User talk:Zolo/archive/2

From Wikidata
Jump to navigation Jump to search

Swedish counties

[edit]

I guess I have reached as far as I can with the limited set of tools we have at the moment. See sv:Mall:Länsfakta. -- Lavallen (block) 18:27, 15 May 2013 (UTC)[reply]

Properties used are: p94, p237, p242, p395, p507, p36 and p35. - Lavallen (block) 18:34, 15 May 2013 (UTC)[reply]
That looks good, however, I see that in items like Västerbotten County (Q104877), head of state (P35) links to a "list" item (at least the English sitelink). I tend to think that pages like en:List of governors of Västerbotten County or en:List of counties in California should not be called list, because there is no corresponding non-list page and it makes sense to add non-list content there. But as long as the English article is called "list of", I am afriad that such items will be considered "lists items", and so not very suitable for use in statements. --Zolo (talk) 07:28, 16 May 2013 (UTC)[reply]
There is not much to tell about the governor of a special county since there is more or less the same work to be a governor of any County. Such an article would tend to only repeat the historical part of the County-article. The only information-bit worth mention is that when Stockholm City and Stockholm County merged 1968, it was the governor of the City who replaced the County-governor. A governor is not elected, they are selected by the central goverment. It's a typical job for a retired politician.
It looks worse in Malmöhus County (Q533915) who was transformed into Skåne län 1998. The English article about the County is there a list-article, while the Swedish has a separate article for the list. -- Lavallen (block) 08:58, 16 May 2013 (UTC)[reply]
BTW, the governor of Stockholms City: "Over-Governors office (Q4994097)" had a different title, and the title is redirected and the list is also included in the article there. -- Lavallen (block) 09:38, 16 May 2013 (UTC)[reply]

Edition data

[edit]

I have expanded the RFC about references/sources to decide how to store edition data.--Micru (talk) 13:28, 16 May 2013 (UTC)[reply]

part concerned

[edit]

Suggested new label here: Property_talk:P518. Danrok (talk) 23:40, 21 May 2013 (UTC)[reply]

I see that in the statement "creator: Lysippides Painter" you've put the qualifier "applies to part: Black-figure pottery", and similarly for the other part. However, black-figure pottery (Q629602) is about a style of painting, it's not about a specific part of the amphora. And while us humans can deduce what you really meant, strictly speaking the semantics are incorrect. Perhaps it would be best to create an item for each of the painted sections on that specific amphora? Silver hr (talk) 15:55, 30 May 2013 (UTC)[reply]

Yes, I was wondering about that. It is probably more rigorous to create two items. But I am wondering if it is practical. For humans at least, it makes it rather hard to read. A common practice for ancient vases is to call one side "side A" and the other "side B". So maybe: "technique: black figure pottery - applies to part: side A" and "creator: Lysippides - applies to part side A" ? --Zolo (talk) 16:24, 30 May 2013 (UTC)[reply]
(Late reply since my time for Wikidata has been very limited lately.) I'm not sure why a solution like that would be hard to read. Suppose you had the items "black-figure side" and "red-figure side" which would be linked to the amphora item via part of/has part, and would have the statements "technique: black-figure pottery" and "technique: red-figure pottery", respectively. Surely that wouldn't be hard to read? Plus, the items for the sides could have additional statements, such as "depicts", which would be more precise than having them on the amphora item. Also, I realize that you may have the desire to keep all the information in one item so that it's all visible at once, but please remember that this is a user interface issue, and we shouldn't try to solve it by having a less precise semantic structure--we should eventually have a better UI that will display data from connected items as well as the base item. There is already something along those lines - the Reasonator. Silver hr (talk) 15:24, 12 July 2013 (UTC)[reply]
This kind of "backwards-inferrence", I mean inferring properties about the parts for properties about the whole, seems substantially more tricky than things like "this is an amphora, so it is also a vase". Determining whether something that applies to a part also applies to the whole seems to require some "common-sense". Can the guy who made one side of the amphora be considered a creator of the amphora ? I would say yes. Can the guy who designed a tiny room in the Palace of Versailles be considered the architect of the Palace of Versailles ? I would say: for most purposes no, but it may not be easy to make the distinction clear enough to be determined by a machine. Zolo (talk) 11:24, 13 July 2013 (UTC)[reply]

Re: Précédé par / suivi par

[edit]

Bonjour, j'ai vu que dans Gagik I of Vaspurakan (Q882866), vous aviez mis : "Gagik Ier : prince de Vaspourakan, suivi par roi de Vaspourakan", ce qui pour un robot voudrait dire que le prince du Vaspourakan est devenu roi du Vaspourakan à la place de Gagik. Je pense que pour être plus exact sur la structure, il faudrait mettre "suivi par roi du Vaspourakan" dans prince of Vaspurakan (Q13408970) --Zolo (talk) 13:05, 31 May 2013 (UTC)[reply]

Bonjour,
J'ai ajouté différents champs (follows (P155), followed by (P156), start time (P580) et end time (P582)) sur prince of Vaspurakan (Q13408970) et sur king of Vaspurakan (Q13409232), mais dois-je aussi modifier Gagik I of Vaspurakan (Q882866) ? et comment ?
Merci d'avance de ta réponse. Odejea (talk) 13:19, 31 May 2013 (UTC)[reply]

J'en profite aussi pour dire que si tu peux ajouter des sources, c'est mieux, surout sur des sujets assez pointus comme les dates de dates de règnes de rois arméniens médiévaux. Il y a encore des discussion pour savoir comment le faire exactement, mais une solution simple est de créer un élément sur un livre qui sert de source, et d'ajouter dans le champ source stated in (P248) <nom du livre>, page(s) (P304), <page>. On pourra toujours reformater par la suite. Zolo (talk) 15:29, 31 May 2013 (UTC)[reply]

Voila, j'ai commencé avec les modèles Cyrille Toumanoff, "Les dynasties de la Caucasie chrétienne de l'Antiquité jusqu'au XIXe siècle", 1990 (Q13409386) et Continuité des élites à Byzance durant les siècles obscurs (Q13409396) que j'ai déjà inséré dans Gagik II Artsruni (Q13407257). Est-ce correct ? Odejea (talk) 15:57, 31 May 2013 (UTC)[reply]

et les dates ?

[edit]

J'ai un souci pour certaines dates qui ne sont pas toujours connues avec précision. Par exemple comment renseigner pour les cas suivants :

  • entre 987 et 991
  • 886 ou 889
  • avant 854
  • après 965
Merci d'avance de ta réponse. --Odejea (talk) 16:13, 31 May 2013 (UTC)[reply]

Re: Bot adding of labels

[edit]
Hello, Zolo. 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 11:02, 1 June 2013 (UTC)[reply]

Hello, Zolo. You have new messages at Wikidata:Properties_for_deletion#Property:P387.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

这个属性已经被提请删除。请参与讨论

This property is nominated deletion. see Wikidata:Properties_for_deletion#Property:P387.

--GZWDer (talk) 02:10, 7 June 2013 (UTC)[reply]

Resolved

[edit]
Hello, Zolo. You have new messages at MediaWiki talk:Gadget-AuthorityControl.js#Bug with 'http://www.marinetraffic.com/ais/fr/shipdetails.aspx?MMSI=-$1'.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

--Ricordisamoa 10:43, 7 June 2013 (UTC)[reply]

Property creation request

[edit]

Hi Zolo, could you please create edition of? Consensus has been reached and this property is necessary for connecting work items and edition items. Thanks!--Micru (talk) 20:08, 14 June 2013 (UTC)[reply]

✓ Done, as P:P629. --Zolo (talk) 07:51, 15 June 2013 (UTC)[reply]
Thanks a lot! Could you also create the translator property? The discussion has been open for a long time and now it seems that there is consensus for creation.--Micru (talk) 20:43, 21 June 2013 (UTC)[reply]
✓ Done, you can also request property creator rights ;). --Zolo (talk) 06:25, 22 June 2013 (UTC)[reply]
I did, but apparently I am "not familiar enough with the process" :D But the way, Chapter is ripe to be created ;) --Micru (talk) 05:12, 24 June 2013 (UTC)[reply]

Re: List or Mayors of New York

[edit]

Do you think it will in the future be possible to use a 'Query' to find all "Mayors of New York" and also the latest one? The possible problems I can identify is if the title change by time. Like in Sweden the chef d'état is normally the king or queen, but it could also be a regent (Q477406) in times of interregnum (Q560509). And since this title is still in use (on rare occasions), there can be some confusions. Finding Ingemund Bengtsson (Q707581) in the line of kings and queens of Sweden would suprise many. (In the 1980s all the members of the royal house where on holiday, and the chairman of the parlament was "acting chef d'état" for a few days.) -- Lavallen (block) 06:13, 15 June 2013 (UTC)[reply]

I think that it will be possible to use queries, but perhaps too complex to make it acceptable coding-wise and performance-wise. One possible solution would be to use dedicated items (possibly sv:List of Swedish monarchs). I suppose we need a RFC on that, given that the only relevant high-traffic page is the project chat and it gets archived too often to have a detailed discussion. --Zolo (talk) 07:53, 15 June 2013 (UTC)[reply]
Good question and good answer as well. I'm thinking a lot about data structures and how we could query things later. Just one point: A dedicated item "List of Swedish monarchs" would be wrong in my opinion. Because an item named "List of Swedisch Monarchs" is about a "List of Swedisch Monarchs". Enter all Swedish monarchs under this item would be wrong, because the item it isn't about the Swedish monarchs itself, its about the list. I hope you understand what I mean. So if we should choice the way to use dedicated items, I strong suggest to use a label like "Swedisch monarch" and not "List of Swedish Monarchs" --Nightwish62 (talk) 20:52, 28 June 2013 (UTC)[reply]

Hi! I notice you've added an example for the new property PubMed Health (P653) in myocardial infarction (Q12152), but the linked external page is not found. I have no idea how that website works, or who proposed this originally. So, just letting you know that someone may need to fix something somewhere! Danrok (talk) 15:08, 22 June 2013 (UTC)[reply]

Oops, I got the URL wrong, found it through Google search for pubmedhealth + 0001246. Thanks for notifying. --Zolo (talk) 07:17, 23 June 2013 (UTC)[reply]

Street property

[edit]

Hi Zolo

I notice we've similar opinions about how to handle address information. Please take a look to my comments in Wikidata:Property_proposal/all#Address_.2F_Adresse_.2F_Adresse. It's very frustrating there a little response, as my proposal as well as you proposal. I really like to start enter address information, but it seems the community wouldn't come to an agreement in the next time. Any ideas how we can push this forward? --Nightwish62 (talk) 20:58, 28 June 2013 (UTC)[reply]

Hi Nightwish, sorry for late reply. I am added some comments to the proposal, I hope this will help with clarifying things. --Zolo (talk) 21:46, 1 July 2013 (UTC)[reply]
Meanwhile there are two more pros. So at all there are four pros (with your proposal). There is only one oppose and it seems the person didn't understand something right. The proposal is more than two weeks old. Do you think we can create this property? I think so. --Nightwish62 (talk) 17:53, 3 July 2013 (UTC)[reply]
I think we can create street, but I think "number" should be created at the same time, because it would not be very efficient to fill out the street now and the street number later. An I am unsure about it: should it be street specific or a more general string-type number that could be used for other things like catalog numbers ? I suppose I can create a "street number" property, as it is easier to broaden it later. --Zolo (talk) 18:04, 3 July 2013 (UTC)[reply]
I totally agree. Only street name makes no sense and beginning with street number and maybe broaden (rename) it later would be fine. Thank you for creating those both properties! --Nightwish62 (talk) 18:34, 3 July 2013 (UTC)[reply]
Done, I have also left a message to user:Rschen7754, as some adjustments may be necessary, I think we should not rush using it in the first few days. --Zolo (talk) 19:32, 3 July 2013 (UTC)[reply]
Thanks. Yes, but I'll create some sample entries. One question left, what do you think: Use street number as qualifier or as standalone statement? I think we should use it as qualifier, as in Empire State Building (Q9188). --Nightwish62 (talk) 19:46, 3 July 2013 (UTC)[reply]
Yes, some sample entries are necessary. I agree that it should be a qualifier of street, as it will always be possible to encounter buildings with several street statements. Actually, I had already documented it this way in French, but forgot to fill out the English version. --Zolo (talk) 19:53, 3 July 2013 (UTC)[reply]

RfC on Wikidata's primary sorting property

[edit]

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? 18:22, 30 June 2013 (UTC)[reply]

Succesion of administrative units

[edit]

Haven't been here for a while. Are you aware any good system to describe the evolution in a case like this?

There is many even more complicated cases, but lets start with this! -- Lavallen (block) 16:57, 4 July 2013 (UTC)[reply]

I see your changes! The original founding of these rural communes is that parishes split in a civil and a clergal part. Ytterlännäs parish (Q10723853) split in Ytterlännäs parish (Q12560122) and Ytterlännäs rural municipality (Q10723852) as an example. Until today these parishes, like Ytterlännäs parish (Q12560122), are still the smalest part in the census-division. They will in this matter be replaced by census-districts as of 2016. -- Lavallen (block) 18:17, 4 July 2013 (UTC)[reply]
It seems to me that we should use follows (P155) / followed by (P156) and inception (P571) /dissolved, abolished or demolished date (P576) as main statements for all items (I would not mind replace P571/576 with start time (P580) / end time (P582) but that would amount to the same thing. Should we also use start time (P580) / end time (P582) as qualifiers for P131 and P132, for instance Ytterlännäs rural municipality (Q10723852). I would say yes, even in cases where it is arguably redundant, like Ytterlännäs rural municipality (Q10723852) -> P132 (P132): Q10553927 beginn 1863 end 1951. I would say yes, at least for P132, as it makes things more explicit, and keeps the format consistent with items where the adminsitrative status changes. With located in the administrative territorial entity (P131), it feels a bit awkward, but I think it makes sense, for the same reason (things would sound more natural if we used "part of" instead of P131).
I would say that Q10700445 was replaced by Ytterlännäs rural municipality (Q10723852), but maybe not that Ytterlännäs rural municipality (Q10723852) was preceded by Q10700445, which I find a bit confusing. I think once we have the geographic shape datatype, it will help make things clear that Ytterlännäs was enlarged in 1952.
For the rest, I think things work fine, dont they ?
By the way, I am not sure about the municipality names. Are they called Ytterlännäs and Torsåkers or is the final s just a genitive marker ? --Zolo (talk) 19:17, 4 July 2013 (UTC)[reply]
In Swedish: The ending 's' in 'Torsåkers' is a genitive marker, write 'Torsåker' when the word stands alone, but in the case 'Ytterlännäs', the genitiv and the nominativ looks the same. Words with ending 's' in nominativ looks the same in genitiv, and with ending aouåeiyäö, older grammatical rules can be used, with no ending 's' in genitiv. Older grammer is normally used for municipalities. A few cases even have irregular genitiv, [very unusual today] like Falu kommun and Kalmar kommun. In the later case, I would not recomend you to use the Swedish irregular genitiv in any other language than Swedish. [I have doubt using it in Swedish, since I consider irregular genitiv as dialect (Q33384).]
The full name is "Ytterlännäs landskommun" (at least in Swedish). "Ytterlännäs" can be used as "short name", when there is no doubt that it is the rural comune that is discussed. It is more obvious when you compare Q10714737 with Q10714739, here you have a rural and an urban comune with the same name. They coexisted and in some cases (but not in this case) the urban could be totally surrounded by the rural. The mayors office of both comunes where most likly located in the same locality, maybe in the same building. It can be different for urban comunes, but since we had three kinds of urban comunes, with different laws applied to each kind (at least before 1952), we normally use "stad/köping/municipalsamhälle" as a part of the name in Swedish. As you know, we also have urban areas, a statistical unit, and they can be both larger and smaller than the municipality with the same name.
But on Svwp, there is normally not a separate article for a "municipalsamhälle", therefor it may look like we do not use municipalsamhälle as a part of the name.
I have not modified the Swedish label of Q10700445 yet, it includes a separator, the Province.
As you maybe have understood a "municipalsamhälle" was a sub-municipality with city-laws inside a rural municipality. But not all city-laws where applied to every "municipalsamhälle". They where added when the king/goverment decided it. As time went by, the city-laws was applied to all municipalities, and 1971 all municipalities changed to one single kind. Today some municipalities claims to be urban municipalities, but it is both de jure and de facto no difference today. They can have different titles for the mayors, and a different trademark, but that is all. So if you see somebody change the label from "Stockholms kommun/municipality" to "Stockholms stad/city", it is correct, but it is only the trademark. Juridically, the name is still "Stockholms kommun".
By the way, all munipalities was de jure founded 1863, but because of local trouble some municipalities, where de facto not created until several years later. How are we supposed to describe that?
In some cases it is necessary to use dates as qualifiers for P131/part of, since some municipalities has changed county in history, and will soon in the future. To look more like the rest of EU, several of our counties will be merged in the future. (they are to small) The difficult things is how they will be merged. Where I live, some municipalities wants to merge with the north, some wants to merge to the west and some thinks they are the center of the world, and wants to be unchanged. -- Lavallen (block) 08:07, 5 July 2013 (UTC)[reply]
I made some tests also with Härnösand Town (Q10531751) now. P132 (P132) changes here by 1862/63. A "city" founded 1585 becomes an "urban municipality", and remains so until 1971. -- Lavallen (block) 12:56, 5 July 2013 (UTC)[reply]
Ok, I'll have the look at it. Perhaps not right now, as I won't have much time in the coming three days, so probably at the beginning of next week. --Zolo (talk) 13:40, 5 July 2013 (UTC)[reply]
I have some problems also with Härnö Parish (Q10531730). It was dissolved as administrative unit 1863, but it is still in use as a geographic unit. That looks hard to describe!
The exact date of when it was founded is hard to find. I have not even found which century, but I have found from which other parish it was split. It is also a little hard to describe what is had been a part of/in administrative unit. It has a long history, even longer that Härnösand City. The area here was to some extent terra incognita for the people in Stockholm. -- Lavallen (block) 16:36, 5 July 2013 (UTC)[reply]
Ok thanks for the explanation, that looks good to me. I think the fact that cities can exist for some time without a clear status is a reason why we should merge P131 with P32. --Zolo (talk) 12:44, 8 July 2013 (UTC)[reply]
P31, I guess you mean. -- Lavallen (block) 16:34, 8 July 2013 (UTC)[reply]
Yes, P31, sorry. --Zolo (talk) 19:11, 8 July 2013 (UTC)[reply]

street / address

[edit]

Hi Zolo. I'm still not quite happy about the street property (even I was the one who pushed the creation). At moment, there are no relation between the property 'street' and our intention to collect the address information. One example: If someone like, he could add the "33rd street" and the "34th street" for the Empire State Building (Q9188), as the building is also located on them. One idea which came up to me, is to use a 'container' similar to the "key event" which is also proposed.

(property): key topic
 (value): addresse
  qualifier: street
   value: (street name)
  qualifier: street number
   value: (street number)

This way it would indicate that the values are representing the address (and not just a street which the building is adjacent too). Another benefit would also be, that street and street number are horizontal aligned. What do you thing about this? --Nightwish62 (talk) 15:48, 6 July 2013 (UTC)[reply]

I am not sure to understand what you see as the value for the main statement if the name of the street goes to a qualifier. I agree that we have something broader that just the street but I think it could be left in the main property. ie, if something's official address is on Central Park rather than on the 5th Avenue, then the main value should be "Central Park". But I cannot think of a good label for that. --Zolo (talk) 17:50, 7 July 2013 (UTC)[reply]
The value of the main statement would be 'address' in every case (this was not a variable in the example). As mentioned, I just adapt the method which is discussed for "key event" (eg. key event <publication> start date 2010-05-04), even if the reason is not exactly the same. The function of "key topic" would just act as container, because more than one value which belongs together are expected. However, I'm even not sure if I like my idea by myself. It's just an idea at moment. But I think we must ensure just the street of the address is entered, no other street else. Perhaps labeling "address street" and "address street number" or "street used for address" is the better idea? --Nightwish62 (talk) 21:09, 7 July 2013 (UTC)[reply]
Ok sorry, yes, I had misread your suggestion. Yes, I think that a change in the label would be better, something like "address street" may do, though there are really case where the address is something else as a street (like possibly a park). I do not think we need to change "street number" if it remains a qualifier, as the fact that it is the address' street does not change the meaning of the qualifier. --Zolo (talk) 12:50, 8 July 2013 (UTC)[reply]

It's me again :) Could you please change the Template:Q, so the parameter can also have a leading 'Q'? The appending of the Q happens automatically, but this behavior is either annoying then helpful, because I often write '{{Q|' and then copy paste the item ID from another window. I know it's possible that the template checks if the 'Q' already exist and if not it prepend it, but I'm not familiar with programming template. Could you implement this please? --Nightwish62 (talk) 15:59, 6 July 2013 (UTC)[reply]

I'll try. However, it seems that this templates already slows the loading of long talk pages, I hope new features won't significanlty make things worse. --Zolo (talk) 18:08, 7 July 2013 (UTC)[reply]
Thank you. I don't think this feature would slow down the loading much more, as this would just be a easy string operation. --Nightwish62 (talk) 21:11, 7 July 2013 (UTC)[reply]
done. Yes, with Lua, string operation are relatively efficient, so I think that should be ok. I wished I could have share the code between {{P}} and {{Q}} but for some reason calling a subtemplate breaks the link--Zolo (talk) 06:35, 9 July 2013 (UTC)[reply]
Thank you so far. So even you can't make a link between the two templates: Could you please also add this feature to the P-Template as well? --Nightwish62 (talk) 18:57, 9 July 2013 (UTC)[reply]
✓ Done--Zolo (talk) 05:27, 11 July 2013 (UTC)[reply]
Thank you very much --Nightwish62 (talk) 18:30, 15 July 2013 (UTC)[reply]

Hi Zolo, I conducted test edits like you asked for on Wikidata:Requests_for_permissions/Bot/VIAFbot_4. Please let me know if you'd like me to do more editing or have comments. And you can reply here if you @Maximilianklein:. Thanks Maximilianklein (talk) 20:30, 17 July 2013 (UTC)[reply]

The property cause of destruction (P770) is available now. I saw that you participated in the discussion. --Tobias1984 (talk) 09:39, 6 August 2013 (UTC)[reply]

Thanks for notifying, though I am still not convinced that this is the best way to handle that. --Zolo (talk) 09:42, 6 August 2013 (UTC)[reply]

Peinture à l'huile

[edit]

Bonjour Zolo. Concernant la bonne propriété "peinture à l'huile" à utliser pour La Joconde, j'ai répondu : User_talk:Shonagon#Peinture_.C3.A0_l.27huile. En bref cela n'est pas si évident mais cette ambiguité de propriétés pose surtout un problème de cohérence globale qu'il conviendrait amha de résoudre. Pourrais-tu jetter un œil ? Merci --Shonagon (talk) 10:44, 12 August 2013 (UTC)[reply]

Merci beaucoup Zolo pour toutes tes réponses et remarques. Vais poursuivre mes contributions en tenant compte de toutes ces informations, et puis ruminer tout ça aussi ; au besoin nous aurons peut-être l'occasion d'en rediscuter sur Wikidata:Artworks_task_force :). Bien à toi. --Shonagon (talk) 19:09, 14 August 2013 (UTC)[reply]

The property chapter (P792) that you supported is available now. Please help out by reviewing other people's proposals. --Micru (talk) 17:47, 12 August 2013 (UTC)[reply]

Thanks, we should really try to do something about property proposals... --Zolo (talk) 18:18, 12 August 2013 (UTC)[reply]

The property significant event (P793) is available now. I saw that you participated in the discussion. --Tobias1984 (talk) 17:05, 13 August 2013 (UTC)[reply]

Ok thanks. --Zolo (talk) 18:07, 13 August 2013 (UTC)[reply]

On me dit de contacter un admin

[edit]

Bonsoir,

J'essaye d'ajouter P107 (P107) work (Q386724) à une entrée, et ma modif est filtrée (j'ai le message "Cette action a été automatiquement identifiée comme nuisible, et a donc été empêchée. Si vous êtes convaincu(e) que votre action était constructive, veuillez contacter un administrateur et l'informer de ce que vous avez essayé de faire. Voici une brève description de la règle de filtrage antiabus ayant détecté votre action : Adding main type property.")
Donc ou bien j'ai raté une discussion, ou bien il y a un problème avec le filtre. --Coyau (talk) 22:13, 17 August 2013 (UTC)[reply]

Ah oui, j'ai trouvé. Bonne soirée. --Coyau (talk) 22:20, 17 August 2013 (UTC)[reply]
Oui, ça fait longtemps que cette propriété était critiquée. J'étais personnellement pour sa suppression, mais le passage d'"obligatoire" à "nuisible" a été un peu brusque... Enfin, le filtre a été retiré. --Zolo (talk) 09:01, 18 August 2013 (UTC)[reply]

Update: Move statements

[edit]
Hello, Zolo. You have new messages at Wikidata:Bot requests#Move statements.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

--Ricordisamoa 02:49, 27 August 2013 (UTC)[reply]

Update: Script for property deletion

[edit]
Hello, Zolo. 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 00:25, 30 August 2013 (UTC)[reply]

columbarium

[edit]

Salut,

Je souhaite indiquer sur des biographies le numéro de case du columbarium du Père-Lachaise. J'ai fait un test sur Raymond Bossus (Q3420692) en utilisant inventory number (P217). Cette propriété ne convient pas vraiment, mais je n'ai pas d'idée de propriété à créer. As-tu une idée ? Pyb (talk) 08:03, 4 September 2013 (UTC)[reply]

Salut, à moins de créer une propriété spéciale, ce qui parait un peu encombrant, je ne vois que inventory number (P217) ou catalog code (P528). Ma préférence irait pour généraliser P528 en "numéro d'identification". J'ai suggéré ça sur Property talk:P528#Extension of the property. --Zolo (talk) 09:30, 4 September 2013 (UTC) + Zolo (talk) 13:48, 4 September 2013 (UTC)[reply]
A part ça, il ne faudrait pas créer un élément "sénateur français" ou quelque chose comme ça ? "fonction: Sénat", c'est un peu bizarre. --Zolo (talk) 09:31, 4 September 2013 (UTC)[reply]
Merci, j'ai confondu avec Député qui existait. J'ai créé Senator of the French Fifth Republic (Q14828018). Pyb (talk) 17:19, 5 September 2013 (UTC)[reply]

P6 and some trouble I have detected when I made some tests...

[edit]

How far have we come with the use of head of government (P6)? I made some tests in Kramfors Municipality (Q514815), the municipality I live in.

I have tried to give some answers, but these are difficult questions :]. --Zolo (talk) 13:27, 6 September 2013 (UTC)[reply]
  • Are there any property to identify the municipal board?
  • There are 41 seats in the counsil. But the counsil has 65 members + 1 vacant. (Only 41 is supposed to vote at each time.) Therefor is it not possible to identify the "number of seats/political party" only by listing the members and their political party.
  • Both the counsil and the board has commitees. The chairmens in those commitees are considered important. How to describe the relations between these object. Observe that a commite can be created and dissolved any time, except some non-optional commitees.
  • That seems rather difficult unless we are willing to create items about all and any comittee. If this is not needed for Wikipedia templates, I would rather wait until we are more clear about more basic topics.
  • Agree, it's not top priority. On svwp only the chairman of the municipal board has automatic notability.
  • The political parties in the counsil is organised in "groups of political parties", and each group has a 'groupleader'" (Except for the political party who is represented by an empty/vacant seat.)
  • The "chief executive" is choosen by the counsil or the board, but it is not considered a political position.
  • I would say we can add "instance of: chief executive" as a qualifier, at least until we find a better solution.
  • Are we supposed to create an item for each period? ("Municipal counsil of Kramfors 2010-2014") (Member of municipal counsil of Kramfors 2010-2014.) Such items and articles exist for the riksdag. Re-elections by political reasons are extremly rare in Sweden. And I have never heard of it on municipal level. Re-election becase of trouble with the vote-counting has occured.
  • Perhaps later, but I do not think we have yet reached the critical level of sophistication to embark in this kind of thing. And if reelections, that may not really solve the problem.
  • How to describe "empty/vacant seats"? They occur since we vote on lists of names. Empty and handwritten lists are allowed as long as the full name of the political party is written in the label. If no list in the vote contains anybody who live in the municipality, the seat becomes empty.
  • If we create an item for each period, that could possibly be done through listing the seats in contains the administrative territorial entity (P150). But that would still be debatable. Otherwise, the only other solution I see would be one item per seat, but at the moment, that seems even more like an overkill...
  • A person can represent a political party in the counsil or board, without being member of that political party. It's rare, but it can happen in positions who require special education. Using member of political party (P102) only in the person-related items, can therefor be wrong. And using P102 without a source in the person-item can be troublesome. Using P102 as qualifier in relation with "member of municipal counsil" is less of a trouble, and is easier to give a source for.
Yes, always adding a qualifier to "member of" seems like a good idea. But if the person is also member of the party, that makes sense to also indicate that outside the "member of", and that may be confusing a bit confusing. Perhaps the qualifier should not be P102, but rather something like "represents: ". I think that would be techncally more accurate, and that may be useful in for other cases as well. For example, someone who represents a company in a board of directors.
I guess technically every members of the company board are supposed to represent all of the annual assembly. But in practice, yes, people in the board often represents an (important) fraction of the owners. -- Lavallentalk(block) 11:41, 8 September 2013 (UTC)[reply]
  • How are we supposed to describe the governing majority (or minority in some cases)? The members of the board represents not only the governing majority, but also the important parts of the minority. And there is often a commisioner from the opposition (oppositionsråd) in a municipality.
  • Isn't that sufficient to have the political affiliation of people in items about them ?

-- Lavallentalk(block) 11:47, 6 September 2013 (UTC)[reply]

URL Code ROME

[edit]

Il semble que l'url vers Pole Emploi que tu as mis pour le Code ROME expire. Ce schéma ci ne devrait pas expirer: http://recrutement.pole-emploi.fr/fichesrome/ficherome?codeRome=XXXXX&domaine=Candidat --Teolemon (talk) 12:29, 12 September 2013 (UTC)[reply]

Lua and Qualifiers

[edit]

I see you doing your experiments in the Sandbox.

Do you have any solutions to how we reach the information in the Qualifiers (and References of cource) from WP? I find Lua difficult use and cannot find how we should "read" the information in the Q's and R's from WP. -- Lavallen (talk) 18:33, 24 September 2013 (UTC)[reply]

I have spent part of the day trying to expand fr:Module:Wikidata that was provided by user:HenkvD, but it is not fully ready yet. . I'll think I will have something better by the end of the week The datastructure is provided in mw:Extension:Wikibase Client/Lua. This is something like entity.claims.pXX[1].qualifiers.pYY[1].datavalue.value["numeric id"].
Merci!
Many many tests later. I have partly succeded to use ONE property for several lines in one template. See: sv:Wikipedia:Sandlåda för Wikidata. I never proposed a property for county-letter, but in my test, I have succeded to introduce it! (The last box in the left bottom corner is directly affected by Swedish county code (P507).) -- Lavallen (talk) 14:23, 26 September 2013 (UTC)[reply]
Bien, it's time we start tryiing to use Wikidata in Wikipedia (though I would not try that for everything given that the way we organize item is not very well defined yet). I am mostly done with fr:Module:Wikidata There is surely a lot more to be done, but that is enough for me for now... Note that "GetQualifier" can get the value of a qualifier matching a set of criteria. On fr:WIkipedia:Wikidata/Test, it is used through fr:Modèle:Date clé to retrieve the inauguration date of the item. --Zolo (talk) 16:36, 26 September 2013 (UTC)[reply]
I see that you have a module for Wikidata coordinates. Does it work for non-Wikidaata coordinates too ? I am struggling for finding a way to cleanly integrate Wikidata coordinates with non-Wikidata coordinates. --Zolo (talk) 17:51, 26 September 2013 (UTC)[reply]
The coordinates-module is not designed by me. User:Larske is the creator of the Wikidata-part. It's not designed to handle "novalue" and "somevalue" yet, and not all cases where P625 or any wikibase-entity is present. But I hope User:Larske can help me with that part too, soon.
Actually all coords on svwp go through that module, it has done so since we imported it from enwp. Larske has added the Wikidata-part afterword. It will be called, if the coords are missing in the infobox. Template:Länsfakta (Template:Swedish counties) had no coordinates on svwp, so I proposed him to start to design it for that. The module is designed to handle "precision" in coordnates, and to only write seconds and minutes if the precision is set to show that. Later will we, (I hope) add region: and type: with the help of Wikidata, but we are far from there yet. -- Lavallen (talk) 19:06, 26 September 2013 (UTC)[reply]
The reason I made this test with qualifiers is to make it possible to use one property for several purpose. The talk about the railwaystation-code had no template who could be used as an example, to show that one property can be used for several types of code. P507 is used in Swedish counties, and nothing else. One property for < 30 items looks like a h--l of waste of property-space to me. I think P507 (or something else) could be used to replace a lot of properties for codes for administrative units. Then, we do not have to have to propose new properties for any new code of that kind. Yesterday, I discovered such codes for administrative units in Sweden 1863-1951. The purpose for this code is internal among history-scientists and among genealogists. It has never been used by goverment or for statistical administration, so it's notability as a property here is maybe weak. But if we find a link to genealogy or history-science, it fullfills it's purpose. -- Lavallen (talk) 19:21, 26 September 2013 (UTC)[reply]
Yes, it seems that generic properties can used really well on Wikidata (provided the datastructure is consistent across items and topics). It might deserve some performance analysis though, as it requires looping over all the many possible statements done with the same property rather than just querying a single property. But I suppose the software was developed with that sort of things in mind.
It seems that what you have done is actually very similar to my module, where you can retrieve the NUTS code with {{#invoke:Wikidata|formatStatements|property=P507|qualifier=p31|qualifiervalue=Q193083}}. --Zolo (talk) 19:56, 26 September 2013 (UTC)[reply]
Well, it retrieve all properties P507 who uses P31, if it isn't filtered away. For examle a Swedish municipality is getting a new municipality-code every time it changes county. (Happend in Stockholm County in the 1960's and in several places in 1990's, and more is to come, especially where I live, in the north of Sweden, as soon as the local politicians can agree.) One Swedish municipality can therefore have several municipality-codes here, but only one is relevant on WP. I therefor have to filter away any property using end time (P582). In other cases, like for historical municipalites, it's intresting to see both P131 before the "dissolve date" and P131 today, since "next higher level" when it was "alive" was a county, while it today is another municipality.
I guess we have to wait for phase 3, to see if it is a good idea to merge a lot of properties or not. -- Lavallen (talk) 07:37, 27 September 2013 (UTC)[reply]

Still needed? Multichill (talk) 14:50, 28 September 2013 (UTC)[reply]

No, I have no idea what it was for, I suppose it was used by another template I imported. --Zolo (talk) 14:52, 28 September 2013 (UTC)[reply]
Did it ;) . Greetings, Conny (talk) 20:27, 15 January 2014 (UTC).[reply]

Tableau des métiers

[edit]

Salut Zolo, Je me penche sur le tableau des métiers que tu m'avais indiqué, maintenant que plusieurs codes standardisés ont été acceptés comme propriétés. La colonne des éléments wikidata est partiellement vide (je ne comprends pas pourquoi). Cela pose-il un problème au niveau des correspondances lors de l'import, où pourras-tu effectuer les correspondances manquantes grâce aux libellés (s'ils sont extraits de Wikidata) ?

J'ai récupéré les fichiers de la nomenclature allemande, française et américaine et j'ai un fichier de correspondances que j'ai établi à la main pour le ROME, mais je me pose la question de savoir comment faire ça proprement. https://drive.google.com/folderview?id=0B2Z-7cB-sMWONFZPbldodUN0VG8&usp=sharing --Teolemon (talk) 12:08, 5 October 2013 (UTC)[reply]

Salut, le modèle de Commons datent d'avant Wikidata. Je ne pense pas qu'il y ait de moyen simple d'automatiser la mise en correspondance avec Wikidata, donc il faut tout faire à la main... Si les nomenclatures ne sont pas liées entre elles, ni liées à une base de donnée tiers, je ne pense pas q'il soit possible d'automatiser de manière fiable la mise en correspondance. On peut sans doute imaginer un script qui propose quelque chose en se basant sur les similitudes linguistiques, mais je crains qu'on ait là aussi besoin de vérification manuelle. --Zolo (talk) 14:20, 5 October 2013 (UTC)[reply]

Q506250 and P571

[edit]

Can you take look at how I use inception (P571) in Stockholm Municipality (Q506250)? I try to express what is described in the svwp-article. All primary municipalities in Sweden were transformed in 1970/1971, from several kinds §of municipalties to one single kind. All of that officially (de jure) happend 1971-01-01. But Stockholm was involved in much more than only that kind of transformation. Until 1970 Stockholm was also a landstings-municipality (a consolidated landsting/kommun). Therefor the new munipality-counsil started to work (de facto) already in october 1970, to be able to separate the landsting from the kommun. -- Lavallen (talk) 06:39, 8 October 2013 (UTC)[reply]

That sounds right to me, except that I would not use P31: "instance of de jure" does not seem very meaningful. applies to part, aspect, or form (P518) would seem slightly more accurate but I cannot think of any good qualifier for that. --Zolo (talk) 07:11, 8 October 2013 (UTC)[reply]
Yes, I guess P518 is better. -- Lavallen (talk) 07:16, 8 October 2013 (UTC)[reply]

Bug sur ordre des déclarations

[edit]

Juste une proposition au sujet du bug concernant l'oarganisation des déclarations dans un élément: le plus simple serait un ordre défini pour chaque utilisateur. Chaque utilisateur définirait un ordre sur sa page javascript ou autre page du genre. Cela simplifierait le boulot de développement parce que cela va être dur voire impossible de définir un ordre qui contente tout le monde. Au niveau technique, le plus simple serait de détecter une page qui doit être organisée à l'aide de la propriété instance of (P31) puis une liste des propriétés dans l'ordre d'apparition. Snipre (talk) 10:15, 9 October 2013 (UTC)[reply]

Ca me parait une bonne idée pour ordonner les prorpriétés, mais il resterait à régler l'ordre des déclarations à l'intérieur de chaque propriété. Dans l'exemple que j'ai donné sur bugzilla:44678, il faudrait pouvoir ordonner les architectes par date ou quelque chose du genre, mais on ne dispose pas vraiment de date précise, et la seule solution praticable me parait un règlage manuel. Bien-sûr, ce serait un peu du bricolage, mais je pense qu'en se fiant au sens commun et à quelques conventions, on pourrait règler une bonne part du problème d'affichage. --Zolo (talk) 11:06, 9 October 2013 (UTC)[reply]

Editing translatable pages

[edit]

Hi

When you edit policies and other translatable pages, please do not add the "T:XX" tags manually (like you did here). These tags are added by the translation extension automatically when a translation admin marks an edited page for translation. --Michgrig (talk) 07:38, 18 October 2013 (UTC)[reply]

Ok, sorry, I should give a closer look at the translate extension documentation. --Zolo (talk) 08:45, 18 October 2013 (UTC)[reply]

GZWDer

[edit]

Hi! I have noticed that User:GZWDer is making some mistakes regarding property proposals and creations, but I am not sure that he/she understands my comments posted on User_talk:GZWDer. Perhaps you can help? --Danrok (talk) 21:42, 23 October 2013 (UTC)[reply]

Hi! In this particular case I would think it is ok to create a property proposed by someone else if there has not been any comment for one month (at least for structrually simple properties like geographic codes). We should really find something to make property more reasable and easier to track, but I am not sure what. --Zolo (talk) 06:54, 24 October 2013 (UTC)[reply]

I think we should add a statement to Q5503. Which property can be used?--GZWDer (talk) 06:05, 25 October 2013 (UTC)[reply]

Yes, I was wondering about that too. The main problem is that we do not know the exact meaning of Q5503 (metro line, metro train ?). --Zolo (talk) 08:22, 25 October 2013 (UTC)[reply]
[edit]

Hi Zolo, Just like you did for the ROME code, could you add the authority link for the SOC and NOC codes ? For instance, for truck driver (Q508846) (our best-in class example):

--Teolemon (talk) 10:48, 27 October 2013 (UTC) (still working on getting a magical experience on jobs and occupations :-)[reply]

Hi, I am not very familiar with Javascript; so I will ask for someone's help at MediaWiki:Gadget-AuthorityControl.js. NOC has both a French and an English version, both can be added, but it is a bit annoying that the URL do not match well:
Is there any way to make the French URL more similar to the English one ? --Zolo (talk) 12:56, 27 October 2013 (UTC)[reply]
I'm afraid we're talking about bureaucratic machines, and I'm not an insider :-S--Teolemon (talk) 13:13, 27 October 2013 (UTC)[reply]
Ok. I Do we really need the first digit before the full code, it seems to work just as well without it (http://www5.hrsdc.gc.ca/noc/english/noc/2011/Profile.aspx?val1=7511 http://www5.rhdcc.gc.ca/CNP/Francais/CNP/2011/Profil.aspx?val1=7511 ) . About SOC, is the property restricted to the 2010 version ? That seems to make sense, but the, that should be made explict in the label. --Zolo (talk) 15:32, 27 October 2013 (UTC)[reply]
Turning it into various values didn't do a thing, so I guess it's fine to remove it. I also adjusted the SOC description to be 2010-only (though the structure hasn't changed, more likely additions for new jobs)--Teolemon (talk) 10:11, 28 October 2013 (UTC)[reply]

Pending Array Property requests

[edit]

Hi Zolo, requests in User talk:Legobot/properties.js/requests were not pushed to User:Legobot/properties.js since Oct 15, but only administrators can do that... do you know if there is a systematic scheme for having them processed? WD:AN ? Thanks - LaddΩ chat ;) 22:16, 27 October 2013 (UTC)[reply]

I have done some. We can use {{Edit request}}, though I do not know if many people check it often~, or yes WD:AN if that appears to be too long. --Zolo (talk) 09:34, 28 October 2013 (UTC)[reply]
Thanks for processing a batch. Category:Wikidata protected edit requests is empty, so I guess some admins do check it from time to time... I'll give a try to the template next time. Cheers - LaddΩ chat ;) 22:16, 28 October 2013 (UTC)[reply]

Bonjour Zolo,

Tu évoques sur la page de discussion de P173 la suppression suppression de la propriété. Mais je ne trouve pas trace d'une prise de décision argumentée. J'avoue que j'ai trop de compréhension de la modélisation dans Wikidata mais P173 (P173) me semble plus spécifique que instance of (P31). Le reproche que l'on peut faire c'est que les contraintes sur P173 (P173) n'ont pas été précisées. Dom (talk) 04:58, 30 October 2013 (UTC)[reply]

Suite à la lecture de Dom bot peux-tu m'aider à ajouter la contrainte sur les valeurs possibles sur P173 (P173). Les valeurs possibles seraient : presidential election (Q858439), general election (Q1076105), municipal election (Q152450), primary election (Q669262), European Parliament election (Q1128324), ... . Les Élections régionales ou les élections sénatoriales posent à première vue un problème. Dom (talk) 05:38, 30 October 2013 (UTC)[reply]

Bonjour, j'ai ajouté les contraintes sur la propriété.
La discussion sur p31/p173 est à Wikidata:Requests for deletions/Archive/2013/Properties/1#"Type of" properties. Le résultat était globalement en faveur de la conservation, mais au vu de quelques discussions ultérieures éparses, j'ai l'impression que les avis ont évolué. D'ailleurs deux de ces propriétés ont finalement été supprimées. Disons que vouloir que découper p31 en petites cases du genre "type d'élection", "type de bâtiment" etc. rend la structure mons lisibile et peut poser des problèmes de définition, et par conséquent de chevauchements et redondances, et éventuellement d'instabilité de la structure (si on décide que finalement les ponts ne sont pas des bâtiments, il faut enlever "type de bâtiment" sur tous les ponts, en croisant les doigts pour que ça n'affecte pas trop les sites utilisant nos données). A priori, "type d'élection" ne semble pas poser de gros problème là dessus mais c'est difficile à savoir par avance. D'ailleurs, pourquoi les élections régionales ne devraient-elles pas utiliser cette propriété ? Une propriété "type d'élection" que l'on ne peut utiliser que pour cetains élections me semble assez problématique.
Il me semble que la principale utilité de "type d'élection" serait d'aider à définir la structure de l'élément, en suggérant des contraintes du genre : "une élection doit avoir un élu", mais on n'a pa vraiment besoin d'une propriété spécifique pour ça (les contraintes par type vont déjà dans ce sens). -Zolo (talk) 06:45, 30 October 2013 (UTC)[reply]

Please replace 1014/ by 1014:. --Fomafix (talk) 09:59, 8 November 2013 (UTC)[reply]

Done, thanks. There seems the tool seems to be broken though. --Zolo (talk) 10:14, 8 November 2013 (UTC)[reply]
Hello, Zolo. 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 10:22, 10 November 2013 (UTC)[reply]

P518

[edit]

You are the inventor of applies to part, aspect, or form (P518) as far as I remember.

In Unnaryd municipality and some other items, am I trying to describe both how they were split and how it looks like according to located in the administrative territorial entity (P131) today.

The problem is that Bolmsö civil parish has been split between three different municipalities today, and I have not enough items to describe its parts.

The island of Bolmsö (Q891514) was given to Ljungy already 1971. At the same time, the north non-island part of Bolmsö civil parish was given to Värnamo. 1974 the west non-island-part of Bolmsö civil parish, what was left, was given to Gislaved.

Should I create items about "the west non-island-part of Bolmsö civil parish" and "the north non-island-part of Bolmsö civil parish" or do you have any other idea? The third part, "the island of Bolmsö" have an item: Q891514, but that has a notability in itself, which the other parts haven't.

And is there any good word in English to describe the part of something, that isn't located on an island? See File:Västbo Bolmsö.svg for some graphics. -- Lavallen (talk) 13:28, 23 November 2013 (UTC)[reply]

If I understand correctly, Bolmsö civil parish no longer exists. I am not sure we should try to make it fit into p131 for dates posterior to its dissolution (if the administrative unit borders have moved through time, that may require two date qualifiers and would become really complicated).
I do not think there is any English word for "not on an island" ("continent" is sometimes used by that would not do). However, I think we can use excluding (P1011) here. --Zolo (talk) 14:49, 23 November 2013 (UTC)[reply]
Bolmsö civil parish has no administrative function anylonger, except as a geographic entity. Therefor, Bolmsö civil parish exists and it doesn't. Swedish National Heritage Board (Q631844) is for example using the civil parish-division of Sweden to describe where all protected cultural heritage locations can be found. There are several hundred thousands of them. On svwp is the articles about the civil parishes used to describe the nature and history of Sweden. Such things as lakes are also divided by civil parish. The civil parishes are also alive in how the people describe their private history. Me for example, was born in Bäckseda civil parish, the hundred of Östra härad and the Province of Småland. Neither of these has any administrative function, but they are alive as geographic and ethnical enitities.
Historical municipalities on svwp have a parameter for what it belongs to today. See for example parameter "nuvarande kommun" (current municipality) in sv:template:Historisk kommun in sv:Ytterlännäs landskommun. In contrast to civil parishes, historical municipalities like Ytterlännäs neither has any administrative or any geographic importance today. Creating a separate property beside P131 for "current municipality" looks farfetched. Instead am I using P131 with start- and stop-dates to describe where they belonged when they were alive and where they can be found today, after they have been dissolved. Unnaryd belonged to Jönköping County until it was dissolved in 1973/74 for example, when it was a live independent municipality.
I'm not sure sure how excluding (P1011) can be used in Q10432623?
P131:Jönköping county.           stop-date: 1862-12-31
P131:Bolmsö rural municipality.  start-date: 1863-1-1, stop-date: 1951-12-31
P131:Unnaryd rural municipality. start-date: 1952-1-1, stop-date: 1970-12-31
P131:Unnaryd municipality.       start-date: 1971-1-1, stop-date: 1973-12-31. P1011: Bolmsö island, P1011 The non-island part in the north
P131:Ljungby municipality.       start-date: 1971-1-1,                        P518: Bolmsö island
P131:Värnamo municipality.       start-date: 1971-1-1,                        P518: The non-island part in the north
P131:Gislaved municipality.      start-date: 1974-1-1.                        P1011: Bolmsö island, P1011 The non-island part in the north
I'm not sure P518 and P1011 can do it alone when it was split in three parts? -- Lavallen (talk) 17:12, 23 November 2013 (UTC)[reply]
Note sure either acutally. But if there is no real hiercharchy between types of units, I do not think we should try to express everything in detail. It seems that in the absence of a geoshape datatype, it cannot be always fully accurate, and something too complicated may be hard to understand and to reuse. I guess that if we want to say in an infobox that a building is both Ljungby municipality and in Bolmsö civil parish, we should add two values to its p131.--Zolo (talk) 18:11, 23 November 2013 (UTC)[reply]
"in the absence of a geoshape datatype" is not the only problem. Somebody have to create those geoshape-files when the datatype is availible. We have users, archives with material and the knowledge to do it. But it will take them a lot of time.
I agree that we should allow several kinds of administrative units in P131. The problem today is that we today cannot get information about the nature of the items in the properties from wp.
Does a "hiercharchy between types of units" exist at all anywhere? In Sweden is there sometimes a competition between municipalities and county counsils. They can argue about who has responsibility to take care of helthcare for elder people, but nobody is above the other... -- Lavallen (talk) 20:27, 23 November 2013 (UTC)[reply]
I am not sure that there is anywhere a nice and tidy hierarchy of roles, but sometimes some types of divisins are always subidivided in other types of juridiction. In Taiwan, there are even types of divisions that are officially called "level 2", "level 3" etc. In France, a region is always subdivided into departements (even though some regions only have one département). That does not mean that the distribution of roles is clear (in fact it is an appalling mess), but it makes it rather straightforward to say that one unit is in another. --Zolo (talk) 08:50, 24 November 2013 (UTC)[reply]

Q57742 reply

[edit]

see User_talk:Droadnaegel

Salut Zolo, excuse-moi pour disparaître pour ainsi longtemps s'il te plaît. Je doit transmettre un ouvrage le mars, donc je suis bien occupé avec ça. C'est pourquoi il me faut de trouver un chemin de me retirer de Wikidata pour au moins quelques mois ;). Tu pourrais simplement me tinter si tu penses qu'il y a quelque chose que j'ai à écrire ou faire.

Quant à User:Marsupium/AAT, nous pouvons resoudre les problèmes ensemble à un temps donné? Je crains qu'il n'est pas raisonnable de travailler au contenu de la liste avant la solution. Malheureusement, c'est mon premier modèle, c'est pourquoi je suis presque lent. Cordialement, --Marsupium (talk) 15:41, 30 November 2013 (UTC)[reply]

Salut, pas de problème, tu reviendras quand tu auras le temps. Pour User:Marsupium/AAT, que voudrais-tu y faire. Je pense que le modèle marche à peu près maintenant, et il parait préférable de remplir les choses à la main, donc il me semble qu'il ne reste plus qu'à remplir les choses par progressivement. --Zolo (talk) 15:53, 30 November 2013 (UTC)[reply]
Merci. D'accord il me faut revoir le modèle.
Je renomme des pages ATM. Ça faisant j'ai trouver Wikidata:Artworks task force/Infoboxes. Je pense que cette page peut être supprimée. Est-ce que tu peux en prendre soin? Merci d'avance, --Marsupium (talk) 14:03, 5 December 2013 (UTC)[reply]
Fait, merci, c'est sûr qu'il y a toujours du ménage à faire ! --Zolo (talk) 18:04, 5 December 2013 (UTC)[reply]

Reminder when fulfilling edit requests

[edit]

Hi! When fulfilling edit requests, please remove or disable the edit request template (example). Thanks, The Anonymouse (talk) 06:49, 1 December 2013 (UTC)[reply]

yes, sorry I forgot it. --Zolo (talk) 18:48, 1 December 2013 (UTC)[reply]
OK :) The Anonymouse (talk) 20:59, 1 December 2013 (UTC)[reply]

Property documentation

[edit]

Hi Zolo, your edit of the Template:Property documentation produces some errors. (Sources are now "Allowed values" etc.) Can you please take a look at it? Thanks --Kolja21 (talk) 17:17, 5 December 2013 (UTC)[reply]

Done, I have tried to clean up the code at the same time as it was rather hard to read, hope I did not miss anything. --Zolo (talk) 18:03, 5 December 2013 (UTC)[reply]
Thanks, that was fast. Minor problem: Data type has still two colons but I don't know how to remove one. --Kolja21 (talk) 18:20, 5 December 2013 (UTC)[reply]
Thanks for notifying. There are also issues with the "proposer" parameter. As it is currently a bit hard to read this template, I am trying to make a new version, hoping it will become more maintainable. --Zolo (talk) 18:25, 5 December 2013 (UTC)[reply]
I have changed the translatewiki message to get rid of the double column (in all cases I have checked it will be an improvement, I hope I didn't miss anything)
Rereading the template, I am not sure I will really rewrite it. I think some things should be done differently, but it will be tedious to change them now that the template is used in hundreds of pages. --Zolo --Zolo (talk) 18:59, 5 December 2013 (UTC)[reply]

Ping Project

[edit]

I tried out ping project, but it didn't work. Did I make a mistake in Wikidata:WikiProject_Molecular_biology/Participants? or by using {{Ping project|Molecular biology}}? --Tobias1984 (talk) 18:19, 16 December 2013 (UTC)[reply]

It worked, I got a notification ;). You didn't because receeive any notification because you do not get notification when you mention yourself. --Zolo (talk) 18:22, 16 December 2013 (UTC)[reply]
Uh actually I don't know how I got notified while I am not in the list... --Zolo (talk) 18:25, 16 December 2013 (UTC)[reply]
Can you use it somewhere, so I can see if I will be notified? --Tobias1984 (talk) 20:57, 16 December 2013 (UTC)[reply]

Erreur de manip ?

[edit]

Bonjour Zolo, il me semble que tu as fais une erreur de manip avec Widar, puisque tu as ajouté sur une vingtaine d'articles sex or gender (P21) = Thierry of Chartres (Q658109) (cf. https://www.wikidata.org/w/index.php?namespace=&tagfilter=Unexpected+value+for+gender&translations=noaction&days=1&title=Special%3ARecentChanges)... — Ayack (talk) 19:31, 25 January 2014 (UTC)[reply]

Effectivement, merci de prévenir. C'est visiblement une erreur de copier-collé de male (Q6581097). J'ai refait à la main, je ne crois pas qu'on ait d'autres solutions --Zolo (talk) 21:46, 25 January 2014 (UTC)[reply]

Gestion des anciennes communes

[edit]

Bonjour Zolo,

Existe-t-il une façon de faire spécifique pour les anciennes communes de France ? Par exemple pour Veneffles Q3555470, j'ai « simplement » ajouté les qualificateurs « date de début » et « date de fin » ; est-ce la bonne façon de faire ? (date de fondation aurait-elle été plus approprié ?). Sinon, faut-il laisser Veneffles en alias sur Châteaugiron Q380027 ? (Help:Aliases/fr ne dit rien sur ce sujet particulier)

Enfin de manière plus générale, existerait-il une commune de France « parfaite » qui pourrait servir d'exemple pour les autres ? J'ai l'impression que cet ajout est une erreur mais je suis incapable de retrouver la règle ou une page de référence pour le confirmer…

Cdlt, VIGNERON (talk) 14:47, 30 January 2014 (UTC)[reply]

Bonjour Vigneron, je pense que les qualificatifs "date de début" et "date de fin" sont la meilleure façon de faire. Je ne sais pas si "date de fondation" conviendrait mieux. En fait je pense que le plus simple serait de carrément supprimer la propriété date de fondation.
Je ne crois pas qu'on ait de commune modèle pour l'instant. En fait, tant qu'on ne peut pas ajouter de données numériques, les entrées sont forcément assez succinctes (enfin on trouve toujours des choses à dire sur Paris (Q90)))
Pour la modification de Dexbot, il y a des discussions pour savoir s'il faut conserver located in the administrative territorial entity (P131) ou la fusionner avec instance of (P31). Les deux solutions ont des inconvénients, et il ne semble pas encore se dessiner de consensus.sur Wikidata:Properties for deletion#Property:P132 (type of administrative division --Zolo (talk) 16:28, 30 January 2014 (UTC)[reply]
I have added some statements to the items listed in Wikidata:Country subdivision task force/Sweden/Municipalities/Ex, who all are obsolete commune suédoises, today all merged with other communes, but that is not described in all items (yet). - -- Lavallen (talk) 18:11, 30 January 2014 (UTC)[reply]
Pour la suppression éventuelle de « date de fondation », je te laisse faire.
Pour Veneffles, j'ai du coup une autre question, faut-il indiquer qu'avant c'était une paroisse et qu'après c'est une portion de la commune de Châteaugiron ? ou bien est-ce too much ? Je vois que Ytterlännäs rural municipality (Q13604045) utilise précédé par/suivi par (or il n'y a pas d'élément pour la paroisse et il n'y en aura sans doute jamais).
PS: j'attend avec impatience les données numériques (je commence dès maintenant à chercher des bases de données sources pour ces données et à réfléchir pour voir comment cela va s'articuler), en attendant je me familiarise déjà mieux avec les bases de WikiData .
Cdlt, VIGNERON (talk) 21:05, 30 January 2014 (UTC)[reply]
I have here used founding date/dissolved date, since all of these entities have had the same status in all their lifetime. That is not always the case. "Norrköping city" was founded as "city" 1863, but later was transformed to a "kreisfree city" and later back to an ordinary "city" before it was dissolved 1971. Both the communes of Gothenburg and Malmö was founded as "kreisfree communes" 1971 but was transformed into ordinary communes a few years ago. -- Lavallen (talk) 07:19, 31 January 2014 (UTC)[reply]
@VIGNERON:. Ce serait vraiment bien si tu pouvais faire qqch. Si on le fait rapidement, ça permettrait en plus d'éviter que les gens commencent à la main, ce qui obligerait à repasser derrière pour harmoniser les formats et les chiffres. J'imagine qu'il faudrait utiliser l'INSEE, et peut-être cassini pour les dates plus anciennes. Cela dit, il y a eu un avis des juristes de la WMF qui disaient qu'il valait mieux ne pas utiliser en masse sur Wikidata, qui est CC0, les donnés publiques européennes qui, ont généralement une licence imposant de citer les auteurs; Si c'est le cas, et qu'on ne peut pas changer la licence de Wikidata, beaucoup de choses risquent d'être compliquées... --Zolo (talk) 15:01, 31 January 2014 (UTC)[reply]
Sauf que je sais pas (encore ?) éditer WikiData autrement qu'à la main… J'ai d'ailleurs demandé de l'aide à Kvardek du pour ajouter les codes INSEE sur les quelques communes de France où ils ont été oubliés. Il a (va avoir) un bot qui pourra peut-être s'en charger.
Pour les données tel que coordonnées, superficie, altitudes max/min, etc., la source officielle est le [professionnels.ign.fr/rgc RGC] de l'IGN est sous licence IO qui correspond à un CC-BY ; or j'ai cru comprendre que le BY posait problème mais in fine je pense que ce sont des données DP (puisque issues du cadastre et équivalent, une sorte de PD-law à la française…). D'ailleurs la plupart des données des infoboxes de communes françaises sur fr.wp sont issues du RGC ; donc importer depuis Wikipédia plutôt que du RGC ne fait que « masquer » la source originale.
Sinon, pour les anciennes communes de France, je peux faire un tableau récapitulatif comme celui Lavallen pour les anciennes communes Sweden. Est-ce que tu saurais ensuite l'utiliser pour remplir les éléments correspondants ?
Cdlt, VIGNERON (talk) 15:58, 31 January 2014 (UTC)[reply]
Effectivement, je ne sais pas pourquoi j'avais en tête les seules données démographiques (peut-être en fait parce que les données de type superficie altitude ne sont pas encore disponibles alors que les populations sont considérées comme sans unité et donc disponibles depuis aujourd'hui.).
C'est vrai qu'on a des incompatibilités potentielles de licence pour tout ce qui est importé de Wikipédia, surtout quand ça provient de source extérieur, et qu'en fait c'est peut-être aussi bien de ne pas en tenir un compte trop métinculeux. J'imagine que si les choses sont vraiment ingérables, il faudra passer Wikidata en licence avec attribution pour la rendre compatible avec Wikipédia et les grandes bases de données.
Je ne sais non plus me servir d'un bot sur Wikidata, mais je pourrai essayer. Ca ne devrait pas être trop compliqué de transformer la table en csv, puis de faire manger le csv à un bot --Zolo (talk) 18:51, 31 January 2014 (UTC)[reply]
Bonjour. Pour l'instant mon robot va comparer les codes Insee entre les articles de Wikipédia en français et ceux de Wikidata (son travail est visible ici). Il va ensuite traiter la liste d'erreurs (a priori un peu moins de 200). Il travaille effectivement (quoiqu'indirectement) à partir d'un tableur. Pour Wikidata, j'ai fait une requête de statut de bot mais pas vraiment d'avis pour le moment. Kvardek du (talk) 19:36, 1 February 2014 (UTC)[reply]

Q848570 - Boulevard du Temple

[edit]

Hello,
Can you please explain me why shouldn't Q848570 be deleted after being merged with Q1851236, or why (huwiki: Boulevard du Temple (dagerrotípia) ) shouldn't be merged into Q848570? I just couldn't understand it. Thanks in advance, Ldorfman (talk) 20:04, 8 February 2014 (UTC)[reply]

Hello Ldorfman, Boulevard du Temple (Q1851236) is about a street in Paris and Boulevard du Temple (Q848570) is about a photograph of the street (mostly known for being the first photo showing a human being). --Zolo (talk) 06:23, 9 February 2014 (UTC)[reply]
OK. Understood. Thanks. Ldorfman (talk) 08:20, 9 February 2014 (UTC)[reply]

Hello, do you use Category:User Zolo/maintenance, or could it be deleted? Thanks.  Hazard SJ  21:12, 22 February 2014 (UTC)[reply]

Can't even remember what the purpose was... Deleted. --Zolo (talk) 14:59, 23 February 2014 (UTC)[reply]

P1198

[edit]

Wikidata:Project_chat#unemployment_rate_of_state_X --Goldzahn (talk) 11:22, 3 March 2014 (UTC)[reply]

Date et Date de fondation/création

[edit]

Bonjour Zolo,

Dans Wikidata:WikiProject Visual arts/Item structure, j'ai changé point in time (P585) pour inception (P571). Une date doit être liée à un événement. Or dans le modèle l'événement création est implicite avec point in time (P585) et n'est donc pas renseigné. Pourtant il pourrait y avoir plusieurs types de dates pour une œuvre, création, restauration, publication... Avec inception (P571), l'événement est explicite et la propriété est faite pour ça. Il me semble donc que la propriété inception (P571) est bien plus appropriée. Bien à toi Shonagon (talk) 01:18, 11 March 2014 (UTC)[reply]

Bonjour, je suis Shonagon et j'ai utilisé inception (P571) souvent. Regards--Oursana (talk) 01:40, 11 March 2014 (UTC)[reply]
Bonjour Shonagon et Oursana, ça parait effectivement plus rigoureux. Cela dit, c'est un peu moins intuitif, et il faudrait sans doute trouver un bot capable de faire épisodiquement la conversion. --Zolo (talk) 10:57, 11 March 2014 (UTC)[reply]
Tout à fait. Si pour une part des contributeurs actuels, l'usage de inception (P571) est déjà présent, il conviendra en effet de reprendre l'existant et de prévoir des remplacements automatiques à l'avenir. --Shonagon (talk) 12:53, 11 March 2014 (UTC)[reply]

Dates de création d'œuvres

[edit]

Bonjour Zolo,

Je prévois de contribuer très prochainement sur les dates de création des items art visuel. Le manque est énorme : en gros sur un lot de 8000 oeuvres d'art visuel que j'ai recencées, il semblerait que moins de 300 aient une date – inception (P571) ou point in time (P585) –. (Peut-être que ma requête est erronée et sous-estime, mais une navigation dans les items prouvent bien que le manque est au moins très important). Je me propose de faire la récupération des années de création depuis DBpedia, réédition en local, publication via bot (au passage je pourrais changer les point in time (P585) en inception (P571)). Pour les dates uniques et précises avec une année, pas de difficultés. Mais j'ai un souci sur les périodes de création. Par exemple pour la Joconde, créée entre 1503 et 1506, comment formaliser cette période ? On pourrait utiliser 1500s-précision décade. Peut-être est-il possible de formaliser "entre 1503 et 1506" mais je n'ai pas trouvé. Mettre 2 dates ? Et pour une œuvree crée entre 1495 et 1505, que faire ? Sachant qu'il y a aussi beaucoup d'œuvres antiques, voire préhistoriques, pour lesquelles la date est quasiment sytématiquement imprécise.

Si tu pouvais éclairer ma lanterne sur le sujet, je t'en serais grandement reconnaissant. Bien à toi --Shonagon (talk) 08:26, 22 March 2014 (UTC)[reply]

Bonjour Shonagon,
Il y a effectivement un souci avec les dates imprécises. A un moment, j'avais cru comprendre qu'il était question d'autoriser les valeurs du type 1595 +/-2 ans mais je ne sais pas si c'est encore à l'ordre du jour. Par ailleurs, peut-être faudrait-il traiter différemment le cas : "peint à un moment entre 1503 et 1506" et le cas "commencé en 1508, achevé en 1512" (Sistine Chapel ceiling (Q844675)). Pour le deuxième cas, je pense qu'il faudrait essayer de trouver avec le développement une solution technique. Pour le premier, il serait peut-être en fait plus correct d'avoir quelque chose du genre : "date de création : inconnue, qualificatif 1 (date de début ?) : 1503, qualificatif 2 (date de fin ?) : 1506. --Zolo (talk) 06:07, 23 March 2014 (UTC)[reply]
Merci Zolo. Oui il y avait aussi la différence "créé ente x et y" et "créé de x à y". La solution "date de création : inconnue" avec en qualificatif les précisons de date ne me convainc qu'à moitié car si la date exacte est bien inconnue, la date dans son extension de période est généralement connue. En revanche la proposition 1595 +/-2 ans est très séduisante et permettrait de préciser au mieux toutes les périodes uniques de création. En effet il conviendrait sans doute de porter ces questions ailleurs. En attendant je vais me concentrer sur le traitement des dates simples et des périodes avec précisions d'échelle (décade, siècle...) et pourrais mettre de côté les cas diffciles. Le gros souci de ce problème de date et que, comme j'ai pu le constater, cela rend le classement chronoligique des œuvres au minimum beaucoup trop incomplet. Or en histoire de l'art la datation est évidemment un élément essentiel de classification et pour la réutilisation des données de Wikidata dans ce domaine une solution, même imparfaite, est vraiment nécessaire. Merci encore pour ces éclaircissements. --Shonagon (talk) 14:18, 23 March 2014 (UTC)[reply]
En creusant un peu la question, me suis rendu compte que le type date devrait permettre, comme tu l'évoquais, de spécifier des valeurs pour avant/précédent ; malheureusement ce n'est pas encore opérationnel pour l'édition, cf after/before_pour_une_date --Shonagon (talk) 20:46, 25 March 2014 (UTC)[reply]
Merci. Oui la solution inconnu + qualificatif ne me convainc pas vraiment non plus mais je ne sais pas trop comment faire la différence avec les deux cas.
Effectivement, on voit qu'il y a une antrée "before" et "after" dans les données de type date ici (je ne sais pas pourquoi cette page a été effacée, je ne vois son équivalent nulle part ailleurs).
Enfin en attendant j'imagine qu'il y a déjà pas mal à faire avec les dates simples. --Zolo (talk) 08:20, 26 March 2014 (UTC)[reply]
Voir Help:Modeling#Time & dates; problème similaire est date de la naissance, vous vous rappelez Wikidata:Bistro/Archive/2013/11-12#« né vers 1132 » ou « né entre 1132 et 1150 »Regards--Oursana (talk) 11:10, 26 March 2014 (UTC)[reply]
Oui, Oursana, j'ai lu beaucoup de choses avant de me lancer. Résumons, oui il est possible d'éditer les dates avec précisions avec un bot, oui c'est un problème pour l'interface web pour l'édition, l'affichage et les modifications. Il ya un bug ouvert https://bugzilla.wikimedia.org/show_bug.cgi?id=61909 auquel j'ai ajouté un commentaire. Tobias m'a dit sur IRC que la correction était prévue mais sans échéance prévue ; n'hésitez pas à vous y inscrire et à le faire remonter pour une correction prochaine, car en effet c'est un sérieux problème pour les datations en particulier pour les œuvres d'art. En attendant j'ai repris les données de DBpedia et vais les publier via bot. Pour exemple pour Mona Lisa (Q12418) à l'affichage on a inception (P571) 1503 mais on peut constater que les données stockées https://www.wikidata.org/w/api.php?action=wbgetentities&ids=q12418&format=jsonfm sont 1503, precision 9, before 0, after 3, ce qui correspond bien à 1503-1506. Voilà on va enfin avoir globalement des inception (P571) pour les œuvres d'art –le manque est criant–, y compris pour les nombreuses dates floues, mais le bug de l'interface reste ouvert et il est vraiment nécessaire de le corriger. Bien à vous. --Shonagon (talk) 16:35, 30 March 2014 (UTC)[reply]

Louvre

[edit]

Hoi, According to the English Wikipedia, the Louvre is in the category of French castles. Can you explain why this is wrong? Thanks, GerardM (talk) 12:38, 26 March 2014 (UTC)[reply]

Hi GerardM, Louvre Museum (Q19675) is about the Louvre Museum, taken as an institution. The item about the building is Louvre Palace (Q1075988). Actually, I am not sure it makes sense to have "instance of castle" in Louvre Palace (Q1075988) either, because the current building is not a castle, it is a 16-19th century palace that replaces a medieval castle, but the medieval castle has been almost entirely demolished.

It was merged item that reused by Dae. Reusing empty items are not allowed. See Wikidata:Administrators'_noticeboard/Archive/2014/02#Reusing_empty_items for details.--DangSunM (talk) 13:30, 6 April 2014 (UTC)[reply]

@분당선M: Oh ok it was a reused item. If you merge item please make sure that everything has been moved. Q1289793 contained statements and an incoming link from Q15829898 that should have been moved to Q15838917 before deletion. -Zolo (talk) 16:26, 6 April 2014 (UTC)[reply]

Widar of zolo - Club-Mate (Q53)

[edit]

Hey,
seems not right [1]. Greetings, Conny (talk) 17:34, 18 May 2014 (UTC).[reply]

Thanks. Can't find how I got this one though. --Widar of zolo (talk) 19:07, 18 May 2014 (UTC)[reply]

bombilla is not a human

[edit]

An edit made in the last hour by Widar of zolo (talkcontribslogs) — bombilla (Q891922): definitely not instance of (P31) => human (Q5). Was it automated? Is there any way to tell what caused this to happen? --Closeapple (talk) 19:31, 18 May 2014 (UTC)[reply]

@Closeapple: It was semi-automated. I was using es:Deportistas por deporte but had tried to check it manually. I had apparently missed a few including this one and Club-Mate (Q53) mentioned above (though I do not see through which subcategory it was). I have rechecked the whole batch, and fixed a few more, the rest seems to be ok, but let me know if you find others. --Zolo (talk) 20:09, 18 May 2014 (UTC)[reply]

Also this timeperiod is not a human being. Sincerely, Taketa (talk) 16:59, 26 May 2014 (UTC)[reply]

P31 = Q5 pour les groupes d'humains ?

[edit]

Bonjour,

Je viens de découvrir le Wikidata Game (et je m'amuse comme un petit fou) mais j'ai un doute : faut-il retirer l'affirmation instance of (P31) = human (Q5) pour les groupes d'humains ? J'ai l'impression que oui mais j'hésite… En tout cas, j'en ai flaggé plusieurs sur Wikidata:The Game/Flagged items.

Cdlt, VIGNERON (talk) 13:18, 29 May 2014 (UTC)[reply]

Bonjour VIGNERON,
Il faut utiliser group of humans (Q16334295). Zolo (talk) 07:12, 30 May 2014 (UTC)[reply]

Hallo Zolo, pourquoi as-tu annulé ceci? Cordialement --Oursana (talk) 15:26, 5 June 2014 (UTC)[reply]

Bonjour Oursana. En fait, j'ai un peu hésité, mais il me semble que occupation (P106) convient mieux (ça me parait être logique de mettre occupation (P106) pour l'activité principale de la personne (qu'elle soit rémunérée ou non) et de garder p31 simple en ne mettrant que human (Q5) pour les humains.
Plus techniquement, je pense qu'il vaudrait mieux faire d'ermite (et de peintre et de boulanger) des sous-classes (p279) d'"occupation" plutôt que d'"humain". Dans ce cas, dire mettre p31: ermite supposerait de dire "Vittore de Camerino est une occupation". --Zolo (talk) 17:49, 5 June 2014 (UTC)[reply]
Merci Zolo--Oursana (talk) 21:30, 5 June 2014 (UTC)[reply]

Bonjour Zolo, cet article était mu plusieurs fois. Est-ce administrative territorial entity of Japan (Q1850442) ou prefecture of Japan (Q50337) [2]? Merci--Oursana (talk) 16:01, 9 June 2014 (UTC)[reply]

Bonjour Oursana. C'est administrative territorial entity of Japan (Q1850442), même si l'essentiel de l'article est occupé par une liste des préfectures. --Zolo (talk) 20:32, 9 June 2014 (UTC)[reply]
Merci beaucoup--Oursana (talk) 20:41, 9 June 2014 (UTC)[reply]

Thanks, and follow up

[edit]

First, thank you for your helpful reply at Wikidata:Project_chat/Archive/2014/06. Since this thread was archived, let me elaborate a bit more. In addition to be being an active article creator, I have another hat - a Wikipedia researcher. I am planning to carry out a research project that would try to investigate gender inequality worldwide using Wikipedia biographies through time and space. Simply put, I want to create nice graphs (which we could host at Commons and which could improve numerous Wikipedia articles, up to and including the country-specific series of 100+ articles on gender inequality in country x), as well as tables, about the disparity between our biographies of men and female by year by ethnicity/nationality. I have already designed a working spreadsheet at [3] to illustrate what can be done. To finish this project, however, I need to extract data from Wikipedia, and I simply lack the skills to do that. Do you know where, or whom I could ask to extract such data for me, preferably in the form of the csv file formatted as in the sample spreadsheet linked? (If you reply here, please echo me - thanks). --Piotrus (talk) 13:17, 21 June 2014 (UTC)[reply]

Thanks for the explanation user:Piotrus, but I do not undesrstand what a "series of 100+ articles on gender inequality in country x" is. From the link you give, it seems that you want to have stats by nationality + sex. It means that you will have to deal with large amounts of data, and I guess I guess the simplest solution for that would be to work directly on the dumps. You can try to get in touch with User:Maximilianklein who has already worked on sex on Wikidata.
Actually, I am a bit unsure that you can get a very usable CSV file combining sex / birth date and nationality. I would imagine that a more practical solution would be creating a custom lightweight database of people in Wikidata and try to play with that, using various criteria. But I am neither a developer nor a data scientist.
Anyway, if you just want stats about gender by year, I have created a mini-script using Wikidataquery and tried it for the years 1000 and 1001. But running it for the whole 1000-2000 period may take several hours.--Zolo (talk) 11:30, 22 June 2014 (UTC)[reply]
Year male (Q6581097) female (Q6581072) Other sex or gender (P21) No sex or gender (P21)
1000 17 1 0 0
1001 9 1 0 0

Caméra Café

[edit]

Salut ! Peux-tu à nouveau nettoyer le bazar entre Caméra Café (Q3654788) et Caméra Café (Q2700478) ? Ça a de nouveau été fusionné, et j'ai un message d'erreur quand j'essaie d'annuler la modification, je me dis que tu y parviendras peut-être ? --Harmonia Amanda (talk) 11:06, 4 August 2014 (UTC)[reply]

Salut Harmonia Amanda, désolé pour le temps de réponse, apparemment ça a déjà été réglé. Je pense que le message d'erreur venait d'un conflit d'interwiki. Comme un même lien wiki ne peut être que dans un seul élément, il faut faire attention à enlever les interwikis ajoutés avant de remettre ceux qui ont été enlevés. --Zolo (talk) 13:08, 5 August 2014 (UTC)[reply]

Wikisource :)

[edit]

Salut Zolo,

En cherchant les Auteurs de wikisource, je me suis aperçue que tu as créé les textes de fr.wikisource... tous ?

En tout cas, j'ai entrepris d'en traiter un certain nombre, sur lesquels je suis tombée par le wikidata game : peux-tu me donner ton avis ? [361%3A%EF%BB%BF%EF%BB%BF474845&mode=undefined&statementlist=P31%3Aq5185279%0AP1433%3Aq17356468%0AP407%3AQ150%0A-P361%3AQ%EF%BB%BF%EF%BB%BF474845&run=Run&label_contains=&label_contains_not=&chunk_size=10000 Le Parnasse contemporain] et Bric-à-brac (Q17355901)

Pour ces poèmes, s'agit-il de travailler au niveau de l’oeuvre, ou au niveau de sa manifestation (au sens du FRBR ? en effet, la plupart figurent ailleurs sur wikisource, avec une page d'"éditions" qui renvoie vers les différentes manifestations (en particulier, les poèmes de Baudelaire).

Je me demandais si, du point de vue de l'Oeuvre, il ne vaudrait pas mieux faire le lien sur la page d'éditions ?

à titre d'exemple : Q17356046 est actuellement lié à fr:Le Parnasse contemporain/1866/Recueillement mais pourrait l'être, en tant que poème sur la page fr:Recueillement

Ne connaissant pas l'avancée des discussions sur les data pour les oeuvres textuelles, je te dépose ici mes réflexions sur le sujet. Qu'en dis-tu ?

Amicalement, --Hsarrazin (talk) 20:21, 8 August 2014 (UTC)[reply]

Discussion transférée sur la page de discussion du projet que je viens de trouver :D --Hsarrazin (talk) 08:00, 9 August 2014 (UTC)[reply]
Salut Hsarrazin, j essaierai de mieux repondre quand j aurai davantage acces a internet. Disons pour l instant que je me suis conforme aux titres Wikisources. Si la page s appelle le Pere Goriot sans precision d edition je le traite comme une oeuvre, si l edition est precisee comme une maniefestation. Il me semble qu on pourrait faire mieux mais je me demande s il ne feaudrait pas aussi reorganiser legerement les choses sur Wikisource elle meme dans le meme temp.s--Zolo (talk) 08:47, 10 August 2014 (UTC)[reply]
il y a certes de la réorganisation à faire sur wikisource... et des décisions à appliquer, avec beaucoup de travail à faire :)
mon questionnement était plutôt sur l'interprétation par wikidata des données relatives à un poème : oeuvre ou/et édition ? et du/des liens à mettre en conséquence… :) --Hsarrazin (talk) 12:22, 10 August 2014 (UTC)[reply]
@Hsarrazin: La recommendation générale est de créeer un élément édition et un élement oeuvre pour les livres, mais de grouper les deux dans le même élement pour les articles. Etant donné qu'un poème est généralement publié dans un recueil, la situation parait se rapporcher de celle des articles, donc je dirais un seul élément pour l'édition et l'oeuvre. Si toutes les publications du poème utilisent le même texte, on devrait pouvoir documenter cela avec published in (P1433). S'il existe des variantes ou des traductions du poème, je ne sais pas comment procéder.--Zolo (talk) 07:06, 13 August 2014 (UTC)[reply]
merci Zolo, c'est tout à fait clair...
il ne devrait pas y avoir de problème pour la majorité des poèmes de wikisource... mais Baudelaire va être un peu plus délicat - je pense que je vais faire ici comme sur wikisource : une seule fiche (item), jusqu'à ce que j'en trouve un deuxième :D --Hsarrazin (talk) 07:17, 13 August 2014 (UTC)[reply]

Bonjour Zolo, j'ai changé de-, cs, pap et svwiki, et je crains, qu'il faut faire la même chose avec les autres contributions du User:UU. Il y a aussi problèmes avec son contributions à Tokyo (Q11199581). Je t'écris pour rassurements, après j'essayerai à refaire ces changements.

J'ai proposé à joindre en:Tokyo Metropolis et en:Tokyo sur en WP.Cordialement--Oursana (talk) 10:39, 11 August 2014 (UTC)[reply]

Bonjour Oursana. Les choses ont l'air effectivement un peu bancales en ce moment. Si tu as le courage de te lancer dans le nettoyage et de documenter tes choix sur les pages de discussion correspondante, ce serait sans doute une bonne oeuvre, mais, vu le nombre d'articles impliquées, ce n'est pas forcément évident. J'avais essayé de le faire, mais il y a quelques mois, un utilisateur avait bougé des liens, et j'ai eu la flemme de me lancer dans une grande discussion. En tout cas, si Tokyo (Q1490) concerne la préfecture de Tokyo, le lien japonais devrait être ja:東京都 plutôt que ja:東京. Je crois qu'on devrait avoir :
  1. Tokyo (Q1490) sur la préfecture
  2. ward area of Tokyo (Q308891) sur les 23 special wards
  3. Tokyo (Q7473516) sur Tokyo en général, sans précision sur le statut administratif. C'est là que devrait aller ja:東京.✓ Done (--Oursana (talk) 02:36, 19 August 2014 (UTC))[reply]
Les articles comme en:Tokyo mélandent un peu #1 et #3, et je ne sais pas trop dans quel élément les mettre. Les mettre dans #3 serait sans doute plus intuitif pour les interwikis, mais moins pratique pour la transclusion des données Wikidata dans l'article.
  1. Peut-être un ou plusieurs éléments instance of (P31) Wikimedia duplicated page (Q17362920) pour Tokyo (Q11199581)=en:Tokyo Metropolis et autres articles apparaissant comme des doublons sur Wikipédia.
--Zolo (talk) 09:58, 13 August 2014 (UTC)[reply]
Merci Zolo, je suis complètement d'accord avec toi. Normalement ce n'est pas trop difficile, mais souvent il y a beaucoup de fautes. Vois-y, si tu veux. --Oursana (talk) 16:30, 14 August 2014 (UTC)[reply]
Urgent!, Vois maintenant. Peux-tu aider. Je craigne un crisis :)--Oursana (talk) 14:55, 16 August 2014 (UTC)[reply]
Désolé, je n'ai pas accès à un ordinateur aujourd'hui, j essaierai de regarder demain. --Zolo (talk) 19:43, 18 August 2014 (UTC)[reply]
Merci, j'étais en shock, on avait fusionneé Tokyo (Q7473516) sur Tokyo en général avec Tokyo (Q1490) sur la préfecture et presque annulé Tokyo (Q7473516) sur Tokyo. Les sitelinks du Tokyo (Q1490) Tokyo prèfecture avaient été mouvés à sur Tokyo (Q11199581)=en:Tokyo Metropolis. J'espère que j'ai reparé tout. A Tokyo (Q11199581)=en:Tokyo Metropolis reste seulement en. J'ai fusionné en Simple English Tokyo (Q1490) sur la préfecture et Tokyo (Q11199581)=en:Tokyo Metropolis.--Oursana (talk) 02:36, 19 August 2014 (UTC)[reply]
Merci, j'ai l'impression que tout est bon. Juste un petit point mineur : les deux sont instance of (P31): financial centre (Q1066984), ce qui fait que si on veut faire une requête "liste de places financières" on aura deux éléments à propos de Tokyo, ce qui ne sera pas très clair, mais je ne sais pas trop quoi y faire. --Zolo (talk) 05:30, 20 August 2014 (UTC)[reply]
Merci--Oursana (talk) 01:59, 21 August 2014 (UTC)[reply]

Talkback

[edit]
Hello, Zolo. You have new messages at Talk:Q17437796.
Message added 04:46, 12 August 2014 (UTC). You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

GZWDer (talk) 04:46, 12 August 2014 (UTC)[reply]

pcall

[edit]

Look at pcall: mw:Extension:Scribunto/Lua reference manual#Changed functions. --JulesWinnfield-hu (talk) 21:05, 24 August 2014 (UTC)[reply]

Thanks JulesWinnfield-hu. There is no explanation of what is changed in pcall, so I suppose they disabled the catching of some errors, like for pcall. I do not find any other way to mute the "invalid ID" error message, except by checking the the ID format is correct before sending it, so doing in the module exactly the same checks as the software does internally. That sounds a bit weird. --Zolo (talk)

Population des communes françaises vs. espagnoles

[edit]

Bonjour Zolo, pourrais-tu me dire comment s'est terminée la discussion sur l'import de la population des communes françaises stp ? Mal, j'imagine, vu qu'il n'a toujours pas eu lieu. Ce qui me surprend est que KLBot2 est en train d'effectuer celui des communes espagnoles alors que leur licence ne me semble guère différente. Ne nous sommes nous pas posé trop de questions ? (ou peut-être est-ce l'inverse...) — Ayack (talk) 10:24, 26 August 2014 (UTC)[reply]

Bonjour Ayack. Il faudrait que je recommence à m'en préocuper... La conclusion semble avoir été qu'il faurait demander l'autorisation directement à l'INSEE. J'ai entendu parler d'un formulaire type de Wikidata qui permettrait de transmettre ces demandes, mais je ne sais pas exactement ce que c'est (c'est peut être d'ailleurs juste un projet). En fait, je pense que la raison pour laquelle KLBot2 fait l'import est que le problème de licence n'a pas été abordé... En attendant, user:ValterVBot devrait importer les données sur les Etats-Unis très bientôt, ça devrait permettre de s'échauffer :]. --Zolo (talk) 11:06, 26 August 2014 (UTC)[reply]
OK, merci. C'est ce que je pensais, il vaut mieux faire les choses dans son coin sans poser trop de questions et ça passe tout seul... — Ayack (talk) 11:18, 26 August 2014 (UTC)[reply]
C'est pas faux. Le risque est quand même d'avoir à retirer les données ensuite, alors qu'on s'en sert (même si j'imagine mal l'INSEE nous faire un procès). --Zolo (talk) 11:20, 26 August 2014 (UTC)[reply]
Et encore, dans Wikidata les données seraient conformes à la license de l'INSEE, donc peu de risques qu'il s'en préoccupe. Le seul problème est la license de Wikidata... — Ayack (talk) 11:33, 26 August 2014 (UTC)[reply]

Module:Tree

[edit]

Salut, un ou deux trucs sur tes modifs au module arbre:

  • je pense que c'est mieux de créer des modules spécifiques pour les cas particuliers comme l'arbre généalogique, ça permettra de suivre et maintenir plus facilement le module Tree principal.
  • juste une note pour signaler que je travaille dessus un peu sur test.wikidata.org, notamment pour lui faire utiliser un module d'algorithmes de graphes. Ça marche pas encore totalement
  • j'ai créé Module:Properties pour éviter d'avoir à harcoder des identifiants de propriétés dans le code, j'aime pas lire des numéros magique. Il y a une table de correspondance entre un nom de propriété en anglais et un id, ça rend le code à mon avis un peu plus agréable à lire de l'utiliser. TomT0m (talk) 15:33, 29 August 2014 (UTC)[reply]
Salut TomT0m, je ne suis pas sûr de te suivre : les fonctions générant l'arbre n'ont pas de propriétés hardcodées. La fonction "familytree" se contente d'appeler la fonction principale avec des paramètres spécifiques (en fait il y a deux fonctions familytree mais c'est juste parce qu'on ne peut pas utiliser la même depuis le Wikitexte et depuis un autre module). Si tu veux les mettre dans un module séparé, ça ne me dérange pas. En fait je me demandais s'il y aurait assez de matière pour faire un sous module "répertoire" avec une liste d'arbres utiles.
Familytree pourrait sans doute être un peu nettoyée, mais vu que le formatage est très customisé, les parammètres sont forcément un peu complexes.
Je ne sais pas si utiliser des noms de plutôt que des numéros est une très bonne idée pour les fonctions Lua. Ca oblige à utliser un module supplémentaire, et c'est un peu difficile à étendre à toutes les propriétés.Un appel direct au numéro de la propriété me parait poser moins de problèmes et ajouter des petits commentaires explicatifs pourrait suffire à rendre les choses plus claires.. En revanche, je retiendrais volontiers l'idée d'utiliser des libellés, mais plus en aval, dans les modèles appelés depuis le Wikitexte.
Cela dit, ça ne concerne pas les arbres, mais j'avoue que j'ai hardcodé (start time (P580), end time (P582), point in time (P585) dans Module:Wikidata, parce que c'est le moyen le plus simple que j'ai trouvé d'avoir une option "classer par ordre chronologique", qui est vraiment souvent utile. --Zolo (talk) 16:26, 29 August 2014 (UTC)[reply]

couple Q219160

[edit]

Bonjour Zolo, peux-tu contrôler ici, si les deux chinois sitelinks vont ici ou à marriage law (Q17542970) ou married couple (Q15052790). Cordialement --Oursana (talk) 13:51, 8 September 2014 (UTC)[reply]

Bonjout Oursana, ça me parait married couple (Q15052790) (je ne comprends pas le zh-classical, mais j'imagine que c'est pareil). Cela dit fr:Couple (droit et sociologie) me parait plus proche de l'article chinois que de it:Coppia (sociologia). L'article français parle des couples de longue durée, vivant sous le même toît, et reconnus par le droit, alors que l'italien parait beaucoup plus large. --Zolo (talk) 14:14, 8 September 2014 (UTC)[reply]
Merci--Oursana (talk) 14:21, 8 September 2014 (UTC)[reply]

Sparse items

[edit]

Hi Zolo, I took a look at Q16738887: a book with P357 (P357): "Excavations at Carthage conducted by the University of Michigan". I guess it is this book. If you don't have time to add any more information please add at least a link to one library catalog. This would be a great help. --P1433-2014 (talk) 11:03, 13 September 2014 (UTC)[reply]

@P1433-2014:, I usually try to add ISBN or OCLC, but it seems that I forgot it here, sorry. --Zolo (talk) 10:41, 15 September 2014 (UTC)[reply]

Bug

[edit]

Voir Talk:Q3581036, il y a un truc de cassé dans {{Item documentation}}, il y a plus rien qui marche on dirait. Ça a l'air d'être consécutif aux changements de modules et l'introduction de l'API pour interroger les items arbitrairement, donc je te relaie vu que tu as déja un peu travaillé sur la migration :) TomT0m (talk) 09:41, 14 September 2014 (UTC)[reply]

Le problème de sérialisation ne touche pas l'accès Lua normal à Wikidata, mais seulement le rendu JSON utilisé par Module:WBHacks. J'ai changé {{Autodescription}} pour lui faire utiliser Module:Wikidata, et pour l'essentiel ça remarche. Il reste des erreurs, mais j'essaierai de voir ça ce soir ou demain.
@TomT0m: En fait, WBHacks avait été créé pour contourner l'impossibilité d'accès aux items par la voie normale, et est donc obsolète, donc il faudrait essayer de s'en débarasser complètement.--Zolo (talk) 10:56, 15 September 2014 (UTC)[reply]
Oui, c'est ce que j'entendais pas migration. Merci pour la correction. TomT0m (talk) 11:17, 15 September 2014 (UTC)[reply]
@TomT0m: J'ai corrigé une erreur sans doute introduite par moi) dans Module:Tree. Je crois que tout refonctionne maintenant. Et WBHacks est pratiquement retiré de la circulation. --Zolo (talk) 08:41, 16 September 2014 (UTC)[reply]

Au sujet des rangs

[edit]

Tu m'a glissé un mot sur les rangs avec le pont des Arts, à ce que je comprend j'ai été dans les patates quand j'ai mis le rang déprécié à l'ancien classement de la Sainte-Croix-de-Tadoussac Mission Church (Q2957267). J'aurai du le mettre au rang normal et mettre les trois autres reconnaissances au rang privilégié? --Fralambert (talk) 01:41, 17 September 2014 (UTC)[reply]

@Fralambert:. Oui les recommandations ont l'air d'être "préféré" pour les valeurs à retourner en priorité (donc il faut qu'elles soit actuelles et importantes), "normal" pour celles qui sont vraies mais soit dépassées soit secondaires (du genre dire que la Joconde est en chêne parce qu'on a ajouté un support en chêne derrière dans les années 50), et "depréciées" pour celles qui sont fausses mais qu'on garde pour mémoire, et pour éviter de les réajouter avec un autre rang. Le mot "deprécié" ne parait pas très bien choisi, mais à part ça il me semble que le système fonctionne plutôt bien. {{#Property:}} retourne les valeurs préférées quand il y en a, les normales sinon. Pour les requêtes il y aura forcément un peu plus de bricolage à faire (quand on veut une liste de présidents de la République, on veut sans doute à la fois les actuels et les passés, donc les normaux et les préférés, mais quand on veut une liste des députés, on voudra peut-être plutôt les députés actuels, donc que les "préférés"). --Zolo (talk) 07:19, 17 September 2014 (UTC)[reply]

Some page changes

[edit]

Hi Zolo! Some page changes:

(Feel free to move them again, if you think other names would be better)

Also now created:

with quite a few weirdnesses for me still to iron out! (A tracking list is at Template talk:Creator/wrapper/test, but probably more to come).


Thank you so much for re-writing Module:MakeLangSwitch for me. The changes you've made make a lot of sense, and it's good to know, most importantly, that it is now a much closer parallel to what Module:Wikidata does to generate a link in the same situation.

Something I might look at later is whether to suppress (at least by default, unless over-riden) language variants like en-gb and en-ca, when both the label and the link are identical to what they would be for the fallback. But probably not soon, as I'm hardly going to have email or net access for the next 7 days.

One other thing I'm wondering: should something like

return entity['claims']['P625'][1]['mainsnak']['datavalue']['value']['latitude']

be considered safe for returning (separately) the decimal latitude of an item? Or should it be considered fragile, and at risk of potential breakage in future?

Thanks again, all best Jheald (talk) 20:25, 19 September 2014 (UTC)[reply]

Hi Jheald. Yes, many of these "languages" do not seem tremendously useful. A bot has actually removed some of them directly in Wikidata.
I do not see why entity['claims']['P625'][1]['mainsnak']['datavalue']['value']['latitude'] would be fragile, but the best value may not always be the first, so better go through wikidata.getClaims that only returns the best ranked values. Also you may want to take the value precision into account so as to avoid displaying 50°20'00.00 N instead of 50°20 N.
Best --Zolo (talk) 06:12, 20 September 2014 (UTC)[reply]