Shortcut: WD:PC

Wikidata:Project chat

From Wikidata
(Redirected from Wikidata:PC)
Jump to: navigation, search
Wikidata project chat
Place used to discuss any and all aspects of Wikidata: the project itself, policy and proposals, individual data items, technical issues, etc.
Please take a look at the frequently asked questions to see if your question has already been answered.
Also see status updates to keep up-to-date on important things around Wikidata.
Requests for deletions and merges can be made here.

IRC channel: #wikidata connect
On this page, old discussions are archived. An overview of all archives can be found at this page's archive index. The current archive is located at 2015/05.

WikiNews links[edit]

Pirate Party Germany (Q13129) should be an item of it's own for the four wikinews articles, that are linked by main subject to this item? In this way we can only have one WikiNews-article about any subject, which is wrong anyways. Edoderoo (talk) 16:27, 17 May 2015 (UTC)

The way Wikinews articles are currently handled is a mess. Wikinews articles are about events, but are getting linked to items about partly-associated entities and then treated both as items about the associated topic, and as items about the Wikinews article. Much of the contents of Special:WhatLinksHere/Q17633526 are just wrong. In my opinion, we need to rethink how Wikinews articles are handled in general, and then do some major repair work on the existing items and links. I don't think that any of these items should link to Wikinews article (Q17633526) or use language of work (or name) (P407), really. --Yair rand (talk) 16:58, 17 May 2015 (UTC)
The way I treated them until now: a Wikinews article has it's own wikidata-entry, always. It is not linked to any of the other projects, to my idea. Then the language makes more sense, as all the listed languages do have a wikinews-article in that particular language. It now goes wrong where a Wikinews article is linked on a Wikidata entry of a person, a building, a city, etc. We indeed need to do (quite) some repair work, but I believe that 99% of the items are actually OK, but the 1% is causing quite some headache. Edoderoo (talk) 18:10, 17 May 2015 (UTC)
as an example: Jamaican Usain Bolt breaks world record 100m to 9.72 sec (Q19931054) is now taken out of Usain Bolt (Q1189) (see history for my edits). I see that the Polish language article is a news article on the same subject, but a different occasion, so I will create a new entry (Usain Bolt sets world record 200m (Q19931072)) for that as well. To my opinion, that is how we should treat Wikinews articles. Edoderoo (talk) 18:36, 17 May 2015 (UTC)
The use of language of work (or name) (P407) in Wikinews items is wrong because it is not corresponding to the general way in WD. We don't add language of work (or name) (P407) to all others items when they are linked to a specific WP (we don't add language of work (or name) (P407): english to an item linked to WP:en). ::: An item about one Wikinews article or about an WP article if linked to english Wikinews or WP:en is ALWAYS in English. This is just a redundant information which is a useless maintenance work to maintain as all redundant informations. Snipre (talk) 19:40, 17 May 2015 (UTC)
Right, that is about the smallest issue. What about the big issue, that a WikiNews article is linked on the WD-entry of it's main subject? Edoderoo (talk) 20:07, 17 May 2015 (UTC)
@Edoderoo: This is a structural problem. Acording to Wikinews structure, everything should be an item but this is not the case in WD. For example if there is a Wikinews article about the birth of a person, we should have an item about the birth of that person. But in the current structure of WD, birth data are added as statements. More we are performing some experiments with WD and more I thing Wikinews can't be included in the current WD. This is not the only issue we have with a database which is used to store data and to store links between projects. We have two different objects and the problem is we try to keep them together and something to create interactions inside the same system. Typical examples are the categories items. Snipre (talk) 21:10, 17 May 2015 (UTC)
The articles you linked to are about events. As such, they should have instance of (P31) event (Q1190554), and they should be linked to articles on other wikis about the event. If there were a Wikipedia article on "Establishment of the German Pirate Party", it could be linked to the Wikinews articles you referred to earlier. These items can have normal relations with other items. For example, the item on Pirate Party Germany (Q13129) could link to the item about its establishment using subject of (P805) as a qualifier of inception (P571). --Yair rand (talk) 20:14, 17 May 2015 (UTC)
@Yair rand: The problem is that WD doesn't store each event as an item. Just think about birth and death of people: most of data about these kind of events are stored as statements. According to Wikinews structure these should be items in order to create appropriated links. This is the problem of mixing data in a structured way with site-links which suppose that everything can be an item in order to create an unique link. Or we have to allow multiples links from the same site to an unique item, but in that case we create a phone book and not more a structured database. Snipre (talk) 21:18, 17 May 2015 (UTC)
Wikidata has plenty of items on events. If you have an item regarding a relation between two topics or an topic and a time/location/etc, and that relation is primarily stored as a statement, use subject of (P805) as a qualifier to the statement to associate the other article. For example, if there was an item "birth of X", then you would add "subject of (P805): birth of X" as a qualifier to X's date of birth (P569). --Yair rand (talk) 21:24, 17 May 2015 (UTC)
The problem is not the fact that some events are represented by items. The problem is that for some events which are defined as statements we need to create empty or data redundant items just to be able to create a link to the Wikines article, the problem is in that mixture of models (sometimes event as item, sometimes event as statement). We have to choose between two models: a all-events-as-item model or we keep the statement for describing some events and we have to find a different way to link the statement directly to the site via a sitelink. Snipre (talk) 23:25, 17 May 2015 (UTC)
Most of the connections can be linked by main subject (P921) out of the wikidata-entry, especially when the news is about persons. In case of events, like accidents, it might be that there is no related article, and there is nothing to be linked. To my idea, the good thing when using main subject (P921) is that the wikinews articles will show up when using Reasonator, without the link they will simply never show up. Edoderoo (talk) 22:09, 17 May 2015 (UTC)
Why wouldn't they show up if the items' topics were considered the article topics instead of the articles themselves, like we do with Wikipedia articles? They could even be linked with significant event (P793) in the otherwise statement-less cases. --Yair rand (talk) 22:46, 17 May 2015 (UTC)
You mean in the way it was in Usain Bolt before I took the Wikinews articles out? Because you can then only link ONE Wikinews article on every Wikipedia-article. Where many politicians have multiple Wikinews articles about them. Major events (Tour de France, Olympic Games, etc) probably have more then one item too. In database language, there is a one-to-many relation, where it is now linked as a one-to-one relationship. Edoderoo (talk) 22:53, 17 May 2015 (UTC)
No, that's not what I mean. I mean having a separate item for the event, with the Wikinews articles linked, but not treating it as an item for a Wikinews article, but rather as an item for the event that is the topic of the linked articles. There should not be a linked Wikipedia article unless Wikipedia also has an article specifically on the event. --Yair rand (talk) 23:17, 17 May 2015 (UTC)
I start to believe that we need some clear examples, on how it's wrong, and how it's correct. Preferably with perma-links, else the items might change over time. I'm afraid that we're not all interpreting the terms in the same way (which might be due to not having all English as primary language, which is as well the case for me) and clear examples might bypass that problem. Edoderoo (talk) 07:09, 18 May 2015 (UTC)
First version, with the topics combined, which is currently the state of many items. This obviously have many problems, and I think we can all agree that it shouldn't be the standard.
Second version, with an item that has the Wikinews article itself as its topic. If I have understood correctly, this is the system supported by Edoderoo. Since it's being considered as a published piece of news, it has main subject (P921) linking to the other item.
Third version, with an item that has an event as its topic, with links to Wikinews articles that cover that particular event. As an ordinary event, it fits into the data model in the ordinary manner, and is linked to from the other item as shown. Should Wikipedia also have in article on the event (which is not true in the example given), then it would also have pages linked to that event's item. This is the system that I support.
--Yair rand (talk) 15:58, 18 May 2015 (UTC)
Given the case there are one or multiple wikinews article about an event, lets say a sports final, it makes no sense to assign language of work (or name) (P407) to the item. Everyone, or even a stupid bot can detect the language(s) of the article(s) based on the wikilink(s). So if there is a link on Jamaican Usain Bolt breaks world record 100m to 9.72 sec (Q19931054) to de.wikinews we can safely assume that the language is German. We also do not assign language of work (or name) (P407) with value German (Q188) to Usain Bolt (Q1189) because there is a German article linked to the item.--Giftzwerg 88 (talk) 21:57, 18 May 2015 (UTC)
I am actually in favour of the second hypotesis: it saves the links between WN versions, but we also can define what the article is about, and the item itself can be used as a qualifier or even as a source. --Sannita - not just another sysop 16:33, 19 May 2015 (UTC)
@Sannita: Can you give an example of when such an item would be useful as a qualifier? --Yair rand (talk) 20:11, 19 May 2015 (UTC)
@Yair rand: I'm sure the idea of the qualifier made sense when I wrote that, but I literally can't remember what I thought. So for the time being, you might want to scratch the qualifier part and keep just the source part. Sannita - not just another sysop 15:13, 21 May 2015 (UTC)
So, regardless of whether these items are to be treated as news pieces or events, are we in agreement that they shouldn't have language of work (or name) (P407) statements? If not, does anyone have arguments in favor of them? --Yair rand (talk) 21:29, 26 May 2015 (UTC)

Open for SI-unit properties[edit]

Progress on implementing numeric properties with units - such as length/distance, area, time and speed - is slow and only rarely appears on the weekly update. A number of properties are on hold due to this.

Would it be possible to allow properties, where an SI-unit exists. The description of the unit could include the SI-unit in use. Examples:

  • Height - the vertical length stated in m.
  • Area - stated in m2.
  • Speed - stated in m/s.

When in the future the units are here, a bot would be able to append them by property.

Off course this will annoy authors used to feet and °C to have to use m and K - and prefixed units (eg. km) are ruled out as well. But it's just for now.

Lua-modules will still be able to display the numbers as the language of the wiki in question might dictate.

Poul G (talk) 17:01, 18 May 2015 (UTC)

I mean, what I think matters is that if people are willing to use a temporary solution, they must be ready and willing to do a mass conversion once units are implemented.--Jasper Deng (talk) 00:26, 19 May 2015 (UTC)
It does sound tempting given that we could easily loose another year or two waiting the for the first implementation of units. Conversion from the temporary solution should be quite easy.
The main issue I see is that the quantity datatype still needs improvement, see: Wikidata:Project_chat/Archive/2015/05#population.2B-0. --- Jura 06:00, 19 May 2015 (UTC)
We are not years but only a few months away from having unit support. In the end it is an editorial decision but please do keep in mind that anyone who uses this data will have to make significant adaptions again later. --Lydia Pintscher (WMDE) (talk) 10:05, 19 May 2015 (UTC)
+1 with Lydia Pintscher (WMDE). If you take the time to follow the work progress of the development team you can have a good estimation of what will happen in the next 2-4 months. Better waiting than doing partial job which will be a mess to handle later. We have already enough things to modify because initial data import or structure was not correct. Instead of forcing to be able to add some hundreds of values better prepare data in an external file and be ready to import all in once with a bot when the datatype is ready. Snipre (talk) 11:37, 19 May 2015 (UTC)
I did check the developers ticket-list, as the number-property rarely surfaces in the weekly reports. And found a few with low activity and relatively few followers. Where should I have looked to see "few months" or "2-4 months"? Poul G (talk) 21:51, 19 May 2015 (UTC)
@Poul G: See here for the general problem of the datatype with unit. Currently the problem is about how the user can select the unit (Unit Selector widget) treated by this bug. A discussion is ongoing about this last problem and you can follow it using this bug which in the current sprint of the development work (see that). Snipre (talk) 22:41, 23 May 2015 (UTC)
@Lydia Pintscher (WMDE): What would the list of predefined supported units be? ("The available units are from a list of predefined supported units." [1]) I guess the base units, the derived units, and their prefixes must be supported, but what about composite units and non-SI units accepted for use with SI? Which other systems will be supported, and to what degree will values be expressed in those systems? The question is really whether the unit system implemented will attempt to avoid data loss, or if the easier solution is chosen. Seems like tracking datum is dropped, can you confirm that? Jeblad (talk) 16:49, 24 May 2015 (UTC)
My hunch is that everything will be stored in SI units with other units converted to on demand. This could prove problematic, however, if the non-SI unit is the source unit and is specified exactly, since unit conversion would then cause loss of precision.--Jasper Deng (talk) 08:04, 27 May 2015 (UTC)
That is a very good point. Sooner or later we also need units for non-convertible units. I have read some early 20th century article recently, and they tells me that this and that was X feet wide and Y feet high. Today, I do not know the length of those feets. -- Innocent bystander (talk) 10:48, 27 May 2015 (UTC)
Yes we'll be supporting convertible and non-convertible units. As it currently looks you'll be able to say something is "12 ducks" ;-) --Lydia Pintscher (WMDE) (talk) 10:51, 27 May 2015 (UTC)

Not sure how to add something "safely"[edit]

I didn't have much time to look through the help pages, and I couldn't quickly see how best to do this; could someone else do it and leave a tutorial (or link to one) on how best to do it? I was afraid I'd break the metadata by introducing something with wrong formatting or wrong content if I just clicked the "Add reference".

Hilbert space, Q190056, can be linked to OCLC ARN 2072176. Could you please add it? Nyttend (talk) 17:32, 18 May 2015 (UTC)

@Nyttend: I took a shot at it but apparently fail at such things since I could not get a link to show up that directed me to the appropriate location. Do you have a URL that would be associated with that number, where$1 is the format of the url and $1 is the ID? It should be numerical only. You can add it yourself to Q190056 in similar fashion to what I just failed to do... (don't forget, this is a wiki, so trying things out yourself is okay).

Otherwise, we may not have the property to make the claim you are trying to make. --Izno (talk) 19:45, 18 May 2015 (UTC)

No, I don't have a URL, unfortunately. When you're using OCLC Connexion Client and browse the LCSH authority file for "Hilbert space", it shows you a subject header for "Hilbert space", and the OCLC record number for this SH is 2072176. It's not the title of a work, so it doesn't have an OCLC number per se. This is precisely why I was hesitant — I didn't even know if we were attempting to link LCSH entries with other data, let alone how to do it if we should. Nyttend (talk) 22:14, 18 May 2015 (UTC)
Two questions: Is the ID you're providing a unique, permanent ID, and, do you know if it can be derived from some other ID (OCLC or other closely related ID)? If yes and no respectively, it's probably a good candidate for WD:Property proposal/Authority control. --Izno (talk) 15:45, 19 May 2015 (UTC)
WD:WikiProject Authority control might be able to help too. --Izno (talk) 16:01, 19 May 2015 (UTC)
Izno, I expect that it's permanent, because the only way to assign this LCSH a new number would be to delete the authority file and create a new one, and that would mess up every single linked OCLC record of every type, whether ordinary bibliographic entries or other things. However, I was proposing using the OCLC number just because it was conveniently available for the LCSH. Is it possible to link LC subject headings to other kinds of records for the same topic? Q190056 currently links to an NDL entry that looks like it might be comparable to the LC topical subject heading that's listed in the print edition of LCSH and the free online edition of LCSH. Is my request now clear? Nyttend (talk) 03:18, 20 May 2015 (UTC)
Your request was cleaer to begin with--I just don't have the domain knowledge. --Izno (talk) 03:25, 20 May 2015 (UTC)
@Multichill, Kolja21: Maybe one of you can help. --Izno (talk) 03:29, 20 May 2015 (UTC)
Okay, thanks. I've only been working in cataloging since January (I didn't have the chance to take a cataloging class in library school; they don't even offer it anymore!), so I'm not as confident on this kind of thing as I wish I were. Not to mention my near-total unfamiliarity with how Wikidata typically operates. Nyttend (talk) 03:32, 20 May 2015 (UTC)
The LCSH can be looked up here. Hilbert space (Q190056): LCAuth identifier (P244) = sh85060803. P244 can be used for Library of Congress Name Authority File (Q18912790) and Library of Congress Subject Headings (Q1823134) (more infos by Gymel). --Kolja21 (talk) 04:17, 20 May 2015 (UTC)
Huh? At Property talk:P244#LCSH I merely stated my observation that this property apparently is also used for LCSH numbers, contrary to the definition at the time of the proposal. Changing the description of P244 might be an easy resolution, however at en:Template_talk:Authority control#At the moment, it is used almost exclusively in biographical articles there are complaints that this extended usage breaks the Authority Control template at en:WP: They also use the P244 value to provide links to OCLC WorldCat and this appears to work for the "n"-valued ids of the LCNAF entries but not for the "s"-valued of LCSH. AFAIK OCLC has developed its own system of FAST headings (Faceted Application of Subject Terminology, cf. en:Faceted Application of Subject Terminology but for the Hilbert Space example (LCCN sh 85060803) the identifier would be aka fst956785 (but this does not provide links to the resources related to that).
As for User:Nyttend's example "ARN 2072176": Keyword search for "2072176" yields mostly titles related to Hilbert Spaces in WorldCat, and searches for something with 060803 or 85060803 from the LCCN or 956785 from FAST result in nothing helpful. Also for subject headings there does not exist an interface like for persons, organizations &c. So I tend to back the claim that "ARN 2072176" is or corresponds to some internal OCLC number for their copy of LCSH for the benefit of their members but there seems no way to research, verify or even utilize this from the outside, i.e. with the means provided by -- Gymel (talk) 00:14, 21 May 2015 (UTC)
This is authority record number 2072176, i.e. record #2072176 in Connexion contains the metadata that allows the LCSH "Hilbert space" to function in Connexion Client and (perhaps) to connect to everything else in OCLC; I don't know if the number is particularly relevant outside the internal Connexion servers. We'd have to find someone with access to Connexion Browser to know whether it's just something with this one client program or something broader, not to mention talking with someone with OCLC to learn if there's a way to match numbers and LCSH with freely accessible resources. Nyttend (talk) 13:16, 25 May 2015 (UTC)

1980's-90's exercise shows on PBS or KVCR[edit]

Does anyone remember the 1980's-90's exercise shows on PBS or KVCR if so, do you remember the names?

Articles mixed with disambiguation pages or categories[edit]

Looking at Wikidata:Database reports/Constraint violations there are a lot of cases where disambiguation pages or categories are merged with articles. Often the property instance of (P31) for disambiguation page or category then gets deleted and everything looks fine. (But of cause it isn't: The different items are still mixed and the descriptions are mixed as well.) Why not block the merging for articles with disambiguation pages and categories? --Kolja21 (talk) 21:09, 18 May 2015 (UTC)

More usually it's the case that a wiki is causing these changes. Two main ways this happens:
  • A wiki takes a disambiguation page and changes it to an actual topic page.
  • A wiki takes a topic page and changes it to a disambiguation page.
Some wikis do this I imagine without moving the pages in question, which Wikidata would otherwise catch. From what I recall, Wikidata probably doesn't catch all of the page moves because there's some lag on the Wikibase client script running, so that might be a 3rd way we're not catching it.

Otherwise, the 4th way is simply that they're extant problems from prior to using Wikidata. I would imagine we've caught most of those by now... In other words, there's nothing we can really do except let the constraints catch the offending items. --Izno (talk) 21:14, 18 May 2015 (UTC)

Sometimes it is really an error where there are disambig pages mixed with some real articles. In some cases the article is a mix between a disambig and a stub. I´ve also found some of these.--Giftzwerg 88 (talk) 22:06, 18 May 2015 (UTC)
Some may be leftovers before everyone has a SUL. If somebody without a SUL moved a page and created a disambiguation page on the former page, the sitelinks would stay the same and would be incorrect. You should compare the maintance category with the statements on Wikidata. Sjoerd de Bruin (talk) 06:41, 19 May 2015 (UTC)
I see an attitude among some Wikipedians that "a not so good interwiki is better than no interwiki at all". That attitude in combination with the attitude that "interwiki should always be done through Wikidata nowdays", makes this a work of Sisyphus (Q102561). -- Innocent bystander (talk) 04:21, 27 May 2015 (UTC)
There are articles which are categorised as Disambiguation pages which correspond to real world topics which need wikidata items. A lot of the 'Surname' pages are this. In some languages the article just lists people with that surname, in other languages the article has an introductory sentence about the surname before the list. In my opinion these articles should be linked to a wikidata item which has statements 'instance of:wikimedia disambiguation page' and 'instance of:surname'. These items are then used as the targets of 'surname' statements. The alternative is to create separate items for the disambiguation articles and for the surname, where the item for the surname has no articles linked to it. this means we have twice the number of items to maintain and we have statements linking to items with no sitelinks. Extra maintenance and less usefulness. Filceolaire (talk) 06:36, 22 May 2015 (UTC)
This is how I've done it and seen it done (two items). It doesn't make sense to me to have an item that is a disambiguation page and a surname when it's easy for the disambiguation item to link to the surname item. Plus, there are some names where the surname is actually a separate article from the disambiguation (list) page. The two-item method allows for consistency and correct separation of data. Hazmat2 (talk) 02:34, 27 May 2015 (UTC)

Using Wikidata on small wikis[edit]

So, I was thinking that we could use start using Wikidata (besides a few isolated uses) in small wikis (with {{#property:P#}}), for things starting with something like the population and head of government/state of countries. For example, by making this edit (except with wikilinks around it). With help, we could make sure all Wikidata items about countries could be manually verified to have the correct and up to date population and head of state/goverment properties, with the correct rank of course. Then we could make that edit to all of the country articles on scowiki (and other small wikis where someone involved with this project would also be an active editor at). We also have to make sure that the items on the politicians have labels, which might make population listings easier, but the problem with the populations is that you couldn't list a date or reference on the wiki. Thoughts or volunteers? --AmaryllisGardener talk 01:58, 19 May 2015 (UTC)

I'm not sure of the schedule, but aren't we only a couple of months away from having arbitrary access on the rest of the Wikipedias? If I understand correctly, after that's available even automatically showing sources and dates would be possible. I recommend waiting and then making a bunch of changes to the infobox templates so that the pages don't need to be edited manually. --Yair rand (talk) 02:47, 19 May 2015 (UTC)
Just be careful - a few years ago ocwiki tried this but in a horrible implementation (bot-created stubs populated with magic infoboxes that broke). [2] is the result. --Rschen7754 03:01, 19 May 2015 (UTC)
@Rschen7754: That does look like something we'd have to be careful not to do, but it looks like they tried to start out with too much. @Yair rand: Well, it can wait a while. --AmaryllisGardener talk 03:09, 19 May 2015 (UTC)
I've also followed the oc experiment. Like Rschen7754 said it's a total mess with empty templates (oc:Modèl:Infobox) or templates that have been changed so often that it's hard to understand how they are linked with each other (oc:Modèl:Debuta Infobox V2). It would be good to start an initiative to clean up the mess and then we see what is the best way Wikidata can help (or destruct) small Wikis. --Kolja21 (talk) 13:00, 19 May 2015 (UTC)
BTW: Is there an admin that can help to delete the pages marked for deletion? oc:Categoria:Wikipèdia:Supression has more than 800 entries. --Kolja21 (talk) 13:15, 19 May 2015 (UTC)
Dropped a request at m:SRM, since I suspect that it may be too much for the 2 active admins. --Rschen7754 13:33, 19 May 2015 (UTC)
I am in pronciple positive but we also need to think about how we guarantee that Wikidata items used across several hundred projects are vandalism-free. I have a dozen of country entries on my watchlist, and they get vandalized on a regular basis.--Ymblanter (talk) 15:53, 19 May 2015 (UTC)
Yeah, vandalism's a problem. I was thinking that those involved in this project could watch a dozen or two country items. That way, we could make sure that any vandalism is reverted. BTW, for heads of state/government, to list that using Wikidata is very simple, and doesn't require dates or refs, just the name. --AmaryllisGardener talk 16:53, 19 May 2015 (UTC)
RE: "for heads of state/government ... is very simple" Maybe it is simple, but to be useful, you sometimes need the title of the person(s). Andorra have two heads of states, with two different titles AFIK. And in the Commonwelth there is both a Governor and a Queen. And the city of Stockholm have 12 mayors, each with a very specific title. -- Innocent bystander (talk) 10:10, 20 May 2015 (UTC)
Isn't this pretty much the same as I did two years ago on ru-wiki? See here for the examples. Same story on es-wiki on tennis related stuff. Edoderoo (talk) 12:16, 21 May 2015 (UTC)
@Edoderoo: Yes. Did your thing work? --AmaryllisGardener talk 17:25, 21 May 2015 (UTC)
Sure, on quite some wiki's that was implemented in the corresponding tennis infobox. Some Wiki's reverted due to several known or unknown reasons. The German wiki because the wikidata entry uses a dash instead of a semi-colon in the statistic numbers, the Dutch because they don't want to use Wikidata at all, the others I don't remember or just don't know. In the mean time those tennis statistics seem to be pretty well updated accross the communities, and someone even tried to get allowance to update part of it by an automated bot, but this gave issues with database rights. Still, for me it's a good show case of how Wikidata can be used to keep often changed data up to date on several wiki's. Some of the smaller wiki's can keep their tennis statistics pretty much updated without extra (or any) effort from their own community, as they benefit from the changes done by the larger foreign communities. Edoderoo (talk) 18:59, 21 May 2015 (UTC)
I might try changing the tennis infoboxes on scowiki, and if anyone complains (which is very unlikely, the whole community's me and two other admins) I can revert. I'll switch the country article's infoboxes individually (as in article by article like this, not changing all articles at the same time by changing the infobox) to insure that it's good. I'll hold off with changing the country infobox for a while. Thanks for the input everyone! :) --AmaryllisGardener talk 19:15, 21 May 2015 (UTC)

Wikidata, cycling and list of participants[edit]

Bonjour à tous. Je travaille sur la version francophone de Wikipédia, mais de manière plus importante à l'illustration du cyclisme (exemple). Entre deux reportages sur le terrain, je prépare la transition vers Wikidata, un projet qui avance petit à petit, mais qui progresse.
Aujourd'hui, ma question va porter sur les équipes et les coureurs participant à une course cycliste, au travers de l'article Grand Prix de Denain 2015 comme exemple. Comment faire (simple) avec Wikidata pour entrer une liste d'équipes, tout en sachant qu'il faut ensuite aller rechercher leur code UCI en trois lettre et leur pays de provenance, et qu'elles ont différentes places dans le tableau selon leur catégorie (UCI WorldTeams, Équipes continentales professionnelles, Équipes continentales...) ? Selon moi, un élément participating teams pourrait être créé. Mais ces mêmes équipes sont reprises dans la liste des partants, où je dois associer à un coureur son numéro de dossart et sa place, voire éventuellement un abandon au cours d'une étape (lorsqu'il s'agit d'une course à étapes comme le Tour de France ; Liste des coureurs du Tour de France 2014). Il faut également mentionner les directeurs sportifs, voire les maillots spécifiques que peuvent revêtir les coureurs. Tout ceci est encore assez compliqué pour moi. je souhaiterais quelque chose de simple à remplir, ce qui permettrait à d'autres contributeurs de générer les listes des partants, et d'effectuer sans trop de mal de nombreuses traductions. Cordialement, JGHJ.

[Google Translate] Hello everyone. I work on the French version of Wikipedia, but more importantly the cycling illustration (example). Between two stories from the field, I prepare the transition to Wikidata, a project advancing gradually, but progress.
Now my question is about the teams and riders participating in a bicycle race, through article Grand Prix de Denain 2015 as an example. How (simple) with Wikidata to enter a list of teams while knowing that they must then go seek their three letter code UCI and their country of origin, and have different places in the table according to their category (UCI WorldTeams, Professional Continental Teams, Continental Teams ...)? I think something participante teams could be created. But these same teams are included in the list of starters, where I have to associate with a runner dossart its number and its place, or possibly a discontinuation during a step (in the case of a stage race like the Tour de France; Liste des coureurs du Tour de France 2014). Mention must also be the team managers or specific jerseys that can take the riders. All this is further complicated enough for me. I would like something simple to fill, which would allow other contributors to generate the lists of participants, and perform without too much difficulty many translations. Sincerely, Jérémy-Günther-Heinz Jähnick (talk) 12:35, 19 May 2015 (UTC)

Teams shouldn't be used in definition of the participants list of a competition but only individual participants. Teams can be deduce from the "CV" of each runner defined in his item. Mainly because teams have often some substitutes which don't take part to the competition. If I know that a runner took part to a competition and at the same time I know he is part of a team at that moment of his career, I can deduce that the team took part to that competition. Snipre (talk) 14:39, 19 May 2015 (UTC)
Oui, je le savais déjà qu'il allait être possible de déduire les équipes à partir des coureurs (tout comme leur nationalité), mais j'ai besoin que Wikidata fasse la différence suivant les compétitions qui alignent 6, 7, 8, 9 voire 10 coureurs sur une course pour que le tableau tel que présenté dans mon exemple soit correctement rempli, tout en sachant qu'il arrive qu'il y ait des trous dans les effectifs présentés. Et enfin, pour aller plus loin, je souhaite que le nom de l'équipe pointe vers sa saison 2015, et qu'il y a ce problème récurrent que les équipes changent régulièrement de nom au fil des saisons. Bref, c'est quelque chose de plus complexe qu'il n'y paraît, d'autant plus que je souhaiterais que tout soit ensuite uniformiser sur les différentes Wikipédia puisque je souhaite pouvoir potentiellement créer des articles de courses dans une vingtaine de langues. Je vais te contacter pour d'autres questions et des points précis. Jérémy-Günther-Heinz Jähnick (talk) 15:51, 19 May 2015 (UTC)
Pas compris ton problème d'équipe avec 6, 7, 8, 9 voire 10 coureurs: c'est la combinaison entre le fait que qu'il existe un lien entre tous les coureurs qui ont participé à une course et le lien entre une équipe et un coureur à un moment précis qui te permet de construire l'équipe qui a participé à une course. Ensuite, le problème de nome est simple: il suffit d'utiliser la propriété "nom officiel" qui doit être définit pour un intervale de temps donné et récupérer le nom correspondant à la date de la compétition. Vu la complexité du sujet, il faudra envisager de faire l'extraction de données hors ligne et de sauver les données en dur dans les pages d'articles et non par avoir un système de connexion à WD à chaque ouverture de page. Snipre (talk) 17:35, 19 May 2015 (UTC)
Conflit d'édition @Jérémy-Günther-Heinz Jähnick: Utilises des qualificateurs (je bosse sur les courses de chiens de traîneau, et j'ai la même question compelxe de gestion des participants à gérer). Moi par exemple j'utilise participant (P710) pour les individus et en qualificateur de P710 sport number (P1618), ranking (P1352), point in time (P585) (mais pas trop utilisable pour le cyclisme où tout le monde arrive le même jour). Pour les abandons j'utilise significant event (P793) avec scratched (Q18595374), withdrawal (Q18609099) ou disqualification (Q18595902) (les 3 possibilités dans ce sport) et j'ajoute alors location (P276) avec le checkpoint (possible d'indiquer l'étape d'abandon à la place puisque je crois qu'il y a un item par étape sur WD pour le Tour). Tu peux sans doute ajouter en qualificateur member of (P463) avec l'item de l'équipe… Concernant le nombre de membres dans l'équipe, je n'ai pas actuellement de solution, vu que la situation n'est pas non plus rélgée pour les chiens de traîneau. Voir 2015 Iditarod (Q19455277) pour un exemple et Wikidata:WikiProject Sled dog racing/Todo pour ma méthode. En espérant avoir aidé… --Harmonia Amanda (talk) 17:46, 19 May 2015 (UTC)
@Harmonia Amanda: merci pour ta réponse, je vois que tu as le même problème que moi concernant la demande de création de propriété où des contributeurs cherchent à faire des économies et à en créer le moins possible au lieu d'écouter ce que demande le proposant, tout en sachant que chaque sport a bel et bien ses spécificités. Grâce à toi, je sais maintenant comment indiquer les abandons et autres.
@Snipre: : ça reste encore trop compliqué pour des contributeurs lambda. Il me faudrait la possibilité d'avoir des qualificatif de qualificatifs.
participant (P710) => Topsport Vlaanderen-Baloise 2015 (Q18746658), avec en qualificatif le listage de six à dix coureurs, et ces coureurs là se verraient qualifier d'un numéro de dossard, et au fil de la course d'abandons, puis de places (Harmonia Amanda : toutes les courses par étapes auront un élément par étape, même les moins importantes, vu qu'ils gèrent également autre chose). Je pense même qu'en qualificatif des équipes on devrait avoir d'abord le numéro de dossart, puis le coureur, et enfin sa place ou son abandon, parce qu'il arrive parfois comme à ma course de dimanche que l'équipe ne vienne pas au dernier moment, et que les dossarts lui aient déjà été réservés. À mon sans, il ne faut pas se contenter d'un système unique pour tous les sports, mais bien tenir compte des spécificités. Jérémy-Günther-Heinz Jähnick (talk) 07:35, 20 May 2015 (UTC)
=> autre équipe, et ainsi de suite.
Ma solution est beaucoup plus simple à gérer pour des contributeurs comme celui du projet cyclisme et reste assez logique et instinctive. Avec ce système, on n'est pas embêté avec le nom des équipes, puisque ça revient à suivre la solution actuelle où le lien renvoit vers la saison 2015 (les articles de saison sont les articles principaux, ceux des équipes sont des articles de fond, et non plus l'inverse). Par ce système, en plus de générer le tableau de la liste des participants, on peut également extraire les données de l'autre tableau listant les équipes participantes, ce qui fait d'une pierre deux coups.
@Jérémy-Günther-Heinz Jähnick: On ne cherche pas à faire des économies de propriétés, on veut simplement s'assurer que l'on une structure aussi générale que possible pour faciliter l'utilisation des données notamment les recherches: si pour chaque sport ou theme il faut appliquer une recherché en function d'une structure différente, on ne pourra jamais fournir d'outils généraux pour l'utilisation sur WP. Imagine que pour chaque type de liste tu doives créer une requête particulière pour cause de stockage des données selon un format spécifique et tu verras que tu auras du mal à trouver des personnes pour te donner un coup de main sur le travail de maintenance. Il faut donc penser global, mais global sportif: trouver les différents cas pour gérer les résultats sportifs (pas en function du sport, mais du type de données pour représenter tous les détails pertinents du classement final) et ensuite demander les propriétés adéquates. Snipre (talk) 16:24, 22 May 2015 (UTC)
@Harmonia Amanda: Ta manière de gérer les données est intéressante: mais au lieu d'utiliser point in time (P585) qui n'est pas comprehensible et surtout inutile pour les performances qui ne sont pas des dates, mieux vaudrait créer une nouvelle propriété "performance" qui accepte des valeurs avec unite et qui puisse être utilisée pour tous les sports: durée de la course pour le cyclisme et les course de chiens de traineau (je ne pense pas que les résultat sont donnés en jour), de même que pour toutes les courses de l'athlétisme, mais aussi longueur pour le saut en longueur, le lancer du javelot ou du poids,... significant event (P793) pour les abandons est un choix un peu difficile à faire passer, mieux vaudrait une propriété spécifique et voir si on peut la généraliser à d'autres sports pour favoriser son utilisation. Et pour l'équipe, je ne le noterait pas dans le classement, pour éviter les doublons avec les données indiquées sous l'élément du sportif, il faut trouver l'endroit le plus approprié entre le classement des courses ou les éléments des sportifs voire les éléments des equips elles-memes. Le plus important est de bien définir ce paramètre, car sinon chacun va créer son système et on ne pourra jamais avoir un moyen unique de récupérer cette information (il faudra 3 codes différents en fonction de l'endroit où se trouve la donnée équipe). Snipre (talk) 16:37, 22 May 2015 (UTC)

Property for "commander"/"commanded by"/"command of"[edit]

Need a property to link Detlev Krankenhagen (Q4237593) to German submarine U-549 (Q563394). Anyone know of an existing one? --Izno (talk) 15:58, 19 May 2015 (UTC)

Izno: You could use commander of (P598). — Ayack (talk) 16:42, 20 May 2015 (UTC)
Thanks. --Izno (talk) 19:12, 20 May 2015 (UTC)

Merge problems[edit]

Can someone merge en:Category:Federico Fellini (Q16812095) with de:Kategorie:Federico Fellini (Q9156497) ? 22:56, 19 May 2015 (UTC)

→ ← Merged --Pasleim (talk) 22:58, 19 May 2015 (UTC)

of (P642) miga[edit]

this diff just popped up a claim

< band (Q215380) (View with Reasonator) > subclass of (P279) miga < group of humans (Q16334295) (View with Reasonator) >
        of (P642) miga < musician (Q639669) (View with Reasonator) >


This is a big problem to me, as I don't know what this means and the of qualifier is highly unspecified. We should either precise the scope and the intend uses of it or find a better solution. (@Emw, X meta, zolo: Pinging probably interested wikidatans.

For example, I'd write this as

< band (Q215380) (View with Reasonator) > composed of search < musician (Q639669) (View with Reasonator) >

, which is a solution with less statements and a more precise meaning. Clearly a music band has musician members, for sure. TomT0m (talk) 09:50, 20 May 2015 (UTC)

I agree that "of" does not seem to be a very clear property in general, and that it should not be applied here in particular. However,

musician is classified as a profession, not as a subclass of humans, and don't think a group of humans should be Composed of search musicians. I would rather use occupation (P106) either as a standalone statement or as a qualifier "instance of group of humans". --Zolo (talk) 09:59, 20 May 2015 (UTC)

@Zolo: By definition, profession (Q28640) (View with Reasonator) is an activity that someone does for money. So it's not a really good solution to say every music band earns money for it, it's pretty false in the general case. I prefer the more less specific field of work search property. Qualifying instance of seems pretty weird and not really well specified to me. Formally maybe it could be used to precise some parameters of a metaclass, like template (Q1411845) (View with Reasonator) ? Like a group of things could be used as a metaclass (Q19478619) (View with Reasonator) in Protégé (Q2066865) (View with Reasonator) (cf. la section sur Protégé dans l'article Wikipédia), and the of qualifier used to instanciate the things parameter ? I'll ping maybe @Denny, Markus Krötzsch: on this :) TomT0m (talk) 10:45, 20 May 2015 (UTC)
@TomT0m: The French label of occupation (P106) is inconsistent with other languages. The subject item of occupation (P106) is occupation (Q13516667) and not profession (Q28640) so we may have to change the French label from profession to occupation --Pasleim (talk) 12:12, 20 May 2015 (UTC)
@Pasleim: Yet another use of of : . I'd prefer an instance of social sciences concept, but we need a clarification on concept anyway. Help:Classification could help, but I'm not sure how. Clearly when I make a hat, it's an event who can be classified in the hatmaking class of events. That someone often makes hat is possible using the occupation property for sure to link the person to those kind of classes. I think then that of course
< hatmaking > instance of (P31) miga < occupation >
makes sense. I'm not sure how the concept concept fits here. I would be happy with a metaclass social science class with a kind of statement
< social science class > classes used by search < social sciences >
< social science class > subclass of (P279) miga < social science class >
. TomT0m (talk) 12:55, 20 May 2015 (UTC)
I forgot
< occupation > subclass of (P279) miga < social science class >
. This would make occupations a subset of all the classes used by social sciences. TomT0m (talk) 12:58, 20 May 2015 (UTC)

I think the example looks rather understandable to me. I wonder if it works in other languages, too. That would be for me the much more important measure whether it makes sense or not. Another important measure are the use cases for such a statement in the other Wikimedia projects. That's what we should be measuring against. Considerations about Metaclasses, Protege, Reasoning and Class subsumption are for me secondary concerns, to be honest. --Denny (talk) 16:28, 20 May 2015 (UTC)

@Denny: That it make sense in natural language is indeed a nice feature, but it's like it is more a gut oriented usecase than a structured one. This implies the problems with several ways to say one thing, language interpretation dependancies, misunderstanding and so on, cultural differences problems, ... . As far as I'm concerned, this defeats a lot of the purposes of structuring datas and mimics a lot of problems with natural languages. The problem here is mainly that we have a multi purpose qualifier with no idea on how it is used by consumer, how it can be used to model things and a really poor documentation on the talkpage. As far as I can tell, it usually used to say something similar that the subclass statements says, I do not always see the added value. TomT0m (talk) 16:41, 20 May 2015 (UTC)
@TomT0m:, as I said, if it works in the other languages too then I assume it is not a natural language thing. Natural languages have so few universals that I would be very much surprised if we stumbled upon one here. So, if it works not only in English and German and related languages, but also in Chinese, Turkic, Arabic, etc., then I am inclined to say that the property has some reasonable conceptual interpretation. --Denny (talk) 16:58, 20 May 2015 (UTC)
P.S.: Actually, thinking about my limited understanding of Turkic languages, I would be rather surprised if it worked there, but I would like to want a more advanced speaker to weigh in. --Denny (talk) 17:00, 20 May 2015 (UTC)
I'm actually trying to find a good interpretation :). Not convinced yet there is some (that is not redundant), for example in the two examples I provided it seems that there is a lot of work to do to know, for example if the qualifier makes the members of the bands musicians, if the group has musician as an occupation ... whether in the second it seems pretty clear that a concept has a domain of relevance. In one case it's used on an instance of statement, in the second it's used on a subclass of statement. Clearly instance of and subclass of have a lot of literature written on them ... but with this qualifier it is a whole new world. TomT0m (talk) 17:46, 20 May 2015 (UTC)

Full names and non-country nationalities[edit]

I've been participating off-and-on for a while, and there are two things that I'm not too sure about when I edit items about people and they keep coming up.

First, what do we do with full names? Many people go by shorter variations of their official birth name, including both people who sometimes include their initials or middle names and individuals with multiple last names (as in Spanish and Portuguese naming customs, where usually only one is needed for day-to-day life). Do they all get put in aliases? This seems like it would not scale very well, considering most languages would have the same value but we have to add them separately for each. Part of me has always wanted to use birth name (P1477), but that's typically use for name changes, not for people who just have a longer "official" name. Do we have an existing procedure for this?

Second, how do we handle non-country nationalities, like Galician, Catalan, Scottish, English, etc.? This goes for both in the item description and statements. I typically use the more specific nationality as an item descriptor (e.g., Scottish politician, Catalan businessman), but then put the actual country in the statement country of citizenship (P27). I don't see any places where we could add "Catalunya" or "Scotland" in the actual statements list for that item, but it feels weird to leave it out, especially when a few Wikipedians actually put that information in the "nationality" spot of infoboxes.

How do other users typically handle these? Cbrown1023 talk 18:18, 20 May 2015 (UTC)

"Nationality" is a very ambiguous concept. To add where a person was born, use place of birth (P19). To add their ethnicity, use ethnic group (P172). To add where they have lived, use residence (P551), preferably with start time (P580)/end time (P582). To add what country or region they've been a politician of, add it to the item about their positions. Is there any other data that determines what nationality a person is? --Yair rand (talk) 20:40, 20 May 2015 (UTC)
I'm mostly referring to autonomous regions like Catalonia (Q5705) and Scotland (Q22) which are technically part of a larger country, but have some degree of autonomy and regional identification. (Those don't really seem the same as an ethnic group (P172) to me, but others may disagree.) There is no Scottish or Catalan passports, but most people identify as being from those regions rather than the broader countries like United Kingdom (Q145) or Spain (Q29). Look at w:ca:Artur Mas i Gavarró, for example, the nationality is indicated as "Catalan" rather than "Spanish". If we pulled Wikidata information for infoboxes on cawiki, we would lose that distinction.
On a similar note, how do we handle country of citizenship (P27) for British citizens? Should it be the broader United Kingdom (Q145) or something more specific like England (Q21), Scotland (Q22), or Wales (Q25)? Technically the latter options are more specific and are still countries, but "citizenship" applies more to the United Kingdom than its constituent countries. Cbrown1023 talk 15:07, 21 May 2015 (UTC)
What determines whether a person is of Scottish or Catalan nationality? Just being born there? Ancestry? Self-identification? Is there anything that can clearly show that a person is of a particular non-country nationality, or are the infoboxes just populated by editorial decisions based on a variety of factors?
Re country of citizenship (P27), presumably United Kingdom (Q145), unless there actually is a formal citizenship for UK constituent countries. --Yair rand (talk) 17:52, 21 May 2015 (UTC)
Re aliases: All used versions of an individual's name should be aliases, IMO. That's what aliases are for: a list of anything that someone would reasonably use to refer to the subject. (Sorry if I've misunderstood the question.) --Yair rand (talk) 20:44, 20 May 2015 (UTC)
birth name (P1477) is the correct property for the full name (at birth). To make it searchable, the same would need to be added as alias. --- Jura 04:45, 21 May 2015 (UTC)
Thanks to both you, that's mostly in line with what I typically do. Even though birth name (P1477) says it's only supposed to be used when the birth name is different (i.e., when there was a name change), I sometimes ignore that. Wikidata should really show the birth name results in search results too though, not just aliases. Cbrown1023 talk 15:07, 21 May 2015 (UTC)

put the same refference to multiple properties at once[edit]

Hello. Is there a way to put the same refference to multiple properties at once? For example in Q17442744 I want to put the same reference for all teams in p:P710 at once. Xaris333 (talk) 22:29, 20 May 2015 (UTC)

Not yet. The developers are working on it: phab:T76233. --Yair rand (talk) 22:55, 20 May 2015 (UTC)
@Xaris333: you can use described by source (P1343) and describe your source with authors/urls/pages once, then only specify stated in (P248) as source for all entity properties. Another option is to create separate Wikidata Item for your source, put all your source properties into this new item, and make reference to this item as value of stated in (P248) for statements in original item. -- VlSergey (трёп) 09:29, 21 May 2015 (UTC)
@Vlsergey: Why do you use described by source (P1343) if there is a reference section ? Use the reference section of the declaration and point to the item containing the data of the source using stated in (P248) and then add the specific data like page under the reference section. Snipre (talk) 12:26, 21 May 2015 (UTC)
@Snipre: sometimes it is more convenient. have a look at Arthur Conan Doyle (Q35610) for example. There are duplicate sources at date of birth (P569) and date of death (P570), but there are only a "reference" mention of those sources using stated in (P248). Full "addresses" (title, page, url) of those sources are placed in described by source (P1343) value. We will describe those sources in described by source (P1343) anyway (like "Literature" part of Wikipedia article), so why not to use it? Thus we removing duplication and we don't need to introduce new wikidata items. Well, it works nicely (only?) when encyclopedic articles is used as a source, because single item is described in encyclopedia once. For more complicated cases like article in the book or magazine, it's better to create separate item and link to it using stated in (P248) in reference, as you (and me -- second option in my message above) described. -- VlSergey (gab) 12:44, 21 May 2015 (UTC)
@Xaris333: Please have a look at Help:Sources and if you have other questions please use the talk page. Snipre (talk) 12:26, 21 May 2015 (UTC)

Constraint Irreflexive[edit]

I propose a new Template:Constraint "Irreflexive". Applies to properties that can't connect an item to itself.

Eg it would be very useful for A different from (P1889) B: such claim is recorded when A and B are often confused, so I DO get confused when entering it for A, and have made the mistake of picking up A.

Could also be useful for familial relations.

--Vladimir Alexiev (talk) 07:54, 21 May 2015 (UTC)

@Vladimir Alexiev: All properties (whether actually irreflexive or not) are considered irreflexive in constraint violations report. See Wikidata:Database reports/Constraint violations/P17#"Self link" violations (although it is not real violations). In addition, there're Special:AbuseFilter/49 plus self-referencing tag.--GZWDer (talk) 09:44, 21 May 2015 (UTC)

Al Rayyan: municipality AND town in Qatar, three different data items here, Persian language difficulty[edit]

Hallo, many language versions of WP do not differentiate the municipality of Al Rayyan from the town of the same name. Until now, Wikidata has had three different items for this disentangled set:

  1. Q311272 which I have used for all articles that are concerned with the municipality only
  2. Q18576873 which I have used for all articles that are concerned with the town only
  3. Q11686604 which used to contain links to the Polish and the Macedonian language versions of articles for the municipality (I have moved those to item no. 1).

Now, I think I have identified the articles (in Latin and Cyrillic alphabets which I can read) for either item correctly but I am unsure about the articles in arabic languages. Especially, the items in Persian are a riddle to me:

Q311272 links to the Persian Wikipedia article fa:الریان and to the Persian Wikivoyage article voy:fa:الریان, Q11686604 links to the Persian Wikipedia article fa:ریان and to the Persian Wikivoyage article voy:fa:الریان (استان). I have no idea what either of them is concerned with.

Could anyone please help me disentangle these? Or could you direct me to a more specific help desk? Thanks, --Andropov (talk) 13:43, 21 May 2015 (UTC)

done Michiel1972 (talk) 14:27, 21 May 2015 (UTC)
Thank you! --Andropov (talk) 14:29, 21 May 2015 (UTC)

Aliases in other-language scripts[edit]

Hello. I'm new to Wikidata but have been trying to clean up some of the items for Korean music artists, primarily by adding descriptions. However, I've also been cleaning up the English- and Korean-language "also known as" fields by, for instance, removing the names of individual members from group entries, which apparently were pulled in by accident. My question comes from the following: I've also removed names in the alias field of English entries in Korean and Chinese scripts (they are already present in those languages' entries, so there was no actual loss of information). Additionally, I've removed copious amounts of alternative romanizations from these fields, as these aren't usually included in the subjects' Wikipedia entries unless they are especially prominent or confusing. I figured all these things had been pulled in automatically when information was pulled into Wikidata. However, I belatedly checked the article histories and found that the info had been added manually by another user. I left a message for that user but it appears he/she is only active here sporadically. I've read the guideline for aliases but found no help. Could anyone provide clarification about inclusion of non-roman scripts and alternate romanizations as aliases in the English-language Wikidata? Thank you so much for any help! Shinyang-i (talk) 17:35, 21 May 2015 (UTC)

@Shinyang-i: Aliases are intended to help users find an item; if the user types one of those other romanisations, they will only find the item if it is present as an alias, Please restore those you have removed. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:43, 21 May 2015 (UTC)
@Pigsonthewing:Can you provide a guideline as to how many romanizations should be included? There can be sometimes 30 or 40 for a single name. It will go far, far beyond what were originally present. Also, are aliases providing the same functionality as redirects do on Wikipedia? Shinyang-i (talk) 17:45, 21 May 2015 (UTC)
@Shinyang-i: See Help:Aliases. --Yair rand (talk) 17:55, 21 May 2015 (UTC)
@Yair rand: Already did that, as stated above. Can anyone actually engage with me in a discussion of this topic instead of admonishing and pointing? This is worse than Wikipedia for biting the newbies. Shinyang-i (talk) 17:58, 21 May 2015 (UTC)
This seems to get high rather quickly here. There is no limit to the number of alias an item can have, and this is not a priority to set one. Aliases are just here to help, nothing more. I see no reason to set a limit. Do you have only aesthetics concerns ? If it ain't broken, don't fix it. TomT0m (talk) 18:09, 21 May 2015 (UTC)
Of course I don't have "aesthetics concerns", I was merely trying to get information; I don't know why you're endowing me with emotional motivations on this. I have zero problem with including dozens of romanizations. But I was told to restore what I'd removed, yet what I'd removed was only a drop in the bucket of what apparently was supposed to be there, and thus the answer wasn't exactly clear. The documentation on Wikidata is pretty sparse and there aren't a lot of historic discussions to refer to. Navigation is a little more challenging than at Wikipedia. A lot of the existing items are not "clean" at the moment, and there are no such things as "Good Articles" to look to for best practices. It's impossible to know if changes previously made to items are "correct" or not without asking. Thus, "stupid questions" like mine are inevitable and I think should be treated with a tad more respect than what I feel I've received. I'm pretty blunt myself but this has been a bit much. Not very motivating. Shinyang-i (talk) 18:17, 21 May 2015 (UTC)
Is it clear now ? TomT0m (talk) 18:35, 21 May 2015 (UTC)
Well, no. Further questions:
  • The "fuzzy" spelling issue mentioned in the Aliases documentation - is it live yet and what kind of "fuzzy" spelling does it recognize? For instance, will "nyeo" and "nyuh" in romanization of Korean be seen as "fuzzy" spellings?
  • Should other-language-script names be included, and if so, how many languages? Those from all Wikipedias? Previously only Korean and Chinese were added. How about hanja (subject's name in Korean Chinese-origin characters, which may or may not be the same as their name in Chinese)? Thai? Russian? Etc.
  • Clarify capitalization - in my experiences on Wikipedia, capitalization of a single word makes no difference but it does for all-caps words; i.e. btb = Btb but btb or Btb =/= BTB. Is that correct?
  • "Alias" implies it's an alias for the item in question, but on Wikipedia redirects are often used to send a user who types in the name of a non-notable band member to the band's article. So should non-notable band member names be included in aliases of the band, as well? (Aliases documentation indicates "no", yet they were there and I was told to restore what I'd deleted...)
  • If a person's entry name is their full name but they are often called by just one of those names, does the single name need to be added as an alias? Example: Item name: John Smith. Is "John" needed as an alias? Again, it seems like "no", but I was told to restore what'd I'd removed, and I removed a lot of those...
These are the major issues I've run into so far. I'm sure more will pop up. Thank you for your time. Shinyang-i (talk) 18:51, 21 May 2015 (UTC)
This is not major issues from my point of view, and has a very low impact. So my answer will be I don't really care. And most of Wikidatans as well. That's why you can't find a lot of documentation on this, and as far as I can tell it's a good thing :) So my advice : don't take too much time on this and just focus on something else. TomT0m (talk) 19:09, 21 May 2015 (UTC)
Except maybe for the band members one : it should be clear that an item about a band must not be used to speak on one of its member who should have its own item. So if there is no item for the band member, remove the alias so that this do not happens. On the other hand if the band member item exists and that it is far better know than the name of the band, it could help to find the band item. TomT0m (talk) 19:13, 21 May 2015 (UTC)
@Shinyang-i: From what I can see, the aliases are there to make it easier to find the items. There is no sematic statements based on them. There is no way to add a source to them and there is nothing who tells that they always have to be correctly spelled. I would prefer to say that there is no definite truth about what the aliases should include, except that they should include things that will help the users and the search-engins to find the correct item. My opinion is that you should make a judgement of your own, and not try to identify an exact policy for every issue. Of course, there are aliases that should not be there. If you search for Antichrist, you should normally not be linked to Barack Obama. -- Innocent bystander (talk) 19:18, 21 May 2015 (UTC)
@Innocent bystander: Thank you for the informative and polite answer, and for not telling me no one cares about my contributions. It's much appreciated. As it is, I will just do what I think makes sense and not interact with other uses. First time I screw up and get yelled at, I'll be gone. Thanks again. Shinyang-i (talk) 19:42, 21 May 2015 (UTC)
Hi @Shinyang-i: :) I'm a bit late to the discussion, so I hope you're still around.
I think you have a very good question. The impression I get from both Help:Aliases and the prominence of aliases in the interface is that aliases are only designed for commonly used things which could be considered alternative labels, yet the impression I get from quite a few users is that aliases are just a place to put as many hints for the search as you feel like adding. Those are actually two separate concepts, so it's not a surprise to me that people are disagreeing.
I think that if aliases are not actually important, are just hints for the search and are not really of interest to humans, they should be made a lot less prominent in the interface. When the aliases take up so much space on the screen (above all of the interesting data even!), a direct consequence of too many aliases is that it makes the site harder to use: They get in the way of viewing and editing statements and they make editing labels slower.
If aliases are supposed to be alternate labels which are of interest to humans, then there is the question of how people can add data which is only for helping the search.
Either way, Help:Aliases does not seem to match how many people are currently using the field, which is a problem, because it will lead to more scenarios like this one.
- Nikki (talk) 01:17, 22 May 2015 (UTC)
Hi Shinyang-i. Welcome to Wikidata. You mention you have been looking at Korean and English language labels and aliases so you are familiar with the labels in other languages. If you have a Babel Box on your home page then that can help control which other language labels you see. See User_language for details. Feel free to drop by my page if you have questions you don't feel comfortable about asking here. Hope this helps. Filceolaire (talk) 06:09, 22 May 2015 (UTC)

Time to ping @Denny: here :) And also some of the major writer of Help:aliases, @Thepwnco, Sven Manguard, Bene*: (from the numbers on this history, who else ? ) to check community consensus. TomT0m (talk) 06:23, 22 May 2015 (UTC)

@Shinyang-i: Sorry for the not so great experience you had so far with the project. I hope it will get better. I usually have a really fun time on Wikidata. Thanks, TomT0m for pinging me in. Aliases were designed to support the search, this is also the reason why they don't have qualifiers and references, e.g. "Pope Francis" would be a good alias for "Francis" and the other way around, when referring to the current pope. It is merely meant as a technicality to help people find entities easier. They should not be considered to be a proper part of the knowledge base, but more meta-data about the topic that helps with searching.

Ideally search should get better over time with new software features. Romanizations or other transliterations, fuzzy, soundex-like search terms, etc. should automatically be created and enrich the search for entities, as well as some of the monolingual properties, like ISO codes, etc., translations, etc. But we are unfortunately not there yet. This would make great volunteer contributions, or summer of code-like projects!

I would, in general, try to use common sense in the absence of data. The more prominent a topic, the more care we should give to make sure it is found. The UI is indeed a bit misleading (also, I never been entirely happy with the "also known as" wording). But also less prominent topics deserve to be treated with the Aliases that will help surfacing them. Basically what I do is when I use Wikidata, and something does not show up with the search term I have used when I expected it, I go and add it. It would be great if we could actually gather the data and see where we are missing out: what are the search terms that do not deliver even though we could? @Lydia Pintscher (WMDE):, any chance of getting that data? This would allow us to make further rational decisions on the usage of aliases. --Denny (talk) 18:08, 22 May 2015 (UTC)

These are all very interesting points, @Denny, Filceolaire, Nikki:. I appreciate the response. I'm starting to run into situations like this one: Jung Eun-ji - and that's not even a quarter of the common romanizations for that name ('Chung' and 'Cheong' are also common spellings for that surname, plus the word orders all need to be reversed for Western style). It's hard to pick out which ones users are most likely to search, so I've been adding tons of them. But then, as already noted, you get a page dominated by alternative spellings, completely dwarfing the "good stuff". I'm interested in the efforts to program in fuzzy spellings and such. I can attest that whatever Google does, it works well for Korean. Even Ebay is pretty good with alternative romanizations. I'd be willing to help with such a project however I can. Meanwhile, I may lay off on adding a gazillion alternate spellings and stick with a few basics, plus actual aliases, and focus more on adding descriptions and such, since it seems maybe there will be some evolution in intent and functionality of the alias area. Thanks for all the feedback. This is all a lot of food for thought. :)
Mmmm while I'm here, can I ask another little question? I'm working with singer-related items, and it appears that the way to list band members on a band's item is to use the "has part" property. The reciprocal of that is "part of", but there is also a "member of" property. Which of the latter two is appropriate for a band member's item to designate their membership in a band, for consistency's sake? If anyone knows... :) Thanks! Shinyang-i (talk) 21:26, 22 May 2015 (UTC)
Looking at Wikidata:WikiProject Music, it seems like the suggestions there are member of (P463) for bands and part of (P361) for things like "Person 1 & Person 2" duos. - Nikki (talk) 01:41, 23 May 2015 (UTC)

Sweet kate
Pictogram voting comment.svg Notified participants of Wikiproject Music

Very helpful link, @Nikki:. Thank you! Shinyang-i (talk) 05:07, 23 May 2015 (UTC)
Aliases is partly of work around for cross language searching, partly for fuzzy searching. I've always wondered if the search result should include other languages, or whether that would clutter the search result to much. It could be done by raking close languages higher than the more distant, transliterated strings even less, and then duplicates should be removed. We could then reuse both labels and aliases from other languages to broaden the search, thereby increasing the chance for a useful result. Jeblad (talk) 16:25, 24 May 2015 (UTC)

Items for Wikisource[edit]

We have Q19063070 which is a document about a chap called James Green, but he also has an item at Q6134904. Can we not just have one item for him, and link all these together in the sitelinks? If this is not possible, then what should the label and description be for the Wikisource item? MSGJ (talk) 08:22, 22 May 2015 (UTC)

These items should be/could be linked together, but not necessarily by the site-project links. If one items is about a person, and another about an encyclopeadic article about that person, it's not the same thing, but they can be linked. This is pretty much analoge to the problems I brought up earlier this week on Wikinews. A news article about a person is not the same as that person, and should have different Q-items in Wikidata that are linked. Edoderoo (talk) 11:23, 22 May 2015 (UTC)
Vlsergey: thanks for your detailed answer. I have a few comments / suggestions:
  • Is this "convention" documented anywhere? It is official?
  • I have relabelled a few of these by hand. Perhaps you could look at my recent contributions and tell me if they look okay?
  • I think it is the responsibility of those creating / importing items to do so correctly, and to make sure they have a proper label and description. The current arrangement is polluting the database making it hard to find other items.
  • Would it be a good task for a bot to relabel all of these items accordingly?
MSGJ (talk) 11:31, 22 May 2015 (UTC)
@MSGJ: 1. this convention is not documented yet, it's only implemented in some JavaScript code used by ruwiki and ruwikisource. We are expecting the formal process of discussion to start after arbitrary access is enabled on ruwiki & ruwikisource, because we will need bot work to handle ~300k links like above. 2. Actually, title of item in Wikidata doesn't matter a lot. Current labels looks good (and correct according to convntions above), but the most important point that we have prefix ("DNB00" or "DNB12"). Thus noone will made a mistake linking those items instead of original topics items. 3. Well, may be. But my bot is blocked :-) I assume one need to make those conventions "official", find some consensus among all users of Wikidata, may be open some RFC, after this bot can do his work. -- VlSergey (gab) 11:39, 22 May 2015 (UTC)


and not:

which seems more precise? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:26, 22 May 2015 (UTC)

    • @Pigsonthewing: the reason behind this is simplification for bot, LUA and template processing. When we need article from, for example, DNB00, we looking for statement with described by source (P1343) value of DNB00 and extract link to article from stated in (P248). Second reason if unification of described by source (P1343) values, both for encyclopedias present in wikisource and ones that doesn't: all of those will have encyclopedias Q-item in described by source (P1343) field. Also, "precise" is not an argument here. Precise item is specified as qualifier:
  • To make this easy to handle, a solution that doesn't require the use of qualifiers would be helpful.
    The problem with the second approach with P|1343 is that it mixes DNB with any other resource that happens to be at Wikisource.
    Both approaches don't make it easy to find DNB merely because it has the disadavantage of being at WikiSource. --- Jura 06:13, 23 May 2015 (UTC)

interwiki text[edit]

How can I access a specific interwiki link? Lets say I want to get the link to for Q1524 (Athens). If I could get just a text string I would be happy aswell. Thnx in advance -- Spiros790 (talk) 18:55, 22 May 2015 (UTC)

I was looking for that as well. Module:Wikidata doesn't seem to offer it yet. --- Jura 06:16, 23 May 2015 (UTC)

how about that?

  • {{#invoke:Wikidata|getLabel|entity=Q1524|lang=it}} --> Atene
  • {{#invoke:Wikidata|getLabel|entity=Q1524|lang=de}} --> Athen
  • {{#invoke:Wikidata|getLabel|entity=Q1524|lang=el}} --> Αθήνα

thats not an interwiki text string but it should be OK in most cases.

-- Spiros790 (talk) 06:51, 23 May 2015 (UTC)

If that was reliable, we could do away with half of Wikidata ;) --- Jura 06:56, 23 May 2015 (UTC)
I need a way to get the name of an entity, in english, german, etc. It doesnt have to be a reliable link or something :) That would be cool though. I m creating infoboxes that use wikidata properties for, currently tryin to retrieve the name of a city in its mother tongue. -- Spiros790 (talk) 07:40, 23 May 2015 (UTC)
There is a property "native name". I think it's meant for this, but I'm not sure if it's frequently used. An alternative would be "official name". --- Jura 07:48, 23 May 2015 (UTC)
a "native name" property is exactly what Ι am looking for. I don’t think Ι have ever seen it though. Thnx for your input :) -- Spiros790 (talk) 07:53, 23 May 2015 (UTC)
it exists for people name in native language (P1559)… but for places, I'm not sure… would be useful though… :)
besides, there would be a question of name at a certain period - Athens is not problematic, but Istanbul/Constantinople, etc. - would require date qualifier :) --Hsarrazin (talk) 10:11, 23 May 2015 (UTC)
There is an item for the old norse name of Constantinopel: Q2008228 (Q2008228) Face-smile.svg
I have used official name (P1448) to Q13106534 (Q13106534), but I do not exactly know what is official with it, so I am not so sure about my choise in that case. And what the "native" name is, is difficult to say, since we do not really know who is "native" here. It depends on your pov. -- Innocent bystander (talk) 10:54, 23 May 2015 (UTC)
I don't think "official" is correct. The native name of "Athens" is "Αθήνα". Thats how Greeks call it. No pov here... -- Spiros790 (talk) 11:59, 23 May 2015 (UTC)
The pov-problem is what you consider as "native language" in this case. People have lived there for thousands of years, but we do not know which language they spoke and definitly not what they called the place. We know that today Swedish-speaking people uses the name "Hälla" and Sapmi-speaking people uses the name "Hïella". Some people say that a Germanic people have lived there, other Sapmi while some think a third group of Creole people lived there. My pov is that it was a Uralic people which later changed ethnicity and language into Swedish. Consider what would happen if you discovered that Sokrates used the name "Atina" for your "Αθήνα", since the place originally was Turkish and not Greek. -- Innocent bystander (talk) 16:26, 23 May 2015 (UTC)
Socrates called it ΑΘΗΝΑΙ and sure as hell it wasnt "Turkish" back then, thats just retarded. Anyway what I was lookin for was a way to get the name its current inhabitants use... i.e. your statement is irrelevant anyways.-- Spiros790 (talk) 17:32, 23 May 2015 (UTC)

As a rule of thumb, we should call "native" the name given to it by its current inhabitants. I call Istanbul "Constantinople", because I am Greek and my ancestors used to live there. But we should use "Istanbul" for a "native name" property here. -- Spiros790 (talk) 12:03, 23 May 2015 (UTC)

what I meant was that the native name can change through time… ;) — see
as for Istanbul (Q406), its current native name woul be İstanbul (in turkish) , and not Istanbul ;)
native meaning as the natives call/write it is different from official - but probably official name (P1448) would be a good approximation, since the intent was to get the "in the language of the country" writing, more than the "legal pov" :)
the nickname given by natives would certainly not be what you want to get here, (like Paname for Paris (Q90)) --Hsarrazin (talk) 13:42, 23 May 2015 (UTC)
yes it can change through time. But what we should care about is now. Please stick to the points that are being made, not to irrelevant spelling errors like "İstanbul". Official name is wrong. Most of them are like "Municipality of X". Nicknames are also wrong for obvious reasons.
btw I am well aware about the history of Istanbul/Constantinople, my family used to live there until the 30ies... Some of them are still alive and well today.
yep the native name of this city would be İstanbul
thnx for your feedback :) -- Spiros790 (talk) 14:00, 23 May 2015 (UTC)

another approach would be to record all native names through history and rank=best the one used today. For İstanbul: Βυζάντιον --> Κωνσταντινούπολη --> İstanbul. I don’t think that this is useful info though. There are also serious POV problems -- Spiros790 (talk) 14:07, 23 May 2015 (UTC)

BTW, would anyone have an answer to the initial question? --- Jura 07:24, 24 May 2015 (UTC)

yep that would be intresting, we are not here to discuss history... -- Spiros790 (talk) 08:26, 24 May 2015 (UTC)

Property for 1st line of a poem ? (incipit)[edit]


Is there a property to enter the 1st line (verse) of a poem… many poems have no title, and are recognized using their 1st line.

I remember a discussion about it, a year ago, but I cannot find it any more… so, is there a property to add it, or do I just input it in title (P1476) ?

Thanks for your help, there are already thousands of poems (from various wikisources) to complete on wikidata --Hsarrazin (talk) 03:42, 23 May 2015 (UTC)

Sounds like a good idea! No, I am, not aware of any property for that. IF you propose it, ping me and I will support! -- Innocent bystander (talk) 06:04, 23 May 2015 (UTC)
@Innocent bystander: done : here - hope I did it right, feel free to add info if you think I forgot some - thanks :) --Hsarrazin (talk) 10:04, 23 May 2015 (UTC)

Wikidata:WikiProject Elections[edit]

I have set up Wikidata:WikiProject Elections, feel free to tell us your ideas on the talk page. -- Innocent bystander (talk) 08:09, 23 May 2015 (UTC)

Please, long time request : when will we be able to add redirection pages to links ?[edit]

In wikisource, we often use disambiguation/redirection pages to store data from the work, and redirected pages for editions…

When there are more than one edition, the page is disambiguation, but as long as there is only 1 edition, we use a plain redirection, to spare the reader a clic :)

Now, to link with wikidata work and edition items, we have a problem…

This is problematic, since, when other editions will be added, it will be necessary to change wikidata fr link, just because of the fact that the redirection is not kept… 

I have seen many of them on wp, but I think those were created before the pages became redirections… and on ws, we have an automatic change of link when that happens…) :/

so, now, my question stands : when will we be able to add redirection pages to links ? pretty pretty pretty PLLLLEEEEEAAAAAAAAAASEEEEEEE --Hsarrazin (talk) 14:13, 23 May 2015 (UTC)

Template:Link LA (Q17609506)[edit]

Can you help? Thank you, Conny (talk) 14:30, 23 May 2015 (UTC).

  • But it's up to the local communities if they want to keep the template. --Stryn (talk) 12:22, 24 May 2015 (UTC)
Stryn only 1 project uses it. -- Magioladitis (talk) 14:28, 24 May 2015 (UTC)
I know many pages (including templates) that doesn't exist on other wikis... --Stryn (talk) 14:29, 24 May 2015 (UTC)
It has no tranclusions. It will be deleted from the wiki soon. WE can only wait. -- Magioladitis (talk) 14:31, 24 May 2015 (UTC)
It's still up to the local community whether they want to keep the template for historical purposes or not. We on Wikidata can't make the decision. --Stryn (talk) 14:35, 24 May 2015 (UTC)

Eurovision Song Contest 2015[edit]

How do we take care about the songs which will be performed today in the finals and how do we denote that a song didn't make it to the finals even though it competed in the semi finals? --Jobu0101 (talk) 16:09, 23 May 2015 (UTC)

As far as I can see we don't. See Q15953308 for example. Last years Belgian entry. Song didn't make it to the final, but that isn't denoted in the item. Mbch331 (talk) 16:24, 23 May 2015 (UTC)
I think all the tables which exist in the Wikipedias concerning the ESC should be creatable using only Wikidata. So we need appropriate properties. In particular Wikidata should know which songs competed and how they performed. --Jobu0101 (talk) 17:45, 23 May 2015 (UTC)
@Jobu0101: What extra properties do you think we need? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:11, 23 May 2015 (UTC)
I added the points for the final. For the semi final results, I think we should create sepearte items for each show. --Pasleim (talk) 09:30, 24 May 2015 (UTC)

I think, also the songs should be linked somehow. But it's already a good basis what Pasleim did. --Jobu0101 (talk) 14:41, 24 May 2015 (UTC)

Eurovision Song Contest 2015 (Q15215081) has some good stuff. My suggestions:
  • We should have a separate item for each of the two semi-finals.
  • The 'participant' statements should link to songs, not to singers. Remember that the winning songwriters get invited on stage too at the end (at least they used to - Didn't see them this year).
  • Statements re singers/songwriters etc. for each song should go on the item for the song.
  • "country', 'ranking' and 'number of points/goals scored' are the appropriate qualifiers for the participants.
  • we can add a winner (P1346) statement. The semifinals can each have 10 winners (i.e. songs that went through to the finals).
This does raise a question about the item for the final. Can this also act as an item for the whole contest or do we need a separate item for that?
OK? Filceolaire (talk) 21:01, 24 May 2015 (UTC)
@Filceolaire: I totally agree on your points. Concerning your last question I think the correct way is to have a separate item for the final since the Eurovision Song Contest 2015 (Q15215081) is the union of all three shows and not only the final. Once we have worked out the Eurovision Song Contest 2015 (Q15215081) it can serve as show case for the other 59 ESC events. --Jobu0101 (talk) 08:30, 26 May 2015 (UTC)

Entering information on new item not in Wikipedia yet[edit]

I created a new item. Where do I enter text?  – The preceding unsigned comment was added by Blaze2453 (talk • contribs) at 20:35, 23 May 2015‎ (UTC).

You'll have to create a new article on any of the other Wikimedia projects, this is also the place where you can enter text. Wikidata is meant to connect the several articles on the other projects, this is done with the links on the right hand side. Edoderoo (talk) 08:37, 24 May 2015 (UTC)
As you can imagine, it is a wee bit more complicated.. Text, sure. Statements are very much information that can be added to Wikidata. Thanks, GerardM (talk) 09:36, 24 May 2015 (UTC)
@Edoderoo: You misanderstand Wikidata, Wikidata is here to build a database of structure datas. This interwiki feature is a small part of this project :) Wikidata is here to build statements like  who will be readable from any language, and also by machines. What text is obviously not :) TomT0m (talk) 09:51, 24 May 2015 (UTC)

About official name (P1448)[edit]

Talking with @Ahoerstemeier: there is a doubt about this property. I have added Italian version here but but for Ahoerstemeier isn't correct because English description of the property say «official name of the subject in its official language». So how to use this property? --ValterVB (talk) 18:54, 24 May 2015 (UTC)

ValterVB, this property is for the official name as defined in some official source (usually a law or parliamentary resolution). Some countries and cities will actually define official names in more than one language. That is why the "official name of 'Thailand' is something in Thai which I cannot read because I don't understand Thai. Unofficial translations should go in the Label or in the aliases. If you don't have a source to confirm that 'Regno di Thailandia' is officially endorsed by the government then it should be deleted.
There is some discussion of 'transliteration' properties or a 'translation' property which could be used as qualifiers to 'official name' but these properties have not yet been approved. Filceolaire (talk) 20:40, 24 May 2015 (UTC)
ราชอาณาจักรไทย translates to "Kingdom of Thailand". The translation of this term, which then would be in all possible languages, should not use the mono-lingual datatype, but the still planned multilingual datatype, that's why that property is still pending. By the way, there's also quite a big overlap between official name (P1448) and native label (P1705), in most cases the official name would be the same as the native label. Ahoerstemeier (talk) 20:49, 24 May 2015 (UTC)
In I read: "Thailand officially the Kingdom of Thailand", in I read: "La Thailàndia ufficialmente Regno di Thailandia", in I read "Thailand offiziell Königreich Thailand", in I read "La Thaïlande, en forme longue le Royaume de Thaïlande", so every language have an official name that is also used in infobox. --ValterVB (talk) 20:56, 24 May 2015 (UTC)
Since Thailand has embassies in many countries, and those embassies will communicate "officially" with the governments of those countries in their respective languages about matters concerning the Kingdom of Thailand, I presume there exist official names for the item in many, many more languages than the official ones "of" the item. So what to do with "official" names of the item in "unofficial" languages? -- Gymel (talk) 21:08, 24 May 2015 (UTC)
These are official translations but not official names. Snipre (talk) 07:21, 25 May 2015 (UTC)

Ratified date, effective date[edit]

Using the example of Continental Association (Q4231641), the date created is October 20, 1774, the date ratified is the same, and the effective/operative date is December 1, 1774. For the first, I'd use inception (P571) but I can't seem to find any for the other two. Before I go requesting new properties, I'd like to assume that something is already in existence. What have others used? Thanks, Hazmat2 (talk) 22:36, 24 May 2015 (UTC)

Instead of creating new properties, I would use significant event (P793), i.e.
Thanks for the info. Second question then: how would one go about calling these qualifier statements in an infobox? w:Template:Infobox document is the specific one I'm using at the moment. Thanks, Hazmat2 (talk) 17:40, 25 May 2015 (UTC)
Look at your module module:wikidata in your WP (ex.: fr:Module:Wikidata). If someone develops the code you should be able to filter results according to some qualifiers or some values. You should perhaps code a little in lua to obtain the data you want. Look if someone already develops module:infobox. The main of advantage of significant event (P793) is the possibility to group all data (time, location, person,...) in one statment. Snipre (talk) 19:51, 25 May 2015 (UTC)

Combat Flight Simulator WWII Europe Series (Q1735558)[edit]

Combat Flight Simulator WWII Europe Series (Q1735558) seems to be mixed up, can somebody split it? Sjoerd de Bruin (talk) 13:07, 25 May 2015 (UTC)

Testing Flow?[edit]

Hey folks :)

I'm currently at the MediaWiki hackathon in Lyon and talking to lots of people about Wikidata and other projects. One of the people I talked to is Danny, the product manager of Flow. There have been discussions here about using Flow on Wikidata here before. At that time Danny said he wants his team to finish a few more things before Wikidata could use it. Those blockers have now been resolved and I think it'd be great if we can help another project with our testing and feedback to make the discussion landscape around Wikimedia better. Danny set up a test page for you to go wild on and there is a page with the current features/benefits of Flow. If you like it and are willing to give it a try I think item talk pages are actually a nice test case since not too much is happening there and the discussions that do happen are not too complex. And depending on how that goes we can look at other spaces. What do you think? If you want to give feedback on the state of Flow you can use Wikidata talk:Flow.

Cheers --Lydia Pintscher (WMDE) (talk) 13:41, 25 May 2015 (UTC)

Does Flow handle templates in section headings better? (anchors, etc.) Sjoerd de Bruin (talk) 14:55, 25 May 2015 (UTC)
Flow doesn't support section headings yet, but you may be able to use separate threads to do what you want. Can you describe an example of what you'd like to do? DannyH (WMF) (talk) 16:25, 25 May 2015 (UTC)
I use a template in the section above this one, would that be possible in the thread name? Sjoerd de Bruin (talk) 16:30, 25 May 2015 (UTC)
I have a very bad impression of FLOW on the English Wikipedia. It is substandard and is being aggressively pushed through by WMF. Typical discussions of FLOW tend to go around the fact that it is not compatible with the current markup and can not be turned on in the opt-out regime.--Ymblanter (talk) 17:49, 25 May 2015 (UTC)
@Ymblanter: I think it would be damaging to take a position only wrt. the WMF. I don't get the incompatibilities in the markup you refer to because it seems based on the Visual editor libraries. Do you mean it would not be possible for old talkpages to be converted to Flow ? The only valid question is imho : can it fits our discussion patterns scheme ? I think it fits like a charm and we get better notifications and answers tracking. TomT0m (talk) 18:11, 25 May 2015 (UTC)
My understanding is that converting is not possible. The old talk pages will be archived, and the new pages will use these Facebook-style discussion tools.--Ymblanter (talk) 11:19, 26 May 2015 (UTC)
Pictogram voting question.svg Question What about talkpages that are used partly for documentation, like Talk:Q5 or Property Talk:P31 ? @DannyH (WMF): TomT0m (talk) 18:11, 25 May 2015 (UTC)
I don't think Flow is ready for Wikidata yet, and I've been following Flow since the beginning. Mainly per TomT0m.--Jasper Deng (talk) 18:27, 25 May 2015 (UTC)
@Jasper Deng: The question might not be is Flow ready for Wikidata but Is Wikidata ready for flow. Using talkpages for documentation seems to me like a workaround. TomT0m (talk) 18:32, 25 May 2015 (UTC)
So, addressing the broader issue of where to put property/item documentation then, where are we going to put it? Note that the English Wikipedia keeps a lot of internal information like WikiProject affiliations on article talk pages too.--Jasper Deng (talk) 18:34, 25 May 2015 (UTC)
That's a question for @Lydia Pintscher (WMDE): :) What are precisely the issues on enwiki, so that we know what we're talking about ? TomT0m (talk) 18:39, 25 May 2015 (UTC)
I think the issue of "documentation" on talk pages is much smaller on item pages. That's why I suggested them to start. The stuff that is on property talk pages can hopefully soon be moved to a large extend to statements on the property page. The students' project to improve constraints reports is making good progress and can hopefully go live soon in a first version. That being said the Flow developers have this on their radar and I have seen a version that offers a space for this as well. If I understood Danny correctly this will be out in the next week or two. --Lydia Pintscher (WMDE) (talk) 18:46, 25 May 2015 (UTC)
I wonder if we should try to make editing items closer to the way talk pages are edited (with or without flow). It seems that some are at ease filling up talk pages, but seem to have difficulties adding statements to items. --- Jura 06:30, 26 May 2015 (UTC)
Symbol oppose vote.svg Oppose at least until phab:T93883 is resolved. Wikidata is not a test site. --Ricordisamoa 10:04, 26 May 2015 (UTC)
The first point is fair. I'll bring this up with the Flow team. The second point: Please remember how much Wikidata has and still does depend on other projects being willing to try something new. Without projects being willing to test what we have we would not have gotten anywhere. --Lydia Pintscher (WMDE) (talk) 19:30, 26 May 2015 (UTC)

How to get the value of a qualifier from a property if an other qualifier has a specific value?[edit]


I'm trying to extract the publication date from version if version type is either stable version, beta version or alpha version. Is there a way to do that? I'm trying it out on the Wikidata Sandbox with Modèle:Infobox Logiciel/Bac à sable and Wikipédia:Wikidata/Bac à sable The RedBurn (ϕ) 18:19, 25 May 2015 (UTC)

@The RedBurn: Before answering such questions, I'd like a discussion aboout the model … What's the bigger picture about software items model you have in mind. I'd like to see if it would be better to have one item per realease … But the answer anyway : use a scribunto module and code in luas. TomT0m (talk) 18:27, 25 May 2015 (UTC)
Thank you for the quick answer! What do you mean by one item per release? No alpha / beta version? Or either an alpha or beta version? Or only a stable version? Or no history of the versions? The RedBurn (ϕ) 18:32, 25 May 2015 (UTC)
I mean items like for example <Linux 3.0> with statements like
< Linux 3.0 > instance of (P31) miga < Linux stable Release >
and in a similar model (and maybe the same properties) of the Wikiproject Book (see also Functional Requirements for Bibliographic Records (Q16388) (View with Reasonator). This allows to put statements like added features and so on.
Well, it would probably be too much for the current intended use: the latest stable and unstable versions of a software for the French software infobox. The RedBurn (ϕ) 07:47, 26 May 2015 (UTC)
@The RedBurn: We need to think bigger imho :) We got an effective model for this, and migrating to this later will be harder. The devteam has already deployed the feature needed to make this work on infoboxes on some projects (including Wikidata for tests). This overall is to reduce redundancy. TomT0m (talk) 08:06, 26 May 2015 (UTC)
@TomT0m: I guess you mean that for pages like Android_version_history, but those aren't the majority and it could be done differently just for them, without adding complexity to the majority. The RedBurn (ϕ) 08:22, 26 May 2015 (UTC)
@The RedBurn: I does not add complexity to have only one model, it makes things more simple. It just add a little extra work : creating an item for a release. That's all. TomT0m (talk) 08:32, 26 May 2015 (UTC)
@TomT0m: I think it also adds complexity to have multiple items for one page. And that extra work isn't needed for 99.9 % of the pages, so I personally think it's overkill and I won't try to find a consensus to change the current system, but feel free to do so. The RedBurn (ϕ) 08:38, 26 May 2015 (UTC)
@The RedBurn: There is very often several items for one page (see the discussion about Wikivoyage earlier ;). What's the current system by the way ? We should create Wikidata:WikiProject Software anyway. I think it's a good thing in Wikidata to have, in general, precise items. For example in the opensource world, Red Hat Enterprise Linux (Q215273) (View with Reasonator) is an item that refers to tons of versions, tons of feature. It quickly needs a more sophisticated model if we want to be expressive, like pointing to a specific version of a sotware included into a version of RHEL. It turns out we have one, at a very low cost. With adding features to version made easy later. Plus the power of Wikidata who will make this available to all Wikipedias so the maiftenance cost will be overall reduced … It would be a shame to do less :) TomT0m (talk) 08:57, 26 May 2015 (UTC)
@TomT0m: Well, the infobox of Red_Hat_Enterprise_Linux doesn't seem like it needs that. So it would seem like fixing something that isn't broken and/or killing a fly with a cannon (against Confucius' advice). The RedBurn (ϕ) 09:38, 26 May 2015 (UTC)
@The RedBurn: Infoboxes are not very sophisticated … Wikidata is far more expressive. Are not such changes a time to think a little bit ? Maybe the infobox is like that just because it was very hard to do better at that time. Now things have changed, opportunity to think again … We're into no rush ;) Maybe we should involve the local communities ? TomT0m (talk) 09:53, 26 May 2015 (UTC)

Wikidata weekly summary #159[edit]

Adding coordinates[edit]

It states that en:Newfound Regional High School doesn't have a coordination on wikidata, so how'd I put that on here? The coordinate is 43° 36′ 25.2″ N, 71° 40′ 37.2″ W Nick2crosby (talk) 22:06, 25 May 2015 (UTC)

Hi Nick. Welcome to wikidata. Here are my comments for you:
To link to an item on wikidata you need to use the q number Q19962872 as we can have lots of items with the same label. There are a couple of cute templates for doing this too. Newfound Regional High School ( Q19962872) shows the label in the readers preferred language. Newfound Regional High School (Q19962872) (View with Reasonator) includes a link to Reasonator - a cool interface for displaying our data.
I changed to the statement you added. Most thing on wikidata are an 'instance of' something. Getting that right helps the property suggester find the most helpful suggestions. Now that there is an 'instance of:high school' statement you will find that if you try to add more statements our property suggester has 'coordinate location' as one of the top suggestions so I'm going to leave it to you to add the coordinates (unless someone else gets there first).
Wikidata:Showcase_items is a good place to see what good items look like, though we don't seem to have any other high schools there yet.
Hope that helps. Filceolaire (talk) 03:15, 26 May 2015 (UTC)

Tracking changes that just impact Wikivoyage?[edit]

Wikivoyage has 18 separate language communities, and we are using WikiData for our banner images so that (for example) a banner created for one destination can be automatically reused everywhere.

One identified scenario is that if somebody on 'Language X' Wikivoyage uploads a banner that is of worse quality than the one it is replacing on English Wikivoyage, then it will not appear in our wiki's 'recent changes' and basically go unnoticed.

A deeper discussion can be seen here

Does anyone here have any suggestions about how we can monitor these changes that impact Wikivoyage without tracking the 'recent changes' of all WikiData items? --Andrewssi2 (talk) 01:28, 26 May 2015 (UTC)

You can go to Special:RecentChanges on Wikivoyage and monitor the changes (you may have to turn on "Show Wikidata edits" in Preferences, or in the header). --Rschen7754 02:22, 26 May 2015 (UTC)
@Andrewssi2: gadget from ruwiki let's you monitor changes even with extended watchlist. Another option is to create bot that updates special page s with, let's say, 100 last updates of «page banner» property on all item. -- VlSergey (gab) 06:21, 26 May 2015 (UTC)

Catholic bishop (Q611644)[edit]

I'm adding Catholic bishop (Q611644) for example here Joseph Hii Teck Kwong (Q1707237) do you think is the right way? occupation (P106)->priest (Q42603) position held (P39)->Catholic bishop (Q611644) religion (P140)->catholicism (Q1841) --Rippitippi (talk) 04:16, 26 May 2015 (UTC)

Catholic bishop (Q611644) is good for 'position held'. This can have qualifier (of (P642):Roman Catholic Diocese of Sibu (Q876489)).
Religion should be 'Catholic Church (Q9592)'. catholicism (Q1841) is a wider term that includes people that are no longer part of the Roman Catholic Church. Look at other more famous bishops for additional properties to add.
Hope this helps. Filceolaire (talk) 06:52, 26 May 2015 (UTC)
I have a added the qualifiers for position held (P39) as well as adding his previous position as auxiliary bishop (Q75178) and therefore as Titular see (Q15217609), as well as the properties Catholic Hierarchy person ID (P1047) and consecrator (P1598) which are also quite useful for Catholic bishops. Only thing I am not sure about - should we use the date of appointment or the date of consecration for start time (P580)?
Its the date of consecration that counts. He is not a bishop before consecration. If the person dies one dey before consecration, he is not a bishop no matter if he is appointed. A bishop can have several appointments, the day he begins his service in the new diocese (Q665487) counts. It is also a good idea to use diocese (P708).--Giftzwerg 88 (talk) 09:47, 26 May 2015 (UTC)

We have also Catholic priest (Q250867) it is not redundant Catholic Church (Q9592) Catholic bishop (Q611644)? --Rippitippi (talk) 10:41, 26 May 2015 (UTC)

In P140, you could hardly add Catholic priest (Q250867). --- Jura 11:40, 26 May 2015 (UTC)

In religion (P140), there are more items with catholicism (Q1841) than with Catholic Church (Q9592). Should we convert them? --- Jura 11:40, 26 May 2015 (UTC)

@Jura1, Rippitippi, Giftzwerg 88, Filceolaire: A possible data structure:

Snipre (talk) 14:22, 26 May 2015 (UTC)

Catholic priest (Q250867) and Catholic bishop (Q611644) are not useful: we can cross religion (P140): catholicism (Q1841) with occupation (P106):priest (Q42603) or occupation (P106):bishop (Q29182) to get the same information. Better to delete these items. Snipre (talk) 14:53, 26 May 2015 (UTC)
The items exist because the wikilinks exist. They should be used because of that. --Izno (talk) 16:41, 26 May 2015 (UTC)
It is not because an item exists that we have to use especially when the item is created for managing links and not based on a classification. Just take the example of bishop: we have auxiliary bishop and catholic bishop. How can I manage auxiliary bishop from the Catholic church ? Do I have to create an item catholic auxiliary bishop ? Snipre (talk) 22:42, 26 May 2015 (UTC)
@Snipre: I'd like to recall a class item could pretty much be associated to a query. It's not a problem to keep them in such a case, it's an opportunity to check if all instances are resuts of a query. TomT0m (talk) 16:59, 26 May 2015 (UTC)
You think with a technical point of view and not with a classification point of view: the problem is if I have a catholic priest and I want to use that level of the classification for person items I should be able to do the same for all religions which have a priest. So for example I should have an item for old catholic priest. This lead to the creation of items which are not necessary because they are just intersections of existing definitions. If I can define a catholic priest, why can't I have an item "dead person" for example ? Or French dead person ? Or French female dead person or perhaps French dead woman is better? For me this kind of question about classifications should be discussed in a help page. Snipre (talk) 22:42, 26 May 2015 (UTC)

deleted property[edit]

What happens here? Thank you, Conny (talk) 14:57, 26 May 2015 (UTC).

As the property P132 (I remember it was "type of administrative unit" and was replaced by instance of (P31)) is deleted, the software cannot now access metadata of the property. Matěj Suchánek (talk) 15:18, 26 May 2015 (UTC)
@Lydia Pintscher (WMDE): Maybe a paper cut? It's a little rough to use the generic error interface message in the edit history. --Izno (talk) 16:39, 26 May 2015 (UTC)
You're right. Not nice. I'll see if we can improve this. --Lydia Pintscher (WMDE) (talk) 18:01, 26 May 2015 (UTC)


Those of you who like to keep track of our milestones might like to note that I have just created EAGLE id (P1900). Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:09, 26 May 2015 (UTC)

Property:P31 with encyclopedic article (Q17329259)?[edit]

Hello, is it advisable (as I did) to use Property:P31 with encyclopedic article (Q17329259)? Jonathan Groß (talk) 16:50, 26 May 2015 (UTC)

@Jonathan Groß: That depndds, onwhich item ? For a Wikipedia article it's not, but in the case of an item about an article of another encyclopedia … it is. TomT0m (talk) 16:55, 26 May 2015 (UTC)
On Q19828642 (Q19828642). Yes, this or a more specific one. Obviously, someone should probably set it for all pages of the same book at once.
More important is main subject (P921). --- Jura 16:56, 26 May 2015 (UTC)
That's what I had in mind. main subject (P921) would be more difficult to fill automatically.
I want to tag all articles from the Pauly encyclopedia. Can I add another property that states that these articles come from this specific encyclopedia? Jonathan Groß (talk) 17:10, 26 May 2015 (UTC)
One solution is to create an item <Pauly encyclopedia's article>, with
< Pauly encyclopedia's article > subclass of (P279) miga < encyclopedic article (Q17329259) (View with Reasonator) >
. An use instance of (P31). It's also possible to use part of (P361). TomT0m (talk) 17:16, 26 May 2015 (UTC)

To add, published in (P1433) with Paulys Realenzyklopädie der klassischen Altertumswissenschaft (Q1138524), you can use this.
As the main topic is already defined in a template, someone might be able to extract that and add it. --- Jura 17:22, 26 May 2015 (UTC)
To scan for articles only, the category "RE:Autoren nach Artikelzahl" might be the most suitable one. --- Jura 17:38, 26 May 2015 (UTC)
Thank you :) I started creating the items with the Item Creator. I suppose it will be running all night. I'll add the properties when all the items are created. Jonathan Groß (talk) 18:31, 26 May 2015 (UTC)