse, mais je ne comprend pas tout... J'ai créé Q30986 pour la commune d'Aclens, mais je ne trouve pas de bouton modifier pour y mettre les interwikis... Est-ce que j'aurais manqué quelque chose ?

Merci d'avance de tes lumières ! Manoillon (talk) 12:46, 12 November 2012 (UTC)

Wikidata:Infoboxes task force/terms

Hi Snipre, thanks for the improvement! --Kolja21 (talk) 20:33, 31 January 2013 (UTC)

Re Property:P20

No one seems to be able to decide if these should be uppercase or lowercase, so I've proposed this for the time being. — PinkAmpers&^{(Je vous invite à me parler)} 20:46, 4 February 2013 (UTC)

Aide

Salut,

Tu peux m'aider sur le fonctionnement de la page de requête. Et me dire ce que j'ai fait de mal pour ça. Merci. Ludo29 (talk) 19:38, 5 February 2013 (UTC)

certes, mais je ne sais pas où regarder. Ludo29 (talk) 19:39, 5 February 2013 (UTC)

Il mettre la propriété dans une des classes d'objets ou alors créer une classe. Dans ton cas, j'ai créé les propriétés dans la classe Country Snipre (talk) 19:42, 5 February 2013 (UTC)

Tu vas peut-être me dire que je suis bête, mais je ne vois toujours pas comment renseigner l'altitude d'un sommet, comme avec Q1374. Et comment gère-t-on les unités ? Ludo29 (talk) 19:50, 5 February 2013 (UTC)

Les valeurs numériques ne sont pas encore disponible comme donnée. Pour l'instant, uniquement d'autres items ou de liens vers commons sont possibles. Snipre (talk) 19:52, 5 February 2013 (UTC)

Ah, ok. Je comprends mieux. Tu as une idée de l'échéance ? Ludo29 (talk) 19:53, 5 February 2013 (UTC)

Aucune idée. On a demarré hier et il y a encore plein de bugs à fixer avant une extension du système. Snipre (talk) 20:03, 5 February 2013 (UTC)

J'abuse de ta gentillesse à répondre à mes questions. Je souhaite avoir un nouveau champ canton pour renseigner le canton suisse dans lequel peut se trouver une ville, un sommet, un lac, un stade, etc. Je m'y prends comment ? Ludo29 (talk) 19:57, 5 February 2013 (UTC)

Pas de soucis. Tu poses une bonne question: je ne sais pas encore comment faire pour créer une nouvelle propriété. Normalement on commence par proposer la propriété et ensuite on fait du lobby :) . Et pour les cantons, il y a encore du flou si chaque pays définira sa propre structure administrative ou si un système générique doit être privilégié. Je te conseille de contacter user:Tpt qui est un admin francophone pour plus d'info: je ne suis qu'une petite main qui essaie de maintenir un peu d'ordre dans ce foutoir :) Snipre (talk) 20:03, 5 February 2013 (UTC)

Merci de tes réponses. J'abandonne pour ce soir, je dois filer en ville. Autant, la phase1 (itw) me passionnait peu, autant celle-ci, je kiffe. Ludo29 (talk) 20:05, 5 February 2013 (UTC)

Salut,

Je reviens à toi. Comment procéder pour obtenir la possibilité de mettre Conseiller fédéral dans le champ fonction d'un homme politique. Par exemple Q115442 Ludo29 (talk) 17:16, 14 February 2013 (UTC)

Et ça ? Ludo29 (talk) 17:47, 14 February 2013 (UTC)

Merci. Je viens de découvrir cette page en remontant tes contributions. Ludo29 (talk) 18:10, 14 February 2013 (UTC)

Soucis d'édition

Salut,

J'ai un soucis avec l'entrée concernant la Suisse : Q39. Diverses personnes s'échignent à vouloir rajouter le romanche comme langue officielle, ce qui n'est pas exactement le cas. J'ai ouvert une discussion sur la pdd de Q39 mais personne n'y vient. As-tu une idée de comment procéder ? Ludo29 (talk) 05:23, 26 February 2013 (UTC)

Books

Hi Snipre, I've seen you worked a lot on the book properties: have you seen this [force]? In the talk page we are discussing few issues about properties. Moreover, I've created a mapping between different infoboxes: not only Wikipedia ones, but also Commons and Wikisource ones. I've seen you are working a lot on this table, and I synchronized my mapping with that table. I think we have just few properties left, but I'd like very much all the book people to discuss in the very same page (like, here) :-) --Aubrey (talk) 14:40, 15 March 2013 (UTC)

Property proposal: was a(n) / war ein(e) / était un(e)

Dear Snipre, I left you a message here and I am looking forward to your answer. Kind regards, NormanB (talk) 23:27, 16 March 2013 (UTC)

Bot

Bonsoir Snipre,
Juste pour m'assurer que j'ai bien compris :

1. pour faire tourner les petits exemples mentionné sur la page Wikidata:Creating a bot, il faut déjà avoir un compte avec un bot flag ?
2. pour avoir un compte avec un bot flag, il faut créer un compte classique, et expliquer ici son "projet" de bot pour le compte en question, sans que son code soit nécessairement déjà écrit ?

--Gloumouth1 (talk) 21:24, 23 March 2013 (UTC)

Hello,

Pas besoin de compte avec flag pour faire des tests: tu peux faire des essais avec ton compte, faut juste faire gaffe de ne pas utiliser en parallèle ton compte pour contribuer manuellement, car en utilisant un navigateur, tu vas créer une deuxième session que les serveurs vont détecter.

Oui, mais encore une fois, si tu comptes faire des modifications ponctuelles (moins de 500 par jour) et que tu fais attention de ne pas faire de modifications toutes les secondes, alors le statut de bot n'est pas utile. A Snipre (talk) 21:31, 23 March 2013 (UTC)

"élément de"

Bonjour, j'ai vu que tu avais changé le libellé de P:P31 de "instance de" à "élément de". Il me semble que c'est juste d'un point de la logique formelle, mais dans le langage courant, "élément de" parait plutôt correspondre à fait partie de (le Mont Blanc est un élément des Alpes plutôt que Barack Obama un élément de personne). --Zolo (talk) 06:48, 29 March 2013 (UTC)

Ok, je reconnais que mon changement n'est pas le plus pertinent. Sinon, il y a Exemple de, Cas de, Sujet de,... mais il faut éviter instance parce que ce terme n'aide en rien à son utilisation. Snipre (talk) 10:08, 29 March 2013 (UTC)

"is a" a été renommé en "instance of" pour faire la distinction avec "subclass of". J'avais traduit par "exemple de" mais en fait ça fait un peu bizarre aussi. Quelqu'un a renommé en "instance de" qui est jargonant mais semble plus correct.--Zolo (talk) 13:52, 29 March 2013 (UTC)

drafted by

Could you please respond at Wikidata:PP/P#drafted_by. Thanks, Legoktm (talk) 17:20, 9 April 2013 (UTC)

Online meeting with Dev team

Hi Snipre, I copy and paste the few quick notes I added in Wikidata talk:Books task force:

• WD team will tackle Sister projects, one by one. They will start in few months, after the end of the present phase. they will start with simpler ones, eg. Wikivoyage.
• They don't knoow which "structure" the Commons file will have in Wikidata. We discussed with them the option 1 (above), but they say it's unlikely to have a WD for every manifestation. We then discussed more indetalis and they understood better the concept involved, and say that it's maybe doable. But it also depends by the community.
• They say a lot of metadata will be stored in Commons and Wikisource, not everything will be in Wikidata.

This means that they will provide tools and stuff, but many data will remain in Sister projects. This is somehow unexpected, but not bad per se. We need to stay tuned on this.

• It is not yet certain if Sources were to be stored in Wikidata or not.
• The only Wikidata ID will be for pages. There will be for subpages/sections. Aubrey (talk) 13:32, 12 April 2013 (UTC)

Regarding Sources, they just said they don't know what they'll do. They are asking for the community to come up with ideas, issues and solutions. So I think we (Wikidata books task force) are doing good, tackling issues before them, and foreseeing problems and solutions. I personally think that Sources will be a huge issue: at the beginning they were saying that they didn't wanto to have all book editions metadata in WD, but then I explained to them that we were talking just of the books present in Commons and used by Wikisource, so they changed their mind; it's few thousands book, not the whole bibliography of the world, so it's feasible.

I think that sources are just the same kind of issue, but in bigger: you need to have much more granular metadata (articles, preprints, books, monographies, working papers, thesis, whatsoever), and I think the possibile entities counts in millions. Template:Cite_web, for example, is used 1367805 times... I think that the gain to have a proper system to store these metadata would be enormous, and whole academic community would give us eternal gratitude. But's something big and difficult, and I don't know if feasible... --Aubrey (talk) 10:43, 16 April 2013 (UTC)

Sorry but it's stupid to import all metadata from books in Commons and Wikisource and to think that we will not do the same for all other books. For me there is only 2 options: we create for each manifestation an item or we create another database only for reference data storage and each time someone wants to source one statement we use a tool which import the whole set of data of the source into the source section (a link or an ID is not acceptable because we can't separate data and their references between different databases).

We facing 2 things: provide reliable data and this is only possible by sourcing with a complete data set for the reference, and reuse of reference data and this is possible only with a database. If Wikidata can't handle the storage of data for all books, articles, media,... we need to source the statements, we have to think about another database and tools to link wikidata and the reference database.

What I still miss are the technical or organisational reasons why we can't handle references with all bibliographic data in Wikidata. And if it is really an issue for wikidata we can stop here the development of wikidata because data without references including all bibliographic data are useless. So if you have any informations about why the development team has some reluctances to integrate all bibliographic data for references it could help us to choose a solution or simply decide that wikidata is not a solution for reliable data. Snipre (talk) 12:36, 16 April 2013 (UTC)

Hey man, I agree with you, I just say it's a very complex issue (from what I understand: but I'm a librarian, so maybe I overcomplicate things), especially as we intend books in different ways, eg as

• works (Wikipedia pages)
• manifestations
  • scanned books (Commons and Wikisources)
  • sources
    • for Wikidata
    • for Wikipedia bibliographies and references (eg Cite web)

I (for myself) am always puzzled by the confusion here in Wikidata, because I sense that sometimes we just not talk about the same thing.

Having said that, I agree that sources are overhelmingly important, and we need to discuss and propose solutions. What I can tell you, for example, is a "sensation": I felt that Denny and Lydia were a bit suspicios at the beginning, when I talked with them of "manifestations", because they probably felt we were suggestiong to replicate OpenLibrary in Wikidata, or to insert all manifestations of a book. That it is not so, and I replied to them we were talking about just the books in Commons and Wikisource (thus, only some-thousands books). I saw them relieved by this point of view, and they were much more easy to convince. I think we should do the same with Sources: have them to face a issue and propose a practical solution, integrated with Wikidata mission and structure.

My opinion is that the are just scared aff all the "projections" people are making in Wikidata, which often is viewed as a "panacea" and a "good-for-everything" project. Right now, they are thinking that "all the WD item are Wikipedia pages", and probably they want to stick with this structure for a while, seeing what happens. I agree with you that we need Sources, and possibly in a structured way. Don't ask me how, right now :-) Aubrey (talk) 13:26, 16 April 2013 (UTC)

Sources for wikidata and wikpedia are the same: at the end wikipedia will use sources from wikidata with each inclusion of data. so no difference. Then sources are not different from manifestations data because a coreect sourcing implies the whole bibliographic data in order to be able to verify exactly the data if needed. We can have some additional information for manifestations data but more or less manifestation data are equal to bibliographic data (publisher, date of publication, language, author,...).

The only limit we can put for references in wikidata is in my opinion the obligation of use of the reference in sourcing statement. So in order to avoid unecessary reference data we can have bots checking the use of reference and if no use a deletion procedure is launched. This means that all data from commons or wikisource can't be handled in wikidata and I agree with the WD team about that objective according to the initial plan. Snipre (talk) 13:48, 16 April 2013 (UTC)

I agree with all. So, if I understand it correctly, is just how to have these data in WD. Will we create an item for each source/edition/manifestation used (as reference or in Commons/WS)? Aubrey (talk) 14:03, 16 April 2013 (UTC)

I asked the development team about that but I think that if a source can be used more than once as reference for a claim we can create an item for a manifestation of a work. That will be the first step until the development team decides what to do with Wikisource and Common. Then I think we will find a consensus about how we can deal with data from commons and wikisource. That what I porpose to do. Snipre (talk) 16:14, 17 April 2013 (UTC)

Authority

Salut Snipre,
Pour Authority, tu nous laisses pas tomber, hein?
Avec 23PowerZ, vous prétendez que c'est facile à faire sans ma proposition, il sagit maintenant de la prouver.
Pas pour donner tord à l'un ou à l'autre, mais par ce que cette information d'authorité est vitale.
C'est un engagement moral que vous avez signé en rejetant ma proposition.
Amitiés Liné1 (talk) 09:03, 20 April 2013 (UTC)

Comme je l'ai déjà expliqué il faudrait commencer par une liste de ce que les biologistes ont besoin comme information pour construire leur authority.

Si on reprend les examples que tu as donné ((L.) L., 1755" for Prunus avium et Tang & F.T.Wang ex L.K.Dai"):

• (L), cela veut dire quoi, qu'est-ce que cela représente comme donnée. (L veut dire Liné, je sais, mais la parenthèse ?)
• L., cela veut dire quoi, qu'est-ce que cela représente comme donnée.
• 1755, cela veut dire quoi, qu'est-ce que cela représente comme donnée.
• ex, cela veut dire quoi, qu'est-ce que cela représente comme donnée.

Là on commence à voir la différence entre données et format. Ensuite, si on tient compte du besoin de sourcer le taxon, ce qui revient au même que le (L.) L., 1755, on commence à envisager d'utiliser les données de la source du taxon pour stocker des informations.

Je ne laisse pas tomber, simplement ta proposition est remise en question car elle est vise le mauvais objectif: ce que je veux comme résultat au lieu de qu'est-ce que j'ai besoin.

Ta proposition n'a pas été rejetée, elle a simplement été démontée par des arguments que tu n'as pas pu contrer. Un bon exemple est les auteurs: tu prétends qu'il est impossible d'introduire 15 auteurs alors qu'il le faudra bien si on veut donner la source complète du document où a été défini le taxon.

Bref, ma proposition est de renvoyer cette propriété au niveau de la task force biology et qu'au niveau de celle-ci la liste des besoins soit définie et ensuite on pourra voir comment on peut essayer d'intégrer ces données dans wikidata. Finalement on reviendra peut-être à ta proposition, mais la simple affirmation que c'est trop compliqué ne suffit guère pour convaincre, voilà tout. Snipre (talk) 09:47, 20 April 2013 (UTC)

PS: et ton affirmation C'est un engagement moral que vous avez signé en rejetant ma proposition. est osé: c'est ta proposition, à toi de la vendre. Je ne suis pas un biologiste, mais lorsqu'un spécialiste essaie de me vendre son produit uniquement sur son expertise, je me méfie. Snipre (talk) 09:52, 20 April 2013 (UTC)

J'ai répondu sur User_talk:Liné1#Authority pour la phylosophie et sur User_talk:Liné1#Détails_de_Authority pour les détails de taxonomie.

Amitiés Liné1 (talk) 09:43, 22 April 2013 (UTC)

Wikidata:Requests for comment/References and sources

I have done some clean up, if you think more is needed, go ahead. I think we still have some issues to resolve, but it is more dependent on implementation choices. Like, how should we address different editions (or translations) of the same book. --Micru (talk) 12:53, 29 April 2013 (UTC)

Different editions means different ISBN, different publication dates, for some books different editors,... so if we want to integrate that data in reusable way we have to create item for each manifestation of a work used as source in a certain number of statements. The only issue for me is to know what are the problems of creating a new domain S beside the domain Q for items and P for properties. Mixing Q item for a work and Q items for its manifestations is disturbing in my opinion and some people won't be able to do the difference and will add data of a manifestation to the work item if they don't find the appropriate item for a particular manifestation. So edition property will be defined in the item of the manifestation. Only the page property has to be defined in the source section beside the link to the manifestation item in order to allow the resusability of the source item. If these guidelines are applied in a mandatory way it will be easy to extract information from wikidata: all information about the source is defined in the source item describing one manifestation and the page property has to be recovered from the statement section for the source. Snipre (talk) 13:20, 29 April 2013 (UTC)

The main problem of having 2 namespaces is that there will be a lot more of items, and that it would be more difficult to maintain, create, update, etc. I am ok having all information in one item if it is possible to address the subsection that represents the edition with a unique identifier. I agree with you that otherwise it makes sense to have the information separated in different items.--Micru (talk) 17:04, 29 April 2013 (UTC)

So you give up about the reusability of sources ? We have no choice: we create an item for each source and we increase the total number of items and we can reuse the sources or we keep the number of item low and we forget about the reuse of sources. The question of a new domain S for sources doesn't change anything in case of the first option: if you create several items for different sources it doesn't modify the number of items if you call them QXXX or SXXX. So that is that choice we have to put in the request for comment. And at the end what is the problem ? The management of items or the size of the memory on the servers ? Because if a source is described by 6 properties and I want to use that source for 100 statements, dose it take more memory than one item with 6 properties with 100 links from the statements (I forget about the page property which is the same in both cases). Do you agree to present the choice like that ? Snipre (talk) 17:39, 29 April 2013 (UTC)

No, I don't give up on source reusability. Even if the sources are integrated into the same item as subsections, they would still be reusable by using the notation "Qxxx->Edition ID->Parameters of that edition (editor, year, etc)" (just an example). The problem is that we cannot guarantee now the uniqueness of the "Edition ID". I left a comment on Denny's talk page about it. I think the developers should decide on which option they think it is better for scalability. I have no idea about how taxing can be each one of the options.--Micru (talk) 17:54, 29 April 2013 (UTC)

That is the first time I hear that solution: is it feasible with the present structrue or we need to wait on new development ? Snipre (talk) 18:01, 29 April 2013 (UTC)

I think that was always an option and probably it was mentioned during the office hour last Thursday again, but yes, it would require some development.--Micru (talk) 18:11, 29 April 2013 (UTC)

Edition data

I left a possible solution on the books task force talk. Still needs to be confirmed if it could be valid, but what do you think?--Micru (talk) 14:57, 5 May 2013 (UTC)

I left a new comment about edition data.--Micru (talk) 17:10, 10 May 2013 (UTC)

I have expanded the RFC about references/sources to decide how to store edition data.--Micru (talk) 13:25, 16 May 2013 (UTC)

Thank you for the work. i will try to spend some time next weekend to help. Snipre (talk) 21:21, 16 May 2013 (UTC)

Hi Snipre, please take a look to my last comments after the Hackathon, if that is an acceptable solution we should aim to close the RFC on June 15.--Micru (talk) 23:55, 30 May 2013 (UTC)

Properties for review

Hey Sniper! I saw that you already commented below the properties "Crystal habit, Twinning, Fracture and Cleavage" Do you have time to review them? Some inorganic and organic chemicals can also form larger crystals that can be classified with those criteria (e.g. aspirin). Would be interesting to hear your opinion. (link) --Tobias1984 (talk) 17:15, 19 May 2013 (UTC)

At Wikidata:Requests_for_comment/References_and_sources#How_to_store_edition_data

Hello,

I observed that you deleted some of my comments at Wikidata:Requests_for_comment/References_and_sources#How_to_store_edition_data. I feel hurt that my comments won't be read there now. I would appreciate in the future that if you think that my arguments are not relevant that you contact me directly on my talk page so that I can address why you think it's not relevant. Maximilianklein (talk) 16:15, 20 May 2013 (UTC)

Re: Rotation period

Yes, probably "Time" datatype is not correct for rotation property. If you want, you can give your suggestions in this discussion. --Paperoastro (talk) 20:35, 30 May 2013 (UTC)

Properties for linking between items

Hi Snipre, your last comment on edition number (P393) suggests me that I might have not explained myself with enough clarity. As far as I know, there are this items links represented by certain properties.

1. Sourced statement -> item representing the source: stated in (P248)
2. item representing an edition -> item representing a work: "derived from" (proposed)
3. item representing a work -> item representing an edition: ???

We cannot use edition number (P393) because it represents "edition number" (not the edition itself) and its datatype is "string". We need a property with datatype item. Which property do you suggest for the third case?--Micru (talk) 13:07, 3 June 2013 (UTC)

Tool

Hi, I´m back online. You asked a question on my discussion page. The reason why I didn´t do anything about sources is that the tool is useful for adding the same data many times. I don´t see that someone would add a source for example ten times into different items. I think a tool for adding sources is needed, but this will be a different tool. Maybe asking for such a tool at the English project chat would be best. --Goldzahn (talk) 21:22, 6 June 2013 (UTC)

Thanks but we solved the problem by using item to store data for sources. Snipre (talk) 21:57, 6 June 2013 (UTC)

The table you linked on my talk page

Hey. I think we're misunderstanding each other. The leagues on that table are all different leagues. I'm talking about subdivisions of leagues, not leagues that are at a lower level than other leagues. TCN7JM 00:55, 11 June 2013 (UTC)

Edition of / subclass of

I have been checking the uses of subclass of (P279) and I cannot say that the meaning is the same as as the proposed edition of. "Subclass of" seems to link a reduced set with a much broader set, while in "edition of" there is not such a big step in the broadness of the items being linked. As for the bottom-up, I asked for a clarification, but the clarification needs to be clarified again :) --Micru (talk) 23:09, 12 June 2013 (UTC)

Thanks, Snipre for your collaboration! I will inform some property creator.--Micru (talk) 19:52, 14 June 2013 (UTC)

E number (P628)

The property E number (P628) that you supported is available now. --Tobias1984 (talk) 16:58, 14 June 2013 (UTC)

route of administration (P636)

The property route of administration (P636) that you supported is available now. --Tobias1984 (talk) 14:06, 17 June 2013 (UTC)

Property proposals

I'm not sure what gave you the impression that proposals are deleted, but that is not the case. Just because a property proposal doesn't have new comments doesn't mean it gets deleted. We just wait until some kind of consensus forms whether or not to create or not, we don't simply delete them. Regards, — Moe Epsilon 21:57, 18 June 2013 (UTC)

This was not the lack of comment which is the origin of my delete proposal but the general trend of using strat and end as qualifer: we have to use an unique way to idicate this kind of information in order to simplify the data extraction in wikipedia. Snipre (talk) 02:45, 19 June 2013 (UTC)

RfC on Wikidata's primary sorting property

You recently participated in a deletion discussion for P107 - main type (GND). The discussion has been closed, as it is clear that a resolution won't come from PfD, and an RfC has been opened on the matter at Wikidata:Requests for comment/Primary sorting property. You are invited to participate there. Please note that this is a mass delivered message, and that I will not see any replies you leave on this page.

Yours, Sven Manguard Wha? 18:24, 30 June 2013 (UTC)

Chemical formula

Hello Snipre. Please use [₀₁₂₃₄₅₆₇₈₉][⁰¹²³⁴⁵⁶⁷⁸⁹][⁺⁻]·☐ on chemical formulas, thx. Regards --Chris.urs-o (talk) 05:43, 10 July 2013 (UTC)

Re: Why ?

I deleted instance of (P31) chemical compound (Q11173) in ethanol (Q153) because ethanol (Q153) had the statement subclass of (P279) alcohols (Q156) which has subclass of (P279) organic compound (Q174211) which has subclass of (P279) chemical compound (Q11173). If A is subclass of B, that means all things that are A are also B. Since subclass is a transitive property, this is true for arbitrarily long subclass chains. In this specific case, the statement that ethanol is a chemical compound is a consequence of the subclass chain I gave, so specifying this directly in ethanol (Q153) would be redundant, and it is classified more precisely as an alcohol as well instead of just as a general chemical compound. Also, ethanol (Q153) is a class of chemical compounds, an instance of which is a specific amount of ethanol molecules somewhere in the world, so using subclass of (P279) is more correct than using instance of (P31). For more information about using subclass of (P279) and instance of (P31) please see Help:Basic_membership_properties. Hope that explains it. Silver hr (talk) 19:55, 12 July 2013 (UTC)

Re: Stop to use imported from Wikimedia project (P143)

Hello, should I simply stop adding them, or use an alternative property? Given that I'm importing data from zh.wikipedia, and most of data were originally from another database source, but might have been edited by zhwp users. Liangent (talk) 13:05, 16 July 2013 (UTC)

There are some discussion about the use of imported from Wikimedia project (P143) which was created before the definition of guidelines for sourcing. Better find a way to import data AND their sources from zh.wikipedia. I know this is not easy from programming point of view, but we can't source data in wikipedia with wikipedia as source. I can't force you to stop using imported from Wikimedia project (P143) right now but this will come soon, see Wikidata:Requests for comment/Sourcing requirements for bots. And just as advice statements like P107 (P107) or "instance of" do not need source: this are arbitrary or obivious definitions in most cases. Snipre (talk) 14:23, 16 July 2013 (UTC)

No there's no zhwiki article... it's there for WD:N#3. Details: we (or I?) made a database about that administrative division info on zhwiki with templates; articles and templates are extracting data from it with (complicated) template and parser function syntax and I'm simply moving that database to Wikidata. Liangent (talk) 15:29, 16 July 2013 (UTC)

Specifically we don't really want to (bot-)create divisions in the lowest level unless one is very famous (eg. some tourist attraction), or ... there'll be too many articles without much info in them. Liangent (talk) 15:37, 16 July 2013 (UTC)

UN number (P695)

The property UN number (P695) that you proposed is available now. --Tobias1984 (talk) 13:19, 18 July 2013 (UTC)

Kemler code (P700)

The property Kemler code (P700) that you proposed is available now. --Tobias1984 (talk) 13:24, 18 July 2013 (UTC)

Space group

Hey Snipre! I saw that you changed some of the space groups from "instance of = space group" to "subclass of = space group". I don't know which is better. I took instance of becasue it says "this item is a concrete object (instance) of this class, category or object group". So for the space groups as a whole I thought that e.g. space group 1 is an instance. It will save us some work if we both rethink it once more because I already added that statement to about 100 items ;). --Tobias1984 (talk) 22:04, 25 July 2013 (UTC)

Hi, as space group 1 is composed of different components, it can't be an instance of. If an item can group several other independent items, it can't be an instance of. You are right according to space group item but if you extend the classification in the other direction you see that space group X is a class of component so you can't have an item defined as a class and as an instance at the same time. This becomes a subclass. Snipre (talk) 22:38, 25 July 2013 (UTC)

I didn't know that an item that is an "instance of" can't at the same time be a "subclass of". Now I am wondering what kind of qualifiers we need to use to sort the symmetry groups, because now all of them will get 2 "subclass of" statements. --Tobias1984 (talk) 07:43, 26 July 2013 (UTC)

Be careful: you can have an instance of and a class for an intem for different classification. That's always a problem when you use a general classification based on instance of, subclass and class with specific properties.

Drugbank

Hey Snipre! Me again :) I saw that you changed the formatting on the DrugBank ID (P715). The DB is not part of the identifier and should therefore not be entered. There was a long discussion about this at Property_talk:P465. A lot of databases have these prefixes but I think we generally discourage saving them because they are just redundant data. --Tobias1984 (talk) 14:41, 26 July 2013 (UTC)

Sorry for the delay but as i saw no discussion in the talk page I assumed this was not discussed. About the format I prefer the format DBXXXX because you can use directly the string to create a link. Without the DB at the begining, you risk to see the numeric part reduced to its minimal significant digits: to be clear if you have DB0012 by putting the letters at the begining people will put the double 0. If you put only the number you risk to have 12 instead of 0012. This can be fixed later when you create the url but I don't like to see several possible formats for the same property. Yes, I know about the constraint check but if people can do the job right the first time they put the data this reduces the maintenance work. Snipre (talk) 13:07, 30 July 2013 (UTC)

P728 (P728)

The property P728 (P728) that you proposed is available now. --Tobias1984 (talk) 12:57, 30 July 2013 (UTC)

Re: Property proposal

Hello, Snipre. You have new messages at Ricordisamoa's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

--Ricordisamoa 22:36, 2 August 2013 (UTC)

Chembot

Bonjour,

J'ai commencé à regarder Chembot et les fichiers dont il a besoin.

J'ai une question : sera-t-il possible de spécifier des qualificatifs de propriétés ?

Pour illustrer ma question, voici un exemple dans les noble title (P97) de l'élément Ashot I of Armenia (Q463419).

Cordialement. Odejea (talk) 09:44, 5 August 2013 (UTC)

Oui, c'est possible, mais pas encore testé: il suffit d'ajouter le label du qualificatif (PXXX) dans le deuxième ligne. Une définition est ainsi composé d'une valeur (val), d'une source (ref) et tous les autres paramètres étant définis sous la même propriété sont considérés comme étant des qualificatifs. Si tu as un exemple tu peux me laisser les données sur ma page et je ferai un test. Le seul problème est pour l'instant les dates: je dois trouver un moyen de convertir une date du format excel dans le format wikidata. Snipre (talk) 10:07, 6 August 2013 (UTC)

Voila un exemple, à cette page : User:Odejea/Chembot. Les deux premiers tableaux sont ceux dont a besoin Chembot, les deux suivants sont des versions des deux premiers permettant de visualiser les valeurs, à des fins de vérification.

Pour les dates, comme ce ne sont que des années avant JC, elles sont sous cette forme 323 BCE, dont je ne pense pas qu'elles posent problèmes.

par contre je me rends compte de problèmes non prévus. Que faut-il faire quand :

• quand une propriété a deux valeurs possible ?
• deux références permettent de sourcer la valeur d'une propriété ?

Cordialement - Odejea (talk) 20:36, 6 August 2013 (UTC)

cela est un choix à faire: pour l'instant j'ai paramétré mon code pour une valeur avec référence. Autrement dit, si il n'y a pas de valeur ou si il existe une valeur sans référence, mon code supprime les données existantes et ajoute la valeur avec la référence fournie par le contributeur. Si la valeur dans l'élément est égale à celle du contibuteur et a déjà une référence, mon code n'ajoute rien, si la valeur de l'élélment est différente du contributeur et a une référence, mon code n'ajoute rien, mais crée un rapport qui un conflit.

Mon code est très conservateur et part du principe que peu de données sont actuellement disponibles et encore moins sourcées. Mais tout cela ne sont que des paramètres qui peuvent être modifiés.

Je te propose de déjà faire un test avec tes données avec le code actuel et on peut ensuite définir une approche pour les cas complexes. Snipre (talk) 21:11, 6 August 2013 (UTC)

Re-Bonjour

Je viens de modifier les tableaux (j'ai enlevé des infos non sourcés)

Par contre, il y a des propriétés qui me semble difficilement sourçables, comme P107 (P107) ou sex or gender (P21). Odejea (talk) 21:33, 6 August 2013 (UTC)

élément de codage des dates

Bonjour,

en faisant quelques expérimentations sur le Wikidata Sandbox (Q4115189), j'ai quelques idées sur le format wikidata des dates :

• tout date antérieure ou égale au 31/12/1582 est considérée comme julienne
• les dates sont converties en date grégorienne
• à partir de la date grégorienne, les dates sont à ce format : S0000000AAAA-MM-JJT00:00:00Z , où :
  • S vaut - si on est avant JC ou après JC (l'an 0 est codé avec un )
  • AAAA, l'année
  • MM, le mois
  • JJ, le jour
• l'année AAAA, sans précision de jour ou de mois, est codé : S0000000AAAA-01-01T00:00:00Z , où :
• Par contre, je n'ai pas compris comment wikidata fait la différence entre l'an 800 et le 28 décembre 799, qui sont tous les deux codés 00000000800-01-01T00:00:00Z :
  • diff entre deux années 800 : il voit que c'est la même date
  • diff entre l'an 800 et le 28 décembre 799 : il voit qu'il y a une différence

Cordialement --Odejea (talk) 09:26, 7 August 2013 (UTC)

Source for Riograndense Republic (Q162192)

I added the source to Riograndense Republic (Q162192). As I have not read the source textbook myself but just copied the details from English Wikipedia I then added 'imported from English Wikipedia' which you then deleted. I think this should be kept until someone checks the original actually supports the statement. Filceolaire (talk) 20:08, 5 August 2013 (UTC)

Wikidata:Communes de Hongrie - task force

Salut, on était un peu mal partis sur le Bistro, mais sache qu'en fait je te comprends car moi aussi je suis très à cheval sur les sources sur WP. Donc comme tu as de bonnes connaissances techniques si je comprends bien, ton apport sera le bienvenu dans le lien en titre, vers où a été renvoyée la discussion technique au lieu du Bistro. Cordialement : Oliv0 (talk) 11:08, 7 August 2013 (UTC)

LAU (P782)

The property LAU (P782) is now available. I saw that you participated in the discussion. --Tobias1984 (talk) 06:37, 8 August 2013 (UTC)

Chembot approved

Bonjour Snipre! I just approved Chembot. --Tobias1984 (talk) 16:00, 11 August 2013 (UTC)

significant event (P793)

The property significant event (P793) is available now. I saw that you participated in the discussion. --Tobias1984 (talk) 17:06, 13 August 2013 (UTC)

Deleting Proposals

Hi Snipre! Please don't delete proposals. It is better to move them to the archive or the corresponding task force page. I copied the 3 proposals to the ship task force Wikidata_talk:Ship_task_force#Deletion_of_some_properties_proposals. They have low activity right now, but they should be autonomous in their decisions to propose properties. --Tobias1984 (talk) 09:30, 18 August 2013 (UTC)

I archived proposals when there is something to archive. These proposals are not complete and there are no comment. And please don't say that contributors look at archive pages before creating a proposal, it is not the case. I put a comment at Wikidata:Ship task force so I think the main persons which are concerned are now informed. Snipre (talk) 09:34, 18 August 2013 (UTC)

You are certainly right about the completeness of the proposals. And leaving a message is very polite of you. On second thought you were probably right about deleting them. all the best, and stay busy. --Tobias1984 (talk) 09:45, 18 August 2013 (UTC)

decays to (P816)

The property decays to (P816) that you supported is available now.--Tobias1984 (talk) 21:48, 22 August 2013 (UTC)

decay mode (P817)

The property decay mode (P817) that you supported is available now. --Tobias1984 (talk) 21:48, 22 August 2013 (UTC)

Archiving talk, closing discussions

Hi Snipre,

If you archive manually content on pages other than user talk pages, please add the content to the corresponding archive. A short explanation of the closure of the discussion would be helpful too. --  Docu  at 11:49, 24 August 2013 (UTC)

Economic Statistics on Wikidata

Hi Snipre!

Would you possibly be interested to comment on these newly proposed properties on economics on Wikidata? Link is here: http://www.wikidata.org/wiki/Wikidata:Property_proposal/Term

Thanks, Mcnabber091 (talk) 04:58, 26 August 2013 (UTC)

New RFC

I have started a RFC on 'Place' related properties. I'm alerting you because of the discussion you started on Project chat. Filceolaire (talk) 17:44, 2 September 2013 (UTC)

TB

Hello, Snipre. You have new messages at Sven Manguard's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

Sven Manguard Wha? 19:51, 11 September 2013 (UTC)

P940 (P940)

Hello! This property you proposed has been created. --Danrok (talk) 00:33, 28 September 2013 (UTC)

Talkback

Hello, Snipre. You have new messages at John F. Lewis's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

Comment sourcer

Hello, merci d'apporter des sources sur wikidata. Pour sourcer avec un livre, il faut créer 2 éléments au minimum si il existe plusieurs éditions du livre, comme c'est le cas de ton livre. L'élément que tu as créé correspond à l'oeuvre et les données qui sont disponibles dans cet élément sont communs à toutes les versions et éditions de l'oeuvre. Mais cet élément n'est pas utilisé comme source dans la section référence: il faut un autre élément qui contienne les données spécifiques d'une édition comme l'éditeur, la date de publication, le lieu de publication ou l'ISBN. J'ai créé cet élément Nobel Prize Women in Science (Q15043316) et les différentes propriétés qui y sont liées. je te laisse le soin de compléter les données manquantes et si tu as d'autres questions n'hésites pas à me contacter. Snipre (talk) 21:24, 4 October 2013 (UTC)

Merci beaucoup pour ton aide. C'est vrai que c'est un peu obscur comme fonctionnement au début mais bon, ça va s'améliorer avec le temps. J'ai ajouté 2/3 informations dans l'objet d'édition. Par contre, je n'ai pas compris à quoi servait, fondamentalement, l'objet d'édition ? L'objet oeuvre Q15043031 ne contient pas de lien avec Q15043316 ; c'est uniquement l'éition qui est lié au livre (ce qui se comprend) mais du coup je ne comprends pas l'intérêt de séparer les deux. Pamputt (talk) 21:51, 4 October 2013 (UTC)

PS : merci pour la correction de l'ISBN Pamputt (talk) 21:52, 4 October 2013 (UTC)

Pourquoi créer un élément pour chaque édition et un élément pour une oeuvre: si on veut donner des données très précises comme l'éditeur ou la date de publication, on ne peut pas avoir un seul élément en cas de plusieurs éditions d'une même oeuvre, car tout serait mélangé dans un seul élément. Et la raison d'un élément pour l'oeuvre se trouve du côté de wikipedia qui souvent a un article pour l'oeuvre, mais pas pour les éditions. Il faut au moins un lien entre l'élément oeuvre et l'élément édition pour pouvoir réunir toutes les édition d'une oeuvre si on veut faire une telle recherche, comme pour lister toutes les éditions sur la page wikipédia de l'oeuvre. En espérant t'avoir renseigné. Snipre (talk) 22:05, 4 October 2013 (UTC)

Removing OCLC Properties

Hello Snipre,

I noticed you have been removing some of the OCLC properties that my bot has been adding (see diff). I was would like to know why you are doing this? I feel that is the right thing to do, as I got permission to do it at wd:RFBOT. Maximilianklein (talk) 18:17, 23 October 2013 (UTC)

Yes I removed the property: but you will understand why I did that if you look at the item. The item The King of Rome (Q863) is about a pigeon not a book. So if you want to add data about the book on the pigeon, please create the appropriate item. An item is not a category which mix different subjects with a common topic. Snipre (talk) 18:25, 23 October 2013 (UTC)

Danke

Hallo Snipre, nur mal ein kurzes herzliches Danke für deine Arbeit, gerade auch in Hinblick auf Help:Sources. Wenn wir über Details, wie die beste Art, die ISBN einzubauen, diskutieren, geht leicht unter, dass ich die Mühen, die du, Max und andere sich hier machen, wirklich schätze. Cheers --Kolja21 (talk) 21:56, 25 October 2013 (UTC)

How to get dat a out of wikidata?

I have arrived here from the archived talk about the en:template:infobox element [1]. I am working on that one, long term. I understand the topic is on hold because we need a number data type here. All fine.

My question now is: how could one use wikidata in say enwiki? How can one get existing data (and source) out of wikidata onto a wiki content page? Any examples? en:User talk:DePiep -08:15, 27 October 2013 (UTC)

What you need is to know lua programming language: an infobox is very complex and the use of the template invoke is not sufficient.

But you can try first: go in the article you want to have wikidata data and be sure that the corresponding data in available in the wikidata item linked to your article. Put the code line {{#property:Pxxx}} where Pxxx is the property used in the wikidata item to store the value you want, save and normally the value is displayed in the article text. Now you can take the infobox template you want to improve and in each field of the template you can add the corresponding property in the code line.

{{XXX |xxx1 = {{#property:Pyyy1}} |xxx2 = {{#property:Pyyy2}} |xxx3 = {{#property:Pyyy3}} }}

But this code can't held the existing data in the infobox and can't manage sources or data selection when multiple data values for the same property are present. And here lua plays a major role.

In WP:fr we already develop some lua modules to generate some infoboxes and I can only propose you to look at w:fr:Undéc-1-ène to see the potential of data extraction from wikidata: the first infobox is filled only with data from wikidata (look at the code source of the article: the first infobox is defined in the code only by {{Utilisateur:Snipre/bac à sable | nom = Undéc-1-ène }} (my test code).

But in reality the system behing is very complex: to be able to generate this first infobox we are using different modules: see w:fr:Module:Wikidata, w:fr:Module:Infobox, w:fr:Module:InfoboxBuilder, w:fr:Module:InfoboxBuilder/Composé chimique and w:fr:Utilisateur:Snipre/bac à sable. This is still draft code and if somebody with very good skills in lua programming can simplify a little this structure this can help a lot for further development.

Just a detail: for the full data extraction we need to wait for the number datatype but for the possibilité to extract data for an wikidata item which is different from the one linked to the wikipedia article. This is already known by the developers and this is on their priority list (see bugzilla:47930) but this is an huge development task. Hope I answered some of your questions and if you are ready to develop some feature in WP:en, I will be interested to follow your improvements. I have an account on WP:en : w:user:Snipre. Snipre (talk) 10:09, 27 October 2013 (UTC)

atomic number (P1086)

Just created atomic number (P1086). --Tobias1984 (talk) 07:11, 31 January 2014 (UTC)

Mohs' hardness (P1088)

Just created Mohs' hardness (P1088). --Tobias1984 (talk) 07:36, 31 January 2014 (UTC)

electronegativity (P1108)

Just created electronegativity (P1108). --Tobias1984 (talk) 17:22, 31 January 2014 (UTC)

refractive index (P1109)

Just created refractive index (P1109). --Tobias1984 (talk) 17:28, 31 January 2014 (UTC)

pKa (P1117)

Just created pKa (P1117). --Tobias1984 (talk) 12:34, 2 February 2014 (UTC)

Gini coefficient (P1125)

Just created Gini coefficient (P1125).--Tobias1984 (talk) 14:05, 2 February 2014 (UTC)

Improved Global Economic Map property page

I have made some large edits to the Global Economic Map Wikidata taskforce page here: https://www.wikidata.org/wiki/Wikidata:Global_Economic_Map_task_force. All of the data used in the G.E.M. are mapped out by their sources on the property page. I also proposed a few new properties here: https://www.wikidata.org/wiki/Wikidata:Property_proposal/Economics#Population_growth. I am currently looking for community support on my grant proposal here (https://meta.wikimedia.org/wiki/Grants:IEG/Global_Economic_Map) and if you had anything to say that would be awesome. Thank you Mcnabber091 (talk) 07:37, 18 April 2014 (UTC)

ChemID project and ChemSpider

Hi Snipre,

You seem to be doing a lot of the work around the ChemID Initiative, and I wanted to know if it is possible to arrange some sort of online chat/skype to learn more about how you are doing this - I also work on the Royal Society of Chemistry's ChemSpider database and would like to look into ways we could contribute to the project. If you send an email to the ChemSpider inbox (details in point two on the page [FAQs]). I hope to hear from you, --The chemistds (talk) 12:17, 12 June 2014 (UTC)

@The chemistds: Hello, I was doing a lot of work but since 4 months I didn't do nothing because I had to finish my PhD thesis. So now I can take a little more time to progress in this project. From now I just have raw data and I have to treat them in order to organise them in table. I extracted data from en, de and fr:WP (name of article for each chemical having an infobox with the CAS number and the PubChem ID if they exist in the infobox and finally the Q number from wikidata). My idea is now to code a script to put this data in an unique table using the PubChem ID as key identifier. Then to analyse the table to find missing information (Q number or PubChem ID) or duplicate and to correct them. Second step is to extract data from free databases and to add their ID number to the table using again PubChem ID as matching key. After that, third step, I was thinking about importation of that table into Wikidata. But I have a problem about the licence. For PubChem ID, it is OK, but for other databases, the licence is not compatible, CC-BY instead of CC0, so now I prefer to avoid the importation. So when I will reach the second step I can include the Spider ID or I can compare my table witho their data if I can access to their database (a dump is enough) then I will need some help to treat inconsistencies.

If you want to help me I think the best would to allow wikidata to use the Spiderchem ID under the CC0 licence. I propose you to enter in contact with the wikidata development team and to propose them an agreement based on OTRS system used in common to allow the use of Spiderchem ID without any restriction. Snipre (talk) 14:51, 12 June 2014 (UTC)

Relecture demandée

Bonjour,

Même si je commence à bien connaître Wikidata, il reste encore de nombreux doutes. Pourrais-tu relire Museum of Brittany (Q3329701) et me dire si 1. il y a des fautes dans mes derniers ajouts 2. si tu vois des informations manquantes qui pourrais être ajoutées. Pour le directeur, quelle est la propriété à utiliser ? chief executive officer (P169) ou director / manager (P1037) ou une autre ?

Cdlt, VIGNERON (talk) 11:45, 25 June 2014 (UTC)

@VIGNERON: Désolé du retard. Cela semble OK, mais je ne suis pas spécialiste pour tous les sujets. Je préfère directeur. Snipre (talk) 11:26, 17 July 2014 (UTC)

Help:Sources

Hello, about my edit in Help:Sources: generally volume (P478) and DOI (P356) properties have to be added to an article item, not to a publication item. Maybe them can be moved to 8. of the list? "author (P50), issue (P433), publication date (P577), page(s) (P304), P387 (P387), or chapter (P792)" can be modified in this way "author (P50), issue (P433), publication date (P577), page(s) (P304), volume (P478), DOI (P356), P387 (P387), or chapter (P792)"? --Sbisolo (talk) 09:36, 11 August 2014 (UTC)

@Sbisolo: As I said in the comment there is a misunderstanding about the concept of article in the point 4 of that paragraph: Point 4 is true for all articles (in scientific journals or in newspaper as well as for magazine). But only scientific articles have a DOI and few magazines of newspapers have a volume. Point 4 specifies "at least" meaning we want to see that properties in all items defining articles (without any difference between scientific, newspaper or magazine articles).

So we can speak about the differences between an article and a publication but DOI and volume properties are specific properties and are not so necessary than the properties named in point 4 so they can be added in a second step.

The problem is points 7 and 8 which shouldn't be divided but merged into one point. Snipre (talk) 14:03, 11 August 2014 (UTC)

Sorry after reading carefully the text we don't have to merge point 7 and 8. You are right: we should put everything in point 8. Snipre (talk) 14:09, 11 August 2014 (UTC)

Physics WikiProject

New userbox at Wikidata:Userboxes. It is also translatable. --Tobias1984 (talk) 12:30, 13 August 2014 (UTC)

UNII Import

Hi Snipre, I'm from Open Food Facts and also a Wikidata contributor. I've just made a Bot Import proposal regarding the UNII property. Could you provide some feedback ? --Teolemon (talk) 13:48, 21 November 2014 (UTC)

@Teolemon: If I can help, yes. But what kind of feedback do you need ? Snipre (talk) 14:04, 21 November 2014 (UTC)

PS: Tu peux écrire en French ;-) Snipre (talk) 14:05, 21 November 2014 (UTC)

Collaboration possible

Bonjour

Je suis en relation avec un laboratoire de chimie analytique qui désire réutiliser les données de Wikidata. Est-ce que cela vous intéresse de participer à une conf call à l'occasion sur la question ?

Sinon, j'ai proposé dans le WP5 une tâche 5.2 Bis 2 qui peut être une première piste de collaboration (même si le projet n'est pas accepté).

--Karima Rafes (talk) 19:04, 15 December 2014 (UTC)

@Karima Rafes: Pourquoi pas. Je recherche des personnes qui veulent faire advancer le domaine de la chimie sur Wikidata, donc je serai heureux de pouvoir travailler en équipe. Pour la conference, cela serait sur Skype ? Snipre (talk) 08:15, 16 December 2014 (UTC)

Pour info, j'ai déjà essayé de lancer un projet sur les identifiants des molecules (Wikidata:WikiProject Chemistry/ChemID). A voir si cela entre dans votre perspective ou si vous avez d'autres idées. Snipre (talk) 08:19, 16 December 2014 (UTC)

Ok, j'organiserai un skype vers fin janvier. Merci pour le lien, je vais l'envoyer à la chercheuse qui porte le projet dans son laboratoire de chimie analytique (et je vais me remettre un peu à la chimie d'ici la réunion) A bientôt --Karima Rafes (talk) 18:56, 16 December 2014 (UTC)

@Karima Rafes: Comme matière à discussion, je pense qu'il serait bien de mettre à plat les objectifs que votre labo aimerait voir réaliser: Wikidata est encore à un stade de développement et il y a encore beaucoup de flou quant à la structure des données. Cela offre des portes ouvertes, mais cela implique beaucoup de travail avant de voir un résultat concret. Snipre (talk) 08:24, 17 December 2014 (UTC)

Thanks for reading my edit summary "why section edits are not allowed in WikiData?"

For some sections in article Help:Sources, when I click on [edit] (e.g. [2]), I get this:

https://www.wikidata.org/w/index.php?title=Help:Sources&action=edit&section=7

which says:

Cannot find section
You tried to edit a section that does not exist. It may have been moved or deleted while you were viewing the page.
Return to Help:Sources.

Is this a bug or intended MediaWiki feature? The only thing which could interfere with headers is this XHTML: <!--T:4--><translate> Kazkaskazkasako (talk) 15:53, 18 December 2014 (UTC)

@Kazkaskazkasako: Sorry but I don't have any problem to edit using section button. Snipre (talk) 11:16, 22 December 2014 (UTC)

Wikidata can not find section 7 here too. --Chris.urs-o (talk) 19:17, 22 December 2014 (UTC)

@Kazkaskazkasako, Chris.urs-o: Sorry, I was trying another section which can be edited. The main reason in my opinion is the translation labelling inside the wikicode: depending on the way the sections are marked for the translation, you can't edit them directly. Snipre (talk) 23:42, 22 December 2014 (UTC)

Liste des propriétés pour les communes de Suisse

Salut !

Je suis en train de réfléchir à la future modification des Infobox pour les communes de Suisse en utilisant les propriétés Wikidata au lieu des fichiers de données locaux. Pour cela, il faudra s'assurer que les informations nécessaires soient toutes disponibles ici.

D'où ma question : y-a-t'il un moyen simple de voir quelles propriétés sont renseignées pour quelles communes ? Cela devrait permettre de compléter les infos manquantes assez rapidement.

Merci d'avance pour ton retour, Manoillon (talk) 09:15, 5 January 2015 (UTC)

PS : Pour info, voici les propriétés auxquelles j'ai pensé pour remplir l'infobox sur FR : pays, canton, district, arrondissement, coordonnées géographiques, numéro OFS, nom du fichier représentant le blason, langue(s) officielle(s), site web, nom du syndic/maire et nom de la fonction, code(s) postal (postaux), altitude, nom du fichier représentant la carte, nombre d'habitants et date de recensement (et historique de la démographie ?), surface

Manoillon Hello, tu arrives un peu trop vite avec ta réflexion: en effet, il manque encore des outils pour pouvoir voir des infoboxes utiliser WD d'une manière large. Il y a 2 obstacles à lever: la création du type de donnée quantité avec unité (actuellement nous avons uniquement quantité sans unité). Cela veut dire que toutes les données qui utilisent des unités comme les altitudes et les surfaces ne sont pas encore disponibles. On espère voir ce problem résolu entre février et mars 2015.

Le 2ème obstable est l'accès à plusieurs éléments depuis un article de WP: les informations sont cloisonnées sur Wikidata via des éléments séparés. On a un élément pour la Suisse, un autre pour le canton, un autre pour une commune. Le problème, c'est qu'actuellement depuis l'article Lausanne de WP, tu ne peux accéder qu'aux données de l'élément Lausanne de WD, et non à celui sur le canton de Vaud ou sur celui de la Suisse.

Tu vas me dire que cela n'est pas un gros problem si toutes les informations sont sur l'élément Lausanne, mais ce n'est pas le cas. Par exemple, on ne met pas le nom du pays dans les éléments des communes, car on structure les données: la commune de Lausanne se situe dans le canton de Vaud, donc on ajoute dans l'élément que le lien vers le niveau administratif supérieur et non pas tous les niveaux genre pays ou régions.

L'équipe de développement travaille sur la question et on espère résoudre ce problème en 2015.

Concernant les propriétés, on n'a malheureusement pas encore de catégorisation des propriétés pour pouvoir créer une liste. Mais plusieurs projet ont fait le travail et je te propose decréer un projet Suisse histoire de pouvoir mettre cette liste en place. Voici les exemples des projets hongrois et français.

Wikidata:WikiProject Suisse est lancé. On peut commencer à s'organiser sur la manière dont on veut stocker les données.

Ah oui, dernier détail, le problème de license: les données de WD sont en CC0, ce qu veut dire que l'on va avoir des problèmes si on importe les données depuis l'office fédérale de la statistique: en CC0, on n'a pas besoin de citer la source, mais le plus gros problème vient de l'aspect commercial, puisque l'office de la statistique demande une autorisation écrite pour tout usage commercial alors que WD peut être utilisé pour des applications commerciales sans demande. Il va falloir peut-être prendre contact avec l'office pour obtenir une dérogation: du boulot pour le chapter Suisse. Snipre (talk) 10:07, 5 January 2015 (UTC)

Merci pour ton retour. Concernant le cloisonnement des éléments, je note que (par exemple) sur Lausanne, je trouve aussi bien le pays (P17) que le district et le canton (deux fois P131, ce qui va probablement poser un problème), ce qui semble en contradiction avec ton 2ème obstacle. Manoillon (talk) 12:57, 5 January 2015 (UTC)

En Suisse, cela ne pose pas trop de problème, mais dès que la structure administrative devient plus complexe, on ne peut plus intégrer toutes les infos au niveau le plus bas. La bonne pratique est de mettre les informations qu'une seule fois et non de multiplier les infos, le cas que tu soulèves est typiquement un des trucs utilisés pour reésoudre le problème de l'accès aux données. Snipre (talk) 15:46, 5 January 2015 (UTC)

Talkback

Hello, Snipre. You have new messages at The chemistds's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

2015-04 Category:Properties by subject

About Wikidata:Project chat/Archive/2015/04#Stupid panels, see Category:Properties by subject. Visite fortuitement prolongée (talk) 20:33, 17 April 2015 (UTC)

Cyclisme et Wikidata

Le Module:Infobox/Descriptif course cycliste, bien que perfectible, est désormais utilisé sur la Wikipédia francophone, quelques articles utilisent déjà les données de Wikidata. Merci pour tes conseils. Jérémy-Günther-Heinz Jähnick (talk) 12:38, 30 May 2015 (UTC)

Salut. Je discute avec Anthony59999 des différents types de classements, et on en est venus à parler de coureurs se voyant décerner le prix de la combativité, ou celui de la supercombativité. Du coup, je pensais faire une requête dans le module précité pour qu'un coureur X soit winner (P1346), et en ajoutant un qualificatif prix de la combativité par exemple, mais je ne sais pas quelle propriété utiliser. Je pensais initialement à topic's main category (P910), mais je me dis qu'il y a peut-être quelque chose de plus adapter pour indiquer qu'un vainqueur gagne tel ou tel prix. Jérémy-Günther-Heinz Jähnick (talk) 10:08, 2 June 2015 (UTC)

Sign

You forget to sign in this edit. This makes other users unable to know the reason to close the request, and who closed the request.--GZWDer (talk) 10:39, 4 June 2015 (UTC)

Traduction

Je suis pas vraiment sûr qu'on puisse traduire «implement» par «importer». En général on parle d'import pour un robot qui copie toute une base de données … Je suis pas très sûr non plus pour «mettre en œuvre les données» qui sonne pas bien (les données, on les recueille, on les transmet, on les copie, on le met pas en œuvre … mais comme je suis pas sûr du sens original non plus je sais par trop. Ce qu'on peut implanter, c'est une spec ou un schéma d'organisation par contre. Un avis ? author  TomT0m / talk page 19:24, 12 August 2015 (UTC)

Use of scholarly article (Q13442814) with instance of (P31)

Salut Snipre,

Thanks for answering me on Help_Talk:Sources. I actually have a much more basic question : is there a community recommendation regarding the use of scholarly article (Q13442814) vs. article (Q191067) with instance of (P31) ? From what I've seen some people use the former (Daniel Mietchen, Infovarius, Vlsergey) and others the latter (AS, Achim Raschka). There's also some discussion at P31 suggesting that the lowest class (i.e. scientific article) should be prefered, although there's also some relevant discussion regarding the class items that exist to make the link with Wikipedia rather than for internal reasons. As you may have seen I am relatively new on WD so there are still things that look very weird to me, and in many cases I wasn't able to the related explanations/recommendations.

Cheers and sorry for the trouble,

Tinm (talk) 19:05, 31 August 2015 (UTC)

Human Metabolome Database ID (P2057)

Human Metabolome Database ID (P2057) is ready. --Tobias1984 (talk) 10:18, 13 September 2015 (UTC)

HSDB ID (P2062)

HSDB ID (P2062) is ready. --Tobias1984 (talk) 12:55, 13 September 2015 (UTC)

Isomères

Bonjour Snipre,

Comment doit-on faire des isomères? Il ya ici racemate (Q467717), enantiomers mixture (Q1339942) et group of isomeric entities (Q15711994)--Kopiersperre (talk) 16:16, 13 September 2015 (UTC)

@Kopiersperre: Pour le label, c'est égal pourvu que l'on fasse la distinction entre un isomèe et le mélange correspondant. Pour la classification, il faut utiliser subclass of (P279): chemical compound (Q11173). group of isomeric entities (Q15711994) n'est pas spécifique au racémate, on l'utilise aussi pour la famille des alcools, des acides carboxyliques. Pour distinguer racemate (Q467717) de enantiomers mixture (Q1339942), il faut connaître la composition. racemate (Q467717) est une sous-classe de enantiomers mixture (Q1339942). Donc mieux vaut utiliser enantiomers mixture (Q1339942), on peut toujours afiner plus tard. Snipre (talk) 19:24, 13 September 2015 (UTC)

Je ne suis pas d'accord avec vous de P279:Q11173. Selon Help:Basic membership properties, applique instance of (P31) pour les produits chimiques. Svp corriger mon français.--Kopiersperre (talk) 08:31, 14 September 2015 (UTC)

@Kopiersperre: Better write in English. My German is not very good (Wortschatz) but if necessary you can use it. You can't use instance of (P31):chemical compound (Q11173) for mixture of chemicals because you can't have for the mixtures instance of (P31):chemical compound (Q11173) and for the individual components of the mixture instance of (P31):chemical compound (Q11173). Especially when you consider the link you can create between the mixture and the individual components.

Just look at that example:

Item A is a mixture of item B and C.

Item A = instance of (P31):chemical compound (Q11173), item B, instance of (P31):chemical compound (Q11173), item C = instance of (P31):chemical compound (Q11173)

But items B and C are instance of (P31):item A (item B is a special case of mixture where item B is 100% and item C is 0%, the inverse for item C). So how can be 3 instances linked together ? Three instances of the same class (chemical compound (Q11173)) should be at the same level in the classification tree and can't have then a relation instance between them. I didn't read Help:Basic membership properties because this kind of documentation was done by one person and pretend to rule all classification schemes of WD. There is right now a RfC about this document to obtain a larger acceptance. Snipre (talk) 09:07, 14 September 2015 (UTC)

molar fusion enthalpy (P2066)

molar fusion enthalpy (P2066) is ready. --Tobias1984 (talk) 13:32, 14 September 2015 (UTC)

thermal conductivity (P2068)

thermal conductivity (P2068) is ready. --Tobias1984 (talk) 17:28, 14 September 2015 (UTC)

boiling point (P2102)

boiling point (P2102) is ready. --Tobias1984 (talk) 20:54, 25 September 2015 (UTC)

decomposition point (P2107)

decomposition point (P2107) is ready. --Tobias1984 (talk) 11:02, 26 September 2015 (UTC)

p2128

flash point (P2128) is ready. --Tobias1984 (talk) 21:43, 1 October 2015 (UTC)

Hello, i added the the constraints, russian label and description for this property you requested. Please, check the property definition. And why are so strange units and template parameter noted at talk page? It must be FlashPt in Chembox and °C... -- Sergey kudryavtsev (talk) 14:50, 2 October 2015 (UTC)

@Tobias1984: Is a mistake while creation of P2128? (There is nothing similar in Wikidata:Property proposal/Natural science.) -- Sergey kudryavtsev (talk) 15:00, 2 October 2015 (UTC)

I correct this. Provide an example, please. -- Sergey kudryavtsev (talk) 15:08, 2 October 2015 (UTC)

solubility

solubility (P2177) and solvent (P2178) are ready. --Tobias1984 (talk) 13:25, 5 October 2015 (UTC)

Removal of P31 from Q214917 & Q11774202

You removed instance of (P31) from Q214917 & Q11774202 on november 6th. Why did you remove this? It caused a lot of constraint violations for P106. Mbch331 (talk) 15:39, 13 November 2015 (UTC)

@Mbch331: Because playwright (Q214917) is a subclass of writer (Q36180) which is a subclass of author (Q482980) which is a subclass of creator (Q2500638) which is an instance of profession (Q28640) (Here we have a problem: how a subclass can be an instance of profession ?). So playwright (Q214917) is already linked to profession (Q28640) so no need to do it twice. Snipre (talk) 16:29, 13 November 2015 (UTC)

IMHO being a creator is not a profession. A profession is something what you do professionally.--Kopiersperre (talk) 17:05, 13 November 2015 (UTC)

@Kopiersperre: I have a similar feeling but I didn't want to change the classification: I just corrected the classification structure. We can define author (Q482980) subclass of profession (Q28640), this is not a problem for me. Snipre (talk) 17:09, 13 November 2015 (UTC)

I think that's better and it will resolve the constraint violations. Mbch331 (talk) 18:15, 13 November 2015 (UTC)

Apology

My apologies; I've no idea how this happened; nor was I aware that it had done so. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:29, 7 January 2016 (UTC)

Removal the PubChem CID from lindane?

Hi Snipre, I was wondering why you removed the PubChem CID from lindane? I checked it, but the ID looks correct to me... Egon Willighagen (talk) 06:46, 16 January 2016 (UTC)

@Egon Willighagen: lindane (Q282003) is a stereoisomer of 1,2,3,4,5,6-hexachlorocyclohexane (Q424459) and the PubChem ID 727 has no stereoisomer characteristics. So this ID should be used only for 1,2,3,4,5,6-hexachlorocyclohexane (Q424459). Don't look at the names in PubChem: they don't correspond to the structure most of the time because they were added by people and not by the PubChem team and those people don't take care of isomer problems most of the time. Snipre (talk) 08:08, 16 January 2016 (UTC)

@Snipre: Carp. You're right. Embarrassing that I overlooked the hydrogens. Yeah, you wonder how that could happened (same here; where is my coffee). BTW, I fully agree that this problem is huge; have you discussed with the PubChem team how to handle this stereo problem in Wikidata? E.g. how to differentiate mixtures from unknown and from unspecified stereochemistry? (PS, I was scanning the recent changes because the number of CAS numbers has been fluctuating, but this is explained by people doing curation, so that is great!) Egon Willighagen (talk) 08:41, 16 January 2016 (UTC)

(4S,5S)-5'-methylthioadenosine (Q3599216)

Hi Snipre,

is it normal to use & #39; for '?--Kopiersperre (talk) 09:40, 19 January 2016 (UTC)

No. Corrected. Snipre (talk) 09:57, 19 January 2016 (UTC)

Defining Formula

Hi Snipre,

at the property proposal for 'defining formula' (here: https://www.wikidata.org/wiki/Wikidata:Property_proposal/Natural_science#defining_formula) you opposed because you want user to use property TeX string instead. But the aim of our proposed property is another. There is a new data type 'Mathematical Expression' we developed in the last month. It launched last week and it allows user to add mathematical expressions exclusively - as TeX-Strings. But in contrast to the property 'TeX string' (which is just a plain string) the new data type comes with a validator to permit formula only. And it has also a formatter to show the formula rendered by the MathML extension. That's a more special story than just having mathematical expression as plain TeX strings without any functionality behind.

If you agree with it please revoke you oppose at https://www.wikidata.org/wiki/Wikidata:Property_proposal/Natural_science#defining_formula so we become able use the new data type.

Kind regards --JulianHilbig (talk) 19:46, 14 February 2016 (UTC)

Validating DrugBank UniprotIDs with Uniprot

Hi Snipre,

I posted a while ago about porting drug-protein interaction data from DrugBank to Wikidata and you raised some concerns about the validity of UniprotIDs from DrugBank. I am currently working to validate this information. Do you happen to have any examples of UniprotIDs on DrugBank that don't exist?

Thank you for your help.

Best,

Crowegian

Crowegian (talk)

@Crowegian: I just gave my opinion about DrugBank: you don't need to take account of it to do data import. But my concern about DrugBank is that I found many mistakes in that databases for CAS numbers. I mentioned those errors to the database team and they corrected them fast. But I was disapointed to discover that over 100 chemicals I analyzed around 10 had wrong CAS numbers based on DrugBank. I don't know about UniprotIDs but I can only advise you to implement a constraint report and to check it after the importation. For me DrugBank is not the first source for chemical data and especially for chemical identifiers. I prefer to use ChemIDplus which has a lower error rate. Snipre (talk) 21:10, 22 May 2016 (UTC)

@Snipre: Got it. Thank you for your input. I will be sure to check whatever data I plan on importing against other sources, like ChemIDplus. Crowegian

Page count

Hi -- thanks for fixing the page count on The Time Machines (first paperback edition) (Q24747339). When I created it I couldn't see how to remove the /-1; can you tell me what I need to do in the future to put in the upper and lower bound? Mike Christie (talk) 12:39, 24 June 2016 (UTC)

@Mike Christie: Save your value and then edit again by setting 1 to 0 for the uncertainty. But you can try to fix it already at the first edition by typing 0 after your number. Check if possible. Snipre (talk) 15:04, 24 June 2016 (UTC)

Got it; thanks. Mike Christie (talk) 00:42, 26 June 2016 (UTC)

Other users' unsigned talk page posts

You may find {{Unsigned2}} useful. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:33, 20 September 2016 (UTC)

heptachlor expoide

Hi Snipre,

why haven't you merged cis-heptachlor exo-epoxide (Q27094563) and heptachlor epoxide (Q27116272)?--Kopiersperre (talk) 09:56, 10 October 2016 (UTC)

@kopiersperre: Because one can be used for the chemical without any defined stereoisomers and the other one for a complete defined stereoisomer. I prefer to avoid to merge now 2 items and to create later new items for each different stereoisomer. Snipre (talk) 10:03, 10 October 2016 (UTC)

Loss of information

Hi Snipre. Please avoid loss of information. I re-added what you had removed. Please do this yourself for (Z)-hexadec-9-en-1-ol (Q27145425) (information removed from (E)-hexadec-9-en-1-ol (Q20054539)). --Leyo 07:54, 2 December 2016 (UTC)

(E)-azobenzene (Q410056)

Hi Snipre. The item does not correspond to the Wikipedia articles (de, en, fr at least) anymore after your changes. They describe both isomers, how they photoisomerize etc. Please fix it. My solution would be to undo your changes. --Leyo 08:13, 2 December 2016 (UTC) PS. You removed a correct German alias (Benzen is an alias of Benzol).

@Leyo: I don't take care of the interwikis: I use the data present in the item. Wikipedias deal with their articles as they want. WD is dealing with its own data and splitting so comparing WD items with WP articles is often just a nightmare because topic is often not the same. Then I analyze only the data available in the items, data importing by bot without any check before importation. I have enough work to curate the data in the different items. Perhaps a different choice in data splitting can be made in some cases but I don't have the time to find always the best solution: I have to curate thousand of items due to bad import of data and multiple creation of duplicates due to not sufficient analysis of existing data. Just have a look at one constraint report once, like Wikidata:Database_reports/Constraint_violations/P231 to understand the consequence of bad importations.

So to come back to (E)-azobenzene (Q410056) I just deleted all data related to mixture of isomers as the majority of current data in the items were about the trans isomer. There is an specific item for the mixture of isomers which is azobenzene (Q27444428). So if you want you can transfer the interwikis dealing with the mixture to that item. Due to identifiers problems we have to split data about each isomer and create different items for the mixture (or the inverse). The only criteria are the original data in the items which are coming most of the time from WPs. This is the only way for WD to keep a coherent system. If WP articles want to treat mixture of isomers and specific isomers in one article I let them the choice to define which item should be linked to their article. Snipre (talk) 09:20, 6 December 2016 (UTC)

You should have done it the other way round, i.e. creating a new item for the trans isomer leaving (E)-azobenzene (Q410056) to match the articles. Anyway, I moved the sitelinks to azobenzene (Q27444428) and re-added the most important information that was lost. --Leyo 20:14, 6 December 2016 (UTC)

@Leyo: No, because I used the data in the item to assess what is the concept behind the data. The problem is that most of the data were imported from the english article and especially from its infobox which is focused on the trans isomer. Then other contributors or bots linked this item with the wrong articles without checking what was the data in the item. So please consider the real order of the events and don't come to me complaining about my work when other contributors are responsible of the mess.

Again interwikis are not the reference to define the concept because WP articles are often mixing different concepts and because interwikis are often not correctly created. So only the data should be considered to define the concept and when the data are mixed I choose the concept represented by the majority of the data and then I deleted the rest.

To finish, interwikis is not the responsability of WD, each WP chooses the item corresponding to their article. Perhaps the choice of linking the article of the isomer mixture to the trans isomer item was a rational choice and your decision to move all interwikis is a complete lack of respect of this choice. I let you choose what is better or worst between deleting wrong data and changing interwikis, for me the choice is done. Snipre (talk) 12:54, 8 December 2016 (UTC)

Just a last word about lost data. No data are not lost because most of the data were imported by bot and that bot will redo the importation later. The problem with that bot is its way to match WD items with external databases, mainly based on wrong data in WD and in the databases. So I won't loose time to import data when a bot can do that in faster way and I will focus on what is worthful: data curation. Why do I have to do bot work when more critical contribution is needed ? Snipre (talk) 13:03, 8 December 2016 (UTC)

Unused properties

This is a kind reminder that the following properties were created more than six months ago: lowest-observed-adverse-effect level (P2718), no-observed-adverse-effect level (P2717), acceptable daily intake (P2542), minimal lethal concentration (P2710), UN packaging group (P876), median lethal concentration (LC50) (P2712). As of today, these properties are used on less than five items. As the proposer of these properties you probably want to change the unfortunate situation by adding a few statements to items. --Pasleim (talk) 19:28, 17 January 2017 (UTC)

@Pasleim: Is there any formal decision about deadline for property use ? Sorry but currently I am working mainly on data curation because bot importations are often bad so I prefer to identify clearly items before to start statements addition to the wrong items. We can't do everything at the same. Snipre (talk) 12:34, 23 January 2017 (UTC)

There is no formal decision about a deadline and I don't have the plan to report these properties for deletion. I just wanted to inform you about the existence of these properties in case you should have never been informed about success of your proposals or in case you have forgotten it. --Pasleim (talk) 12:40, 23 January 2017 (UTC)

Sulphur element

Hi Snipre, I was adding EPA CompTox Dashboard identifiers and noted that it added an ID for sulfur (Q682). It picked it up based on the InChI and InChIKey for H2S... which is not the elemental sulphur. --Egon Willighagen (talk) 19:46, 6 February 2017 (UTC)

Un truc pour dupliquer un libellé

Salut Snipre, en voyant ceci, je me suis dit que l'outil nameGuzzler (libellé "VIP's labels") pourrait t'intéresser. Tu peux configurer tes langues préférées comme ceci. Cheers -- LaddΩ chat ;) 15:19, 13 May 2017 (UTC)

do we need a property  ?

Just saw your removal : https://www.wikidata.org/w/index.php?title=Q11344&diff=488772131&oldid=485904559

At first I did not understand, but after a while I realized what is expressed by this statement is meant to be : « instances of this class are part of instances of that class». Something similar to «metasubclass of» but for parts. I wonder if this is necessary however as there is no ambiguity that they are on the same metaclass level. For exbmple I’d have no problem to use th esame property to express :

• <My arm> <part of> <my body> (instances level)
• <arm> <part of> <body> (class level) Obviously here it does not mean that the class of all arms is a part of the class of all bodies. It means that an instances of arm is a part of an instance of body.

I think we can do the same at the metaclass level. What do you think ? author  TomT0m / talk page 18:37, 23 May 2017 (UTC)

Chembot

Your bot has been listed at Wikidata:Requests for permissions/Removal/Inactive bot accounts as being inactive for over two years. As a housekeeping measure it's proposed to remove the bot flag from inactive bot accounts, unless you expect the bot will be operated again in the near future. If you consent to the removal of the bot flag (or do not reply on the deflag page) you can rerequest the bot flag at Wikidata:Requests for permissions/Bot should you need it again. Of course, You may request retaining your bot flag here if you need the bot flag. Regards--GZWDer (talk) 12:06, 26 June 2017 (UTC)

Compound with Isomeric SMILES but without InChI

Hi Snipre, I check the output from the SPARQL query at Wikidata:WikiProject Chemistry/Tools for compounds without InChI. I had Bioclipse process the (isomeric) SMILES and generate InChIs. This is what I found:

Number of missing InChIs: 2098
InChI too long: 1736
With undefined stereo: 348
Bad SMILES: 2
Compound classes: 12

That is, if the 2098 compounds with an isomeric SMILES and no InChI, all of them don't have them for a reason. By far the most have an InChI which is too long to fit in Wikibase (1736). Another 348 have undefined stereochemistry (these could be added, but not sure). Two have an unparsable SMILES, while another 12 are for compound classes, and have a "*" in the SMILES, and InChI cannot be generated for those. The code can be found at https://gist.github.com/egonw/eb07cf69278c0be78d560c103ef29e5a --Egon Willighagen (talk) 16:10, 29 October 2017 (UTC)

Special:Diff/580498321

Hi Snipre. Your addition results in a format violation. What is your suggestion to get rid of this error? --Leyo 13:57, 18 December 2017 (UTC)

@Leyo: The statement is correct and the constraint is not good enough to handle that case. It is a general problem in WD: does the absence of statement means that a property haven't a value or only that nobody was looking for a value ? And how can we declare that no value exists for one property ? I can delete the statement but I think we should start a discussion because I have similar cases for other constraint violation reports. Snipre (talk) 13:51, 19 December 2017 (UTC)

I mostly agree with you. You may also just take the appropriate measure that the "error" will disappear from the constraint violation report. --Leyo 09:43, 20 December 2017 (UTC)

SPLASH proposal?

Hi Snipre, in your recent edit you removed the SPLASH property proposal, but I cannot find where it is now listed? Can you point me to the new enlisting, please? --Egon Willighagen (talk) 13:32, 15 March 2018 (UTC)

@Egon Willighagen: Sorry, I corrected. Just go to the section Chemistry. Regards Snipre (talk) 14:42, 15 March 2018 (UTC)

Ah, thanks! --Egon Willighagen (talk) 16:34, 15 March 2018 (UTC)

GHS pictograms

Hey, I've created first item for GHS pictogram: GHS01: exploding bomb (Q51080746). Now I'm ready for the rest of GHS pictograms and after a few days I think I'll be ready for H, EUH, AUH and P-phrases (in this case I thought of official name (P1448) for phrase and short name (P1813) for phrase code). But before this, I would like to know if you have any comments on this (especially, if you think it should be done in different way, using different properties etc.). Also, if everything is okay, I'll ask in Project Chat for as many descriptions in EU languages as it is possible (like in English: hazard pictogram in Globally Harmonized System of Classification and Labelling of Chemicals), because adding those in QuickStatements is the easiest way. Best, Wostr (talk) 19:21, 29 March 2018 (UTC)

@Wostr: No comment from my side, just ask if you need support. I am on holydays now, Will be back end of next week

Snipre (talk) 20:10, 31 March 2018 (UTC)

Representation of Wikidata at the Wikimedia movement strategy process

Hi Snipre, I'm contacting you because I would like your support and your comments on my proposal to represent the Wikidata community at the Wikimedia movement strategy process. I'm contacting you in private because you are a member of the Wikidata Community User Group and I thought that this could be relevant for you.--Micru (talk) 18:45, 18 June 2018 (UTC)

on Q74335 P31

Salut, ok, I won't revert your revert of my revert of your edit ;) but. I'm aware of efforts at WikiProject_Books, but my conclusion on A Fire Upon the Deep (1st edition) (Q74335) would be that it should be a work due to all the sitelinks and properties refering to the work, and indeed either deleting edition-specific claims or creating dedicated item linking to A Fire Upon the Deep (1st edition) (Q74335) with edition or translation of (P629). But it seems you turned A Fire Upon the Deep (1st edition) (Q74335) into the edition and created A Fire Upon the Deep (Q58221164) for the work: while this is problematic for books data reusers (such as inventaire.io (Q32193244)) that expect some stability from items P31-based FRBR level ersatz, I guess this could work from a strict Wikidata point of view if you do it fully, that is also moving work-related properties to A Fire Upon the Deep (Q58221164) (nominated for (P1411), Open Library ID (P648), ISFDB title ID (P1274), NooSFere book ID (P5571)), move the Wikipedia sitelinks, move labels and descriptions in other languages, and change or remove them on A Fire Upon the Deep (1st edition) (Q74335) to refer to the 1st English edition. That's to avoid to do that that I would have preferred the option of keeping A Fire Upon the Deep (1st edition) (Q74335) as the work and moving elsewhere or removing the edition properties. -- Maxlath (talk) 08:08, 5 November 2018 (UTC)

@Maxlath: First point: never determine the subject of a WD item based on the sitelinks because WP articles 1) are mixing data and subject and 2) you can always find some particular sitelinks which have a different point of view than the others. So unless defining WP:en as the only reference, sitelinks should not be trusted.

If external users are using WP articles to determine the subject of a WD item, this is their problem and they should be aware of the consequences due to the above explanation. WD has its own classification and users have to refer to that classification only. Currently the distinction between work and edition is provided by the WikiProject Books and the mentioned item was not following classification so defining it as a work or as an edition was a 50/50 choice.

I completely support the need of clear rules but rules exist and are defined in WD.

Second point I only moved properties according to the information provided by Wikiproject Books: but (nominated for (P1411), ISFDB title ID (P1274) are NooSFere book ID (P5571)) are not in the list of the work properties or in the list of edition properties. So I prefer to avoid to do mistakes and I let the responsibility of the people who added those information to do their job correctly. I missed the Open Library ID (P648) so we can change it.

Finally about your proposition of keeping A Fire Upon the Deep (1st edition) (Q74335) as the work item, do the change, but just keep in mind what you said about stability: before my modifications, A Fire Upon the Deep (1st edition) (Q74335) was undetermined (you thought it was the work item but based on unreliable hypothesis), now the split is done and both items are clearly identified even if some data should still be moved. By changing this situation, you just break the stability you wanted.

I tried to split the data according to rational facts: work items have often less data than edition so creating a new work item will require minimum of data moves, the original item was no clear majority to become the work item, again because sitelinks are unreliable to decide what is the subject of an item and only the data present in the item have to be considered. Snipre (talk) 10:50, 5 November 2018 (UTC)

@Snipre: books on Wikidata are a mess but until very lately, the recommanded instance of (P31) for works was book (Q571), so no, it's not a 50/50 choice, the right update would then have been to change book (Q571) values into written work (Q47461344) as people (external users, but also infoboxes) might have been built on this common understanding. -- Maxlath (talk) 16:00, 5 November 2018 (UTC)

@Maxlath: Two or three years ago, editions were not distinguished from work and were classified as book too. How can I suppose that the current classification is referring to the work before the recent change to written work or to the edition before the classification change for edition ? Again I can spend hours to understand what old contributors did or I can use the data of the item to identify the subject. So when I see that all usual data of one edition item (publisher, publication date, ISBN,...), what is the criteria to define that the item is for work ? And usually people complain about the fact of creating a work item as a condition before creating an edition item. So often the initial goal of item creation is edition and work item is missing.

We still can discuss about that topic during hours, my advice is the following: next time you see an edit like mine, changing book to edition, if you consider that the edit is wrong, please curate yourself the item to distinguish edition from work and don't revert to the old ambiguous situation. I can do the wrong choice, but if a choice has to done, do it and don't let the case in an undetermined state. Regards Snipre (talk) 15:12, 6 November 2018 (UTC)

Seconds, literally, seconds!

You caught me in fixing that Niels PubChem CID by seconds!!! :) --Egon Willighagen (talk) 19:21, 28 August 2019 (UTC)

Community Insights Survey

Share your experience in this survey

Hi Snipre,

The Wikimedia Foundation is asking for your feedback in a survey about your experience with Wikidata and Wikimedia. The purpose of this survey is to learn how well the Foundation is supporting your work on wiki and how we can change or improve things in the future. The opinions you share will directly affect the current and future work of the Wikimedia Foundation.

Please take 15 to 25 minutes to give your feedback through this survey. It is available in various languages.

This survey is hosted by a third-party and governed by this privacy statement (in English).

Find more information about this project. Email us if you have any questions, or if you don't want to receive future messages about taking this survey.

Sincerely,

RMaung (WMF) 16:25, 9 September 2019 (UTC)

Reminder: Community Insights Survey

Share your experience in this survey

Hi Snipre,

A couple of weeks ago, we invited you to take the Community Insights Survey. It is the Wikimedia Foundation’s annual survey of our global communities. We want to learn how well we support your work on wiki. We are 10% towards our goal for participation. If you have not already taken the survey, you can help us reach our goal! Your voice matters to us.

Please take 15 to 25 minutes to give your feedback through this survey. It is available in various languages.

This survey is hosted by a third-party and governed by this privacy statement (in English).

Find more information about this project. Email us if you have any questions, or if you don't want to receive future messages about taking this survey.

Sincerely,

RMaung (WMF) 19:51, 20 September 2019 (UTC)

Hugo de Vries

Hi Snipre,

thanks a lot for your explanation concerning work <-> edition. As I understand it, I always have to create two wikidata-items for a book. So I started with work: Q21962241, and edition: Q69576569 for De ademhaling der planten by Hugo de Vries. May I ask you, if you could find some time to carefully check the two items. Did I put everything in the right place? Didn't I forget important things ? etc. So that I know what to do for other books on Dutch Wikisource, that I want to add to Wikidata. Many greetings, --Dick Bos (talk) 19:11, 3 October 2019 (UTC)

@Dick Bos: This is OK. Thank you for taking account of our model. Regards Snipre (talk) 14:33, 9 October 2019 (UTC)

Q1225746

Regarding [3] - I do not think, that PubChem-ID (CID): 13361 is the best match for this compound, because it is denoted as a radical. What do you think about changing the CID to https://pubchem.ncbi.nlm.nih.gov/compound/6327114 ? --Mabschaaf (talk) 13:39, 16 October 2019 (UTC)

@Mabschaaf: Please check the formula of the compound and don't take care of the name/label of the entry in PubChem. The name of an entry in PubChem is a parameter which can be modified by registered users, so there is no guarantee that the displayed name is correct. PubChem manages the structural identifiers (formula, InChI, SMILES,...) so those are relevant to identify a compound. And if you take the time of analyzing the compound, you can see that there is no missing hydrogen or other atom. So this compound is not a radical.

I can only recommend you to use structural identifiers like InChI or InChIKey to compare different databases: don't trust name, formula or even commercial identifiers like CAS.

Finally, even if the chemical in WD is not dimethyl hydrogen phosphite, don't modify identifiers with similar InChIkey but create a new item. The rule in WD about chemicals to link different entries from different databases is to use the InChIKey. This is the truth or at least the best truth element we can have currently. Snipre (talk) 12:53, 17 October 2019 (UTC)

Stringent terminology

Thank you very much for your reminder about the terminology; it's precisely what I had in mind and I hope my discussion partners will agree to that. I believe myself that the ambiguous meanings of English words like rock, stone, form, type, kind and piece are among the root causes of this very discussion, likewise with Swedish words sten and berg (I sometimes find myself confused by the ambiguities of my own language).

I'm however a bit hesitant to continue the discussion on that chat page; as I suspect this is a recurring topic at Wikidata, I would rather see this thread moved over to Help talk:Basic membership properties, but as we are already three people involved in it (four with you), I'm afraid that any unilateral move from me now will merely split that thread in two and we'll start repeating ourselves. Do you have any advice for me in this regard?

Also, I have been working on a longer explanation of how the Stone and Rock articles came about and how the different languages treat the article subject, but I feel this explanation is too long even for the Help talk page. Therefore I'm considering posting it to the Talk:Q22731 page (which doesn't even exist yet), extract my conclusions plus a brief summary from it, and post that either in the chat thread or the aforementioned Help talk page (not both). What do you think? Am I over-complicating things for other readers and discussion partners?

But given my slow editing pace, they may well have concluded that thread before I'm ready with this (I'm working on a table to try to illustrate the instance/subclass relations between different existing and hypothetical WD items, in which I have identified four different items that may reasonably be labelled "stone", and three of them also as "rock"; no wonder we don't understand each other when we communicate using linear text only)...

I see that your own native language is French; do you see the same ambiguities in the English language as I do (those mentioned above plus event, office, letter, right, left, work, note, wing, table, point, period, sense etc), or is English perhaps a bit more homomorphic with French?

Final piece of language trivia: In Swedish, there are separate words for two different kinds of liquid used in cooking: One is sky, from French jus since 1755 and meaning the same thing; the other is juice (sometimes spelled jos), from English since 1950 and referring to what you may get from an orange or lemon, but with exactly the same French origin! The word sky also is an older noun meaning cloud or sky, as well as a verb meaning escape or avoid (from Low German). We have quite a bit of work ahead of us getting all these senses into the lexeme database and linking them to the correct items! --SM5POR (talk) 09:15, 31 July 2020 (UTC)

@SM5POR: Every contributor has to pass that situation. My proposition is not to use Help talk:Basic membership properties but to use a WikiProject to discuss the definitions like in your case Wikidata:WikiProject Geology. In our discussion about chemicals we didn't reach the same conclusions as in Help talk:Basic membership properties for the definition of instance of. The important thing is not to have one general ontology for all specific domains, but to have something coherent. And each domain should be able to define itself the level of granularity to its ontology. When working on ontology, the difficulty is the modeling of reality and often it is not possible to include all details, all differences in a model.

I can't help for rock/stone,... because my domain is chemistry and chemicals. My advice is start to list the concepts,(stone, rock,...), then write a definition for each concept. Good approach is to use definition for well known references. For chemistry we have the International Union of Pure and Applied Chemistry (Q33438) which offers some good and well accepted definition for important concepts. Then you have to check if the definitions are good enough to help you in our ontology building, perhaps you have to modify a little the definition for your purpose and this is the first problem as other contributors can use other references or not accepting the change of definitions. Finally you have to take several examples and test them in your ontology. Here it is important to take general cases and some particular cases to check if the ontology is robust. The main problem we faced is to find rare examples which can't be defined in a ontology we agreed in the past and to start again the revision of the ontology to integrate new definitions and new concepts.

Hoping you find constructive contributors, you can then write the ontology in the guidelines of the WikiProject (see what we did for chemicals in Wikidata:WikiProject_Chemistry/Guidelines) and starting there you can start to modify items. It is good to discuss with the main groups of same interest in the different WPs, I try to be present in WP:en and WP:fr to show what is done in WP in order to have several persons having the same interest as you working in WD. Defining an ontology with 3-4 persons is important to reach a good consensus and to be sure to cover enough cases before you start the classification of item in WD.

French is comparable to English when defining concepts: it is not easy to have one word=one concept and often you need a long sentence to define correctly the concept. See chemical substance (Q79529) as example. Snipre (talk) 14:12, 31 July 2020 (UTC)

Thanks for your advice! I have looked into a few of the Wikiprojects, but find it difficult to locate much activity there (I have received best response in Ontology), which is why I began reading and commenting on the Wikidata chat. Acting on your suggestion now, I looked at chemical substance (Q79529) and added a Swedish description by translating the English and Spanish ones (they are very similar), plus added "structure" to catch also the isomer and polypeptide variability, as well as "melting point" for another physical property (geology isn't my cup of tea either; I prefer organic chemistry though it's been 40 years since I studied it at school, and computer science).

I agree that the ontology has to be discussed within each domain, but right now I get the impression that a lot of random edits are made by editors with effectively no understanding of class logic. A Swedish wikipedian who is working with Infobox layout told me that many WP editors dislike terms like "instance of" or "subclass of" appearing in infoboxes, so they want to use part of (P361) instead! That's total ontological disconnect, if you ask me, and the very reason I got into Lua programming, to help develop a less "brutalistic" mapping between Wikidata and Wikipedia... You shouldn't have to understand class logic to be a WP editor, but the software interface should take care of the structural checks, I think. --SM5POR (talk) 15:24, 31 July 2020 (UTC)

By the way, I just read what you wrote above regarding whether to engage yourself in the interwiki links or not when working with Wikidata. Given the amount of effort you apparently put into Wikidata, you are well excused for not also sorting out interwiki links. I think you have made the correct decision, even as I have taken another approach; I tend to refer to multiple WP pages when trying to interpret the WD item statements. However, I fully agree with you that the identity and definition of the WD item should stay the same regardless of which WP pages currently link to it or what they contain; I look at the edit histories to determine what has happened to the item in the past.

One of my first edits on WD (a removal of a subclass of (P279) link to resolve an annoying subclass loop) was reverted (and the opposite link removed instead) because I didn't pay attention to the history of the item. I had merely looked at the English pages, but upon later re-inspection I realized that the English articles (and some of the other languages) were linked in direct opposition to the Russian article (with the remaining languages), and the Russian article had in fact stayed true to the original item while tracing the English article(s) via WD edit histories switching items and being merged back again felt a bit like tracing the footsteps of a drunk Londoner dropped off in the middle of the night near the Dutch-French border on Saint Martin... I have described my findings in detail on my User:SM5POR#Nematode infections page but not made any further edits; some day I may make an effort of actually resolving that mess and putting the incorrectly linked WP articles back in proper company, but not today. --SM5POR (talk) 11:37, 31 July 2020 (UTC)

I can't change the articles in WP to adjust them to WD items. Then we have to thing that article in WP are evolving so perhaps articles are not matching the WD concept now but will be more precise in the future once sufficient information are added in the article. My interest is to compensate the weakness of Wikipedia in term of data management, not to solve languages conflict. So that why I don't take care of the interwikis. Then most problem came from bad data addition from WPs and the people doing the importation are not more contributing in WD, so it is a time loss to try to correct bad linking. Better to offer a coherent system in WD and later people will correct themselves the interwikis if we have a good system. Snipre (talk) 14:12, 31 July 2020 (UTC)

We sent you an e-mail

Hello Snipre,

Really sorry for the inconvenience. This is a gentle note to request that you check your email. We sent you a message titled "The Community Insights survey is coming!". If you have questions, email [email protected].

You can see my explanation here.

MediaWiki message delivery (talk) 18:46, 25 September 2020 (UTC)

Data removal

Why did you remove property IDs from chemical elements? For example, this for Pt looks correct & well sourced. -DePiep (talk) 17:07, 15 May 2022 (UTC)

@DePiep: Because there is one item for the element platinum (Q880) and one another item for the chemical compound platinum (Q27882222). Most IDs are relevant for the chemical compound and not for the element and separating element from chemical compound is necessary in case of several allotropes or non metallic compounds. This is not a general trend for now but as there are several cases using this format (see Wikidata:WikiProject_Chemistry/Tools#Chemical_element_and_corresponding_simple_substance, then the splitting of data is necessary when the difference is made. Snipre (talk) 19:58, 15 May 2022 (UTC)

I see. Thanks. DePiep (talk) 20:03, 15 May 2022 (UTC)