User talk:Thierry Caro/Archives/2017/December

Active discussions

CGF stuff

prior to many more of the CGF additions we need to run a bot through the lawn bowls and weightliftingh categories at enWP to categorise and make it easier to mix n match. Probably need to do something similar for the polish cats for some others. I am to whacked to do it today, though may over the weekend.  — billinghurst sDrewth 11:32, 24 November 2017 (UTC)

@billinghurst: Yeah. It's becoming more and more complicated now. Thierry Caro (talk) 11:55, 24 November 2017 (UTC)
@billinghurst: And thank you for everything, by the way! Thierry Caro (talk) 12:55, 24 November 2017 (UTC)
and back at you for getting it organised. Evidently it is one that piques interest. :-)  — billinghurst sDrewth 22:51, 24 November 2017 (UTC)
The remains of matched are all for deleted articles from enWP. Issue is that they were transferred through to Draft: ns prior to deletion, so WD system didn't cope. I have been resurrecting the biographies bit by bit and just have up to 100 weightlifting bios to go, and I am pretty certain that they are all matched. Have done lots of petscan matches to get the matching data in place. I am clearly fixated. :-/  — billinghurst sDrewth 07:55, 30 November 2017 (UTC)
Oh, not sure what flexibility we have for matching, however, if we do have any, I know that there are still plentiful matches in the unmatched. There are names there where middle names, or middle initials, are impacting the ability to match. Also as I have add quite a few sport (P641) fields, if we are able to grab that into the matching space, it may pick up others. If you know of scope to do some more fuzzy matching, that would be useful.  — billinghurst sDrewth 08:19, 30 November 2017 (UTC)
aaaaaaaand it would great if there was a synonym search, things like Anthony (Antony, Tony); Michael (Mike); Jonathon (Jonathan, Jon). Do we need to nag MM?  — billinghurst sDrewth 08:27, 30 November 2017 (UTC)
@Billinghurst: I honestly believe that you did your best and that we can't expect too much more from the Mix'n'match tools now. We will have, I suppose, to review all the remaining lines manually. What's nice is that thanks to you we now have a much larger number of matched items here whatever. Thank you for your help again. And just in case you liked doing this, know that here you can find another huge set of IDs waiting for a contributor to match them with our items! Thierry Caro (talk) 12:40, 3 December 2017 (UTC)


Would you mind to explain your recent undos!? -- ZH8000 (talk) 17:36, 1 December 2017 (UTC)

@ZH8000: Hello. mountain range (P4552) is a subproperty of (P1647) both located on terrain feature (P706) and part of (P361) and it is set to Monte Rosa Massif (Q14668) on Dufourspitze (Q3403). Thierry Caro (talk) 17:42, 1 December 2017 (UTC)

Russian-Japan islands

Can you please undo your changes with mountain range (P4552)Lesser Kuril Ridge (Q2093721) as it is not (only) a range but also a group of islands and statements like "island mountain range (P4552) group of islands" is more strange that part of (P361).--Infovarius (talk) 20:46, 2 December 2017 (UTC)

And Greater Kuril Ridge (Q2372442). I understand using this property for mountains, but not for rivers, museums, administrative units. I think you are overshooting. --Infovarius (talk) 22:38, 2 December 2017 (UTC)
@Infovarius: I believe it's mostly the definition of Lesser Kuril Ridge (Q2093721) that needs to be reviewed. As long as it is defined as something to which mountain range (P4552) applies, this property will quite logically be used on it. Bots will come and move the data back to mountain range (P4552) whenever it's removed. So we need a better definition of Lesser Kuril Ridge (Q2093721) first. Thierry Caro (talk) 12:35, 3 December 2017 (UTC)
But it is also island group (Q1402592) so at least part of (P361) should not be removed. Infovarius (talk) 14:21, 7 December 2017 (UTC)

Rami Imam

Hi, Rami Imam isn't dead and I think you add (death place) mistakenly, so I reverted it --Ibrahim.ID (talk) 23:02, 3 December 2017 (UTC)

Link from ID in Listeria

Not sure if you got the ping. Check out the listeria template at Wikidata:WikiProject sum of all paintings/Collection/Art Institute of Chicago

{{Wikidata list
|sparql=SELECT ?item ?articurl WHERE {
  wd:P4610 wdt:P1630 ?formatterurl .
  ?item p:P195 ?collectionstatement .
  ?collectionstatement ps:P195 wd:Q239303 .
  ?item wdt:P31 wd:Q3305213 .
  OPTIONAL { ?item wdt:P4610 ?articid }.
  BIND(REPLACE(?articid, '^(\\d+)$', CONCAT("[", ?formatterurl, " $1]")) AS ?articurl)  .
LIMIT 4000

Works for other id's too, just replace P4610 (and you might want to rename ?articid and ?articurl fields). Multichill (talk) 17:37, 4 December 2017 (UTC)

Ajouter Commons à Template:External links

Salut, sur les récentes propositions d'ID de musées, j'ai cherché en vain Commons, qui est pourtant un utilisateur important de ces ID... penses-tu qu'il serait possible de l'ajouter aux liens de Template:External links ? --Hsarrazin (talk) 09:46, 6 December 2017 (UTC)

@Hsarrazin:   Done ici. Thierry Caro (talk) 15:38, 6 December 2017 (UTC)
merci ! ma syntaxe de #switch n'est pas assez sûre pour que je m'y sois risquée :)D --Hsarrazin (talk) 15:41, 6 December 2017 (UTC)

The usage of P4614 Watershed

God afternoon! I was not aware of this property untill now and would like to ask you some questions regarding the use of this property and its use for Norwegian rivers and water basins. In Norway the official authorities defines a drainage basin/Watershed as A water resource is a limited and significant amount of surface water, that is, a coastal area, lake or river, or a limited volume of groundwater. Several water bodies together form a water area Should then Watershed be a waterresourse?

For the Oise basin (Q35086304) why is that not a part of Seine basin (Q2887515) but has the Seine basin as its Watershed.

Do yuo know if the Seine basin is the same as Bassin Seine Normandie [[1]]

Best regardsPmt (talk) 18:44, 6 December 2017 (UTC)

  1. No. A water resource as defined above is something different from a watershed.
  2. Oise basin (Q35086304) is a part of the Seine basin (Q2887515) but as drainage basin (P4614) is a subproperty of (P1647) our located on terrain feature (P706) property, itself a daughter of part of (P361), there is really no need to specify the relationship twice using part of (P361) on top of drainage basin (P4614).
  3. The Seine basin (Q2887515) is smaller than the area managed by Eau Seine Normandie. It is limited to the green part of this map. They also deal with the blue parts. Thierry Caro (talk) 02:04, 7 December 2017 (UTC)

Located on terrain feature

Thanks for fixing, but a vandal destroyed your edits 15:24, 7 December 2017 (UTC)

Please note

Thanks for your recent edits on Welsh buildings; do note, however, that Wales is a nation. Llywelyn2000 (talk) 10:34, 18 December 2017 (UTC)

described at URL

Hi Thierry, I noticed you are removing some claims of described at URL (P973) for some artworks like Portret of Joseph Benoît Suvée (Q21677829) Why are you doing this? This is not correct. Alina data (talk) 10:02, 20 December 2017 (UTC)

@Alina data: Hello. Such data is stored through Groeningemuseum work PID (P2282) and BALaT object ID (P3293), as those are properties dedicated to doing this. Thierry Caro (talk) 10:07, 20 December 2017 (UTC)
@Thierry Caro: No, not really. Groeningemuseum work PID (P2282) was created for persistent URI's with /id/ in their syntax. These URI's identify the artwork itself and resolve you to a landing page. URI's with the syntax /data/ in them identify 'records about artworks' (different thing) and resolve you tot the record of the artwork published by the museum. The /data/ URI's were thus published by us as described at URL (P973) during our project. As Groeningemuseum work PID (P2282) is a 'persistent identifier for an artwork from the collection of the Groeningemuseum in Bruges, Belgium' it should only store the /id/ URI's. Groeningemuseum work PID (P2282) is temporary also storing /data/ because /id/ is not always working. But this will be changed in the future. I am following this up. Could you temporary leave the described at URL (P973) as it is? Thank you. Alina data (talk) 10:24, 20 December 2017 (UTC)
@Alina data: If the IDs are the same and then only the URL changes, you really shouldn't store them twice. Groeningemuseum work PID (P2282) has two entries as its formatter URL (P1630), one ending with /data/ and the other with /id/, meaning that everything is covered by the specific property. Thierry Caro (talk) 10:32, 20 December 2017 (UTC)
@Thierry Caro: These URL's stand for different concepts so they are different pieces of data which in my opinion both should be stored. We proposed Groeningemuseum work PID (P2282) as a property for identification of artworks. Somebody decided that /data/ is also identification of artworks and added this syntax to the property which is not really our understanding. But maybe I should start this discussion on the page of Groeningemuseum work PID (P2282) and also push the museum to finally figure there technical problems out. Sorry. Alina data (talk) 10:53, 20 December 2017 (UTC)
@Alina data: Two properties or double storage would be OK if the perimeter of each URL system is significantly different. I don't know about these ones but if the ID are the same and the perimeters pretty much the same, then everything on Wikidata urges one to remove the described at URL (P973) declarations with no afterthought, whatever the rationale was when the property was created. On the other hand, if the perimeter varies importantly, I'd create a new property instead of using described at URL (P973) whatever. Thierry Caro (talk) 10:57, 20 December 2017 (UTC)

Cairn author ID

Bonjour Thierry,

je viens de voir cette modification. Peux-tu STP m'expliquer comment il se fait que la propriété marche apparemment désormais aussi avec l'identifiant WD ??

Nomen ad hoc (talk) 12:08, 23 December 2017 (UTC).

@Nomen ad hoc: C'est parce que j'ai modifié la structure de formatter URL (P1630) sur Cairn author ID (P4369). Le nom de la personne n'était pas nécessaire. Il fallait juste un peu de texte, n'importe lequel, et c'est ce rôle de joker que joue wd ici désormais, comme dans plusieurs autres formatter URL (P1630) sur Wikidata. Thierry Caro (talk) 12:42, 23 December 2017 (UTC)
Super, tu penses qu'on peut faire pareil avec ma proposition d'ID pour les revues disponibles sur Cairn ? Nomen ad hoc (talk) 12:44, 23 December 2017 (UTC).
@Nomen ad hoc: Non, là il n'y a rien à raccourcir. En général, c'est quand un identifiant se présente comme une combinaison de lettres et d'un nombre avant ou après ces lettres que l'on peut se débarrasser des lettres. Thierry Caro (talk) 12:47, 23 December 2017 (UTC)
Entendu. Nomen ad hoc (talk) 12:48, 23 December 2017 (UTC).

Imported from is not a reference

Ahum. Multichill (talk) 16:56, 23 December 2017 (UTC)

@Multichill: It is. See property constraint (P2302)used as reference constraint (Q21528959) on imported from Wikimedia project (P143). Thierry Caro (talk) 17:00, 23 December 2017 (UTC)
Come on Thierry, you know well to only add "imported from" if you're actually importing data, not to add it to statements that someone else has added. Also see Help:Sources#Different types of sources. Multichill (talk) 17:12, 23 December 2017 (UTC)
I was adding the data too, actually. So I guess it's OK. Plus it's really a detail that does not harm anybody. Thierry Caro (talk) 17:15, 23 December 2017 (UTC)
@Multichill: Pinged. Thierry Caro (talk) 17:16, 23 December 2017 (UTC)
Doesn't the data have a source from where it come into mix-and-match? ChristianKl❫ 22:55, 23 December 2017 (UTC)
I suppose this was first scrapped from the website itself. I personally did nothing more than download the IDs from Mix'n'match and import them here. Multichill did the same at the same time without imported from Wikimedia project (P143). My run came on his imports and added the small precision there. I suppose it's OK eventually. Thierry Caro (talk) 23:30, 23 December 2017 (UTC)

I need help

Hi user:Thierry Caro, i just found you by recent changes. i need help to put this link lccn-nr97-42696/ as worldcat id for Q5431330. but i can't, seems something wrong. --Rafic.Mufid (talk) 07:15, 26 December 2017 (UTC)

@Rafic.Mufid: It's already there under Library of Congress authority ID (P244). Thierry Caro (talk) 07:26, 26 December 2017 (UTC)

Création d'une propriété


c'est encore moi. Tu es créateur de propriétés, je crois ; aussi, pourrais-tu STP mettre au point Canal-U channel ID ? La discussion a excédé une semaine, et aucune opposition ne s'est manifestée.

Nomen ad hoc (talk) 11:52, 28 December 2017 (UTC).

@Nomen ad hoc: Tu as oublié de remplir le champ topic dans ta proposition, ce qui fait qu'elle n'a été listée nulle part, ce qui fait qu'on va prolonger un peu, pour que chacun puisse s'exprimer, maintenant que c'est corrigé. Thierry Caro (talk) 19:43, 28 December 2017 (UTC)
Ok. Désolé ! Nomen ad hoc (talk) 19:51, 28 December 2017 (UTC).
Return to the user page of "Thierry Caro/Archives/2017/December".