Shortcuts: WD:PC, WD:CHAT, WD:?

Wikidata:Project chat

From Wikidata
(Redirected from Wikidata:Project chat/bjn)
Jump to navigation Jump to 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.
Please use {{Q}} or {{P}}, the first time you mention an item, or property, respectively.
Requests for deletions can be made here. Merging instructions can be found 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 2018/12.

Project
chat

Lexicographical
data

Administrators'
noticeboard

Development
team

Translators'
noticeboard

Request
a query

Requests
for deletions

Requests
for comment

Bot
requests

Requests
for permissions

Property
proposal

Properties
for deletion

Partnerships
and imports

Interwiki
conflicts

Bureaucrats'
noticeboard

Scholarly articles that are book reviews[edit]

If a scholarly article is a book review, should the “main subject” be the edition of the book, or the subject of the book, or both? - PKM (talk) 02:32, 14 November 2018 (UTC)

@PKM: In the absence of a single value constraint on P921, I see no harm in using both. Mahir256 (talk) 03:10, 14 November 2018 (UTC)
@PKM:Typical problem when mixing different classifications: scholarly article is a text format, book review is about the content of a text. WikiProject Books should once fix the classification by analyzing in detail the characteristics of a book. Snipre (talk) 07:37, 14 November 2018 (UTC)
@Snipre: While a "scholarly article" as used in WD is clearly more like an "edition" than a work (based on its properties), currently, "scholarly article" is <subclass of> "article" is <sublcass of> "work" (while also being a subclass of "publication"). If you have an idea of a way to separate "publications" from "works", please lay it out so we can discuss - I agree this area is fuzzy. But I would say that a "book review" is a type of article either way. In any case, I would not recommend or support making separate "work" and "edition" items for every scholarly article in WD. That way lies madness. - PKM (talk) 20:47, 14 November 2018 (UTC)
@PKM: In relation to your original question, I think "main subject" should point to the book-edition. The item should also be identified somehow (genre?) as a book review. I am dubious about "main subject" also pointing to the book-subject. How much would one learn about the book-subject in the round from the book review? Possibly an case for a weaker subject indicator, perhaps the proposed subject keyword property? Jheald (talk) 22:09, 16 November 2018 (UTC)
@Jheald:, that's my first instinct as well. book review (Q637866) is currently a genre, so that part works. Lots of missing books, alas. Also, we have 2789 items with <instance of> "book review". - PKM (talk) 22:18, 16 November 2018 (UTC)

Just going to throw in that IMO academic journal article (Q18918145) is more accurate than scholarly article (Q13442814) for the instance of (P31) statement of a book review. Circeus (talk) 14:24, 21 November 2018 (UTC)

@Circeus: Agreed! - PKM (talk) 20:51, 24 November 2018 (UTC)

My two cents, @PKM: I created Le nom de peuple Rhedones (Q52160525) some times ago. What do y'all think? Cdlt, VIGNERON (talk) 17:52, 24 November 2018 (UTC)

@VIGNERON: That's very nice! - PKM (talk) 20:51, 24 November 2018 (UTC)
Shouldn't "book review" be a property, rather than an item, as Q637866?
Example: I just created Q59319726 for a book. A published review of that book was previously available as Q58565334. There should be a way to link these two. If "book review" were a property, one could then specify the book as a property of the review. ???
We need some way to link a review with the book it reviewed. Please suggest an alternative or endorse this. Thanks, DavidMCEddy (talk) 03:12, 1 December 2018 (UTC)
Another example: Q58570163 is a review of a book for which a Wikidata seems not yet to exist. I plan to create a Wikidata entry for that book.
Shouldn't the "scholarly article" that is a book review have a property = "review of" being the book? And shouldn't the book have a property "reviewed as", being the review?
I see properties "review score", "review score by", and "reviewed by". None of these sound to me like either "review of" or "reviewed as".
Thanks, DavidMCEddy (talk) 19:56, 3 December 2018 (UTC)
@DavidMCEddy: I have no problem using the existing "main subject" to include the meaning "review of", but I wouldn't object to a separate property "review of" in lieu of both "main subject" and <instance of> "book review" for these items. That would be a very clean solution. - PKM (talk) 20:48, 10 December 2018 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── I've also run into a number of "scholarly articles" that are reviews of museum exhibitions. I've addressed these by creating items for the exhibitions and linking from the review using "main subject". (See ‘Opus Anglicanum: Masterpieces of English Medieval Embroidery’, Victoria and Albert Museum, London, 1 October 2016–5 February 2017 (Q57678571).) Whatever solution we come to should probably handle other types of reviews than book reviews (so if we make a new property "review at" or somesuch, it could accept reviews of plays, films, exhibitions, and books. - PKM (talk) 20:44, 10 December 2018 (UTC)

EOL has changed all its URLs/IDs[edit]

Encyclopedia of Life (Q82486) has changed all its IDs and URL formats, making Encyclopedia of Life ID (P830) wrong where it is used. Example: Prunus prostrata (Q1258395) links to https://eol.org/pages/631658 but should link to https://eol.org/pages/11164788. Abductive (talk) 07:13, 27 November 2018 (UTC)

  • I updated the description. You might want to propose a new property for the new scheme. --- Jura 07:23, 27 November 2018 (UTC)
    • What, keep the old IDs live on Wikidata, and add the new IDs? Abductive (talk) 07:26, 27 November 2018 (UTC)
    • By the way, on how many items is P830 used? Abductive (talk) 07:27, 27 November 2018 (UTC)
  • Yes, for stability's sake, we don't mix the old and the new scheme. Otherwise data users couldn't be sure what they are getting. --- Jura 07:35, 27 November 2018 (UTC)
  • I'm going to leave to pesky details on creating a property and editing 1,375,794 items to you guys. Abductive (talk) 20:11, 27 November 2018 (UTC)
@Abductive: Your claim „EOL has changed all its IDs and URL formats” is wrong.
@Jura: formatter URL (P1630) has to be changed from http://eol.org/pages/$1/overview to https://eol.org/pages/$1. That solved the general problem.
@Abductive: EOL ids refer to taxon concepts and are a subject to change. In the past we had EOL ids that refered to internally used ids and caused similar errors. Maybe references to The Plant List 1.1 (Q15628808) got new ids. Not sure at the moment.
--Succu (talk) 20:15, 27 November 2018 (UTC)
In my example the ID is different. That may not always be the case? Anyways, I have already stated that this is not my forte. It's up to you guys to fix or ignore link rot (Q1193907) as you see fit. Abductive (talk) 20:19, 27 November 2018 (UTC)
Sorry, but rechecking the known 2,237,550 EOL ids against our taxa has a low priority on my todo list. --Succu (talk) 20:37, 27 November 2018 (UTC)
Presumably this can be automated? Perhaps by another with the requisite skill set? Abductive (talk) 05:40, 28 November 2018 (UTC)
Regarding your example: Added by User:BotNinja (= User:Termininja). --Succu (talk) 20:56, 27 November 2018 (UTC)
Okay, so you don't want to fix it either. I completely understand; this is a volunteer project and nobody knows how to write a program to fix these things. Abductive (talk) 02:25, 4 December 2018 (UTC)
You are wrong and as I told you above this has a low priority on my todo list. --Succu (talk) 22:15, 10 December 2018 (UTC)
So, you are the only person here capable of doing it? Abductive (talk) 18:40, 12 December 2018 (UTC)
Out of around 200,000 items having both PlantList-ID (P1070) and Encyclopedia of Life ID (P830) around 10,000 EOL ids are wrong now. I will correct them within the next days. --Succu (talk) 19:35, 12 December 2018 (UTC)

GlobalFactSyncRE/DBpedia project proposal[edit]

DBpedia, which frequently crawls and analyses over 120 Wikipedia language editions, has near complete information about (1) which facts are in infoboxes across all Wikipedias, and (2) where Wikidata is already used in those infoboxes. GlobalFactSyncRE will extract all infobox facts and their references to produce a tool for Wikipedia editors that detects and displays differences across infobox facts in an intelligent way to help sync infoboxes between languages and/or Wikidata. The extracted references will also be used to enhance Wikidata. For more see meta:Grants talk:Project/DBpedia/GlobalFactSyncRE

Please let us know what you think, your opinion is important to us! Thank you!  – The preceding unsigned comment was added by 212.27.205.232 (talk • contribs) at 13:38, 30 November 2018‎ (UTC).

Heads-up: Upcoming GeneDB item creation[edit]

The good folks at GeneDB (Q5531047) are the authoritative source for gene annotation data for about 50 species. They are preparing a new web presence, and they would like to base some of it on Wikidata! As part of this, I have test-imported two of their species genes, and accompanying proteins, as items into Wikidata a while ago, and made a prototype browser for that information. Now that proof-of-concept is established, they would like me to go ahead and import their entire database as Wikidata items. That's about 770K genes, and about as many proteins, so we are talking about ~1.5M new items for this, over the next 1-3 months. All items will have a GeneDB ID (P3382) statement, and various referenced statements about genomic location, taxon, orthologs/paralogs, and function (example items: PPPK-DHPS (Q18970312) and hydroxymethyldihydropterin pyrophosphokinase-dihydropteroate synthase (Q56565045)).

We already have complete genes and proteins for several species (human, mouse, various bacteria), and I believe this will be a real boost, both in network effects (think main subject (P921) for publications), and in establishing Wikidata as a serious, reliable cornerstone of modern science and research. --Magnus Manske (talk) 09:56, 5 December 2018 (UTC)

  • Symbol support vote.svg Support This is great. ChristianKl❫ 11:39, 5 December 2018 (UTC)
  • Symbol support vote.svg Support Sounds good to me. Have you talked to the User:ProteinBoxBot people to make sure you're not both importing the same things? Are these items that GeneDB (Q5531047) already assert are not yet in Wikidata? ArthurPSmith (talk) 15:51, 5 December 2018 (UTC)
  • Symbol support vote.svg Support pretty please can someone write a case study for this project? I think it would be super useful to help other organisations understand how Wikidata would be useful for them. I'm very happy to help with a write up John Cummings (talk) 16:15, 5 December 2018 (UTC)
  • Pictogram voting question.svg Question Why don't they set up their own wikibase instance and federate it with Wikidata (when that is possible)?--Micru (talk) 16:30, 5 December 2018 (UTC)
    • I don't have a strong enough concept of the standards for inclusion here to have a strong opinion, but I share Micru's question. Not as an objection, but out of curiosity, to know how more involved Wikidatans tend to view such decisions. -Pete F (talk) 23:07, 5 December 2018 (UTC)
      • They are looking into that possibility, but despite the Docker containers available, this is still early stage technology, and "federation" doesn't really work on wikibase level yet (SPARQL, perhaps). If they set up their own wikibase, they would likely do so to track additional, detailed data that would be unsuitable here. --Magnus Manske (talk) 15:02, 6 December 2018 (UTC)
  • Symbol support vote.svg Support Yay! Thanks! --Denny (talk) 22:20, 5 December 2018 (UTC)
  • How would updates happen? --- Jura 08:27, 8 December 2018 (UTC)

Label when adding a wikipedia page[edit]

Can someone explain me why the title of a wiki page is not automatically used for the label of the item when a wikipedia page is added in that language from the wikipedia? It is this way by default if you create an item, but not if you only add a link to a wikipedia page. Why is that behaviour of the Wikipedia tool? Thanks in advance, Paucabot (talk) 10:53, 6 December 2018 (UTC)

As I recall (and it has been a while), this is because of the Wikipedia problem with disambiguation of names, so for people and places (the bulk of early Wikidata items) you would have "Saint Mary's Church, (XXX)" where the XXX is for location, which is generally placed in the description, not the label. That is just one example, but there are many more. Nowadays it might be worth reinvestiging this decision. Jane023 (talk) 11:57, 6 December 2018 (UTC)
Is it a technical problem? Because I think it is worse to have no label than to have misplaced parentheses ... Paucabot (talk) 12:33, 6 December 2018 (UTC)
It's not a technical problem. It's just that the code to enforce/undertake addition of a label based on the sitelink is not in place, because no-one has thought it a good idea to make it so, largely for the reasons set out by Jane023. I have some sympathy for your view, but equally sympathy for the notion that labels are important enough that we don't fudge them. Next, your idea /might/ sort out a single or a small set of language label values, albeit at the cost of the addition of suboptimal labels, but it's not applicable across the board: the Thai label taken from a Thai article is not much use for the English langauge label. Given there are ~200 languages wanting labels, the solution doesn't go very far. --Tagishsimon (talk) 13:01, 6 December 2018 (UTC)
@Tagishsimon: I'm not asking for the label to be automatically set in every language, only for one language: if I'm adding a catalan sitelink, then this text would be automatically used for the label in catalan only. Paucabot (talk) 19:30, 6 December 2018 (UTC)

I created a ticket for a similar idea at phab:T148762 a couple of years ago, where the label would not be set automatically, but where the user is given the choice to set the label for the connected Wikidata item.

Unfortunately, nothing has come out of it yet, but if more people show their interest in features like this, maybe we could finally get something implemented. --Njardarlogar (talk) 16:40, 6 December 2018 (UTC)

Thanks, @Njardarlogar:. I'll ask it there. Paucabot (talk) 19:35, 6 December 2018 (UTC)

I have seen that there is a bot (@MatSuBot:) that does this kind of work: Special:diff/800581443. If it's done with a bot I don't see any reason why not do it automatically from the beginning. Paucabot (talk) 08:48, 7 December 2018 (UTC)

@Paucabot: I don't agree that a strictly automatic solution is good, because then a lot of labels would get wrong capitalization, and the user would have to go over to Wikidata to correct the labels, anyway, if they want the labels to be correct. Automatic labels should be better than the current state, where no labels are added at all (with the exception of the work of bots); but you should be able to adjust the label in the dialogue where you connect the page to item, such that you can ensure that the capitalization is correct and that no disambiguation is present in the label. --Njardarlogar (talk) 11:16, 8 December 2018 (UTC)
@Njardarlogar: If, anyway, I have to go to Wikidata to get things done, I prefer only having to change capitalization or removing disambiguations than having to add everything from scratch. Maybe even we could get rid of the parentheses automatically ...
Have you seen T58410? They have discarded offering to change the label as an option. Having done so, I think that a strictly automatic solution is better than the current situation. Paucabot (talk) 11:26, 8 December 2018 (UTC)
Having corrected lots of item labels (for various reasons different from Wikipedia titles) I think it's not a good idea to do this automatically, unless it is offered as a yes/no pop-up option when the sitelink is created. Then it is up to the user to have the whole title copied into the label or not. Even this may be a bad idea, I am not too sure how often "Wikipedians with zero Wikidata experience" actually add sitelinks by hand. Jane023 (talk) 11:56, 8 December 2018 (UTC)
@Jane023: Maybe you're right, but this behaviour is not very coherent with the actual behaviour of the tool if the item does not exist (it copies the sitelink to the label) or with the existence of bots that do exactly this work. Paucabot (talk) 12:19, 8 December 2018 (UTC)
@Paucabot: In the ticket I created (which is still open), I actually address the concern about casual users mentioned by hoo (see my reply to Edgars2007): the option for adjusting the label could be under an "advanced" section, or even completely hidden unless the user opts in on the feature in the user settings. I'd also add that connecting Wikipedia pages in different languages is already some distance into advanced users territory, so the addition of an "advanced" section should not be very problematic. If you are really sceptical about such a section and think it will confuse inexperienced users, you could even have it appear only after the user has connected 10 pages.
Regarding the concern about making the linkItem code more complicated, I don't know how important that is; but I doubt it would/should be that much more complicated to e.g. make it launch a new dialogue with its own code that allows the user to set the label after the page has been connected to the item. Again,such a dialogue could not show at all by default, or the user could choose to have the dialogue never appear again.
In short: if there is a will to implement such a feature, there is almost certainly a way. :-) --Njardarlogar (talk) 12:11, 8 December 2018 (UTC)
Thanks for your thoughts, @Njardarlogar:. Paucabot (talk) 12:15, 8 December 2018 (UTC)
How about having a popup in case a new Wikilink gets entered in a language that doesn't already have a label that suggests adding the title as a label but leaving the user the option to adapt it as well? ChristianKl❫ 13:05, 10 December 2018 (UTC)
That would be fine for me, @ChristianKl:. Paucabot (talk) 19:54, 10 December 2018 (UTC)
@Paucabot: I'm also very interested in a (semi-)automatic solution. Speaking for North Frisian (frr:) --Murma174 (talk) 12:28, 15 December 2018 (UTC)

References not to displayed unfolded[edit]

Is there a way the references not to displayed unfolded when there is a constraint violation within the reference? It really makes my constributions difficult. Xaris333 (talk) 10:34, 9 December 2018 (UTC)

Hello,
Displaying the reference unfolded when there is a constraint violation is done on purpose: this way, we encourage the community to fix either the content or the constraint. However, if many people complain about it and wish for a way to fold the references, we're ready to think about it further. I hope this answer helps you. Lea Lacroix (WMDE) (talk) 11:48, 10 December 2018 (UTC)
@Lea Lacroix (WMDE): I understand why that was applied. But for me it is very difficult. Because of that problem Wikidata:Project chat/Archive/2018/11#single-value constraint with title property, in many items that I am adding properties, the references are display unfolded. Xaris333 (talk) 16:10, 10 December 2018 (UTC)
I think we really need to clean up the constraint then. I know this sucks but it's better to fix the underlying issue than glossing over it :/ --Lydia Pintscher (WMDE) (talk) 09:45, 11 December 2018 (UTC)
@Lea Lacroix (WMDE): How to clean up? The references are correct. The constraint is correct. @Jura1: Xaris333 (talk) 19:35, 11 December 2018 (UTC)
It seems a bit of an overkill to add the title twice at Q6377735#P1082, but not at all the item about the source itself Cyprus census 1976 (Q29639032). If you use "stated in", I don't think the reference should even include title. Maybe Help:Sources needs to clarify this. --- Jura 19:49, 11 December 2018 (UTC)
Some references have titles in more than one language. Is not a solution just not to add them. Xaris333 (talk) 20:30, 11 December 2018 (UTC)
Can you give a couple of samples? --- Jura 09:27, 12 December 2018 (UTC)
Wikidata:Project chat/Archive/2018/04#Title Xaris333 (talk) 20:56, 12 December 2018 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Thanks. Looks like a non-issue then. I found three samples there:

  1. Q6377735 with 262 in population. It was discussed above. The titles were incorrectly added in references instead of the item about the source (Q29639013). If you move them there it should take care of it.
  2. Q25991162 and Q25991163 which don't have the problem as the title (P1476) is used as main value.

That said, maybe non-mandatory constraint violations don't need to have the references displayed unfolded. --- Jura 08:39, 13 December 2018 (UTC)

And how that can be done? How the non-mandatory constraint violations don't need to have the references displayed unfolded? (I disaggreei that the titles were incorreclty added). Xaris333 (talk) 15:02, 13 December 2018 (UTC)

  • It would need to be a development, but I'd rather see a couple of samples that match Help:Sources. --- Jura 18:48, 13 December 2018 (UTC)

@Lea Lacroix (WMDE):, @Jura1: Problem still exist, and no solution. The changes make my contributions difficult and no solution... There is no respect to users that constribute to wikipedia. It was supposed that the changes would make our work easiest, no more difficult... Xaris333 (talk) 17:31, 14 December 2018 (UTC)

I feel like there is less an issue with the software than an issue with the definition of the constraint. I can't intervene on content discussion. And changing the software is not the best way to solve content issues. However, I can create a ticket asking to keep references with non-mandatory constraints folded, if this is what you request. Lea Lacroix (WMDE) (talk) 17:51, 14 December 2018 (UTC)

Q11263220 and Dumbing down (Q5313720) - are they the same and should they be merged?[edit]

Q11263220 refers to policies like Fado, Football, and Fatima (Q3539497) or 3S Policy (Q10854388) or Bread and circuses (Q845658) that make people either unable to or doesn't want to care about politics, usually for the purpose of maintaining or strengthening the control of the ruling government on a country/society. Can it be said as equal to dumbing down and thus can the two wikidata items be merged? C933103 (talk) 16:30, 9 December 2018 (UTC)

  • I think you want said to be the same as (P460) here, not a merge. - Jmabel (talk) 22:41, 9 December 2018 (UTC)
    • I see. Then, how should Q11263220 be named in English? C933103 (talk) 01:42, 10 December 2018 (UTC)
      • It seems close to George Orwell's prolefeed (Q17104684). --EncycloPetey (talk) 05:10, 10 December 2018 (UTC)
        • Close to but not close enough to be use as translation, especially one is fictional, and then also the English term was used to refer to the media while the wikidata item describe a policy.... C933103 (talk) 19:28, 13 December 2018 (UTC)

Miniature map for coordinate properties[edit]

Hello all,

After we deployed the Commons miniature pictures for the image properties, some editors suggested to do something similar with maps. From Wednesday, December 12th, we will also have a miniature map displayed for the properties containing coordinates, like coordinate location (P625). This feature uses mw:Extension:Kartographer and OpenStreetMap.

Viewing the map directly in the item will help the editors to check quickly if the coordinates seem correct, and therefore improve the data quality.

Some useful features:

  • Double-click on the miniature map: display it in full screen
  • On full screen view: click on “external maps” to display different map services and links to many external maps
  • In edit mode: when adding or editing coordinates in the field, the maps is automatically updated to fit the coordinates you entered
  • The coordinates are still displayed under the map and link to Geohack if you have the gadget AuthorityControl enabled

If you have any questions or issues with this feature, let me know. You can also see the related ticket. Lea Lacroix (WMDE) (talk) 15:22, 10 December 2018 (UTC)

@Lea Lacroix (WMDE): Very nice! Will this display for all coordinates, or only for coordinates with globe=Earth (Q2) (the latter is better)? Thanks. Mike Peel (talk) 16:41, 12 December 2018 (UTC)
Ah, looks like it displays for all coordinates. I'll ask about this on phabricator. Thanks. Mike Peel (talk) 16:43, 12 December 2018 (UTC)
@Mike Peel: No this is just for coordinates on earth, others are exempted (see for example Q499158#P625). Cheers, Hoo man (talk) 16:48, 12 December 2018 (UTC)
Aah, I was confusing it with User:Aude/mapview.js in those cases, thanks for the clarification. I've made a couple of other suggestions on the phabricator ticket anyway. Thanks. Mike Peel (talk) 16:50, 12 December 2018 (UTC)

Linking musical settings with source texts[edit]

What are the correct properties to use when linking a source text with a musical setting of that source text? and when linking an original melody with a set of alternate lyrics written to that tune at a later date? In these cases the melody and the lyrics will be separate items. Sorry if this has been asked before. Beleg Tâl (talk) 15:27, 10 December 2018 (UTC)

I've found has melody (P1625) which looks like it addresses some of what I was looking for... is there a reverse property for back-linking? Is it appropriate to use has melody (P1625) if the musical setting comprises more than only melody? Beleg Tâl (talk) 13:58, 13 December 2018 (UTC)

Wikidata weekly summary #342[edit]

Request for filling out a questionnaire on Wikidata[edit]

Dear Wikidata enthusiasts, I have a huge favor to ask of everyone reading this --

TLDR: Please fill out the questionnaire linked here!

And now in more details: As some of you may know, I'm PhD candidate at the School of Education, Tel Aviv University, and my research focuses on Wikidata as a learning platform. To succeed in my research, I need lots of data about how users (such as yourselves) interact with Wikidata, so it would be of great help if you could take the time to fill out the questionnaire linked above. It's estimated that it should take between 15-30 minutes, depending on how detailed your answers would be. I'm striving for at least 100 replies, but this is one of those cases of "the more, the merrier", so really, every single person filling it out is of huge help.

If you have any questions, or are willing to participate in a follow up interview, please feel free to ping me via my talk page, privately or by email (shani.even at gmail.com). Thanks in advance for considering filling it out. I will be forever grateful to anyone who can help and promise to update you on my progress. :) Shani Evenstein (talk) 22:46, 10 December 2018 (UTC)

Property:P443[edit]

How about to move all instances of this property from items to appropriate Lexemes? (Only a word can have pronunciation, not a notion) --Infovarius (talk) 14:41, 11 December 2018 (UTC)

Yes, that sounds like a worthy project. ChristianKl❫ 15:12, 11 December 2018 (UTC)
Some items are about words. For example: Donau­dampf­schiffahrts­elektrizitäten­haupt­betriebs­werkbau­unter­beamten­gesellschaft (Q20167), fuck (Q31928). (And if the items have sitelinks, we can’t delete them.) --Lucas Werkmeister (talk) 17:18, 11 December 2018 (UTC)

How do I describe population of a species[edit]

Hi all

I'd like to add the population of African elephants to African bush elephant (Q36557), the only thing I can find is population (P1082) which is specific to humans, are we missing a property or am I just missing the magic words?

Thanks

--John Cummings (talk) 17:26, 11 December 2018 (UTC)

@John Cummings: I think its uncommon to see items about species with their amount of individuals. I've searched on Special:ListProperties and I found nothing. Esteban16 (talk) 20:12, 11 December 2018 (UTC)
This was asked recently at Wikidata:Project_chat/Archive/2018/10#Number_of_peoples_or_animals_among_a_nationality_/_population_/_breed. How to model this appears to be unresolved. (A solution to this would also need a way to indicate species populations within a particular area at particular times.) --Yair rand (talk) 20:16, 11 December 2018 (UTC)
@John Cummings: quantity (P1114) View with SQID. Taxons are classes of individuals and number is supposed to give the number of instances, hence individuals, of that class. author  TomT0m / talk page 20:30, 11 December 2018 (UTC)
@TomT0m:, thanks, could you give an example item where this has been used? Quantity seems like a strangely worded way of saying it. --John Cummings (talk) 23:52, 11 December 2018 (UTC)
@John Cummings: Maybe the english label is just wrong because if I remember correctly the proposal time, it’s the exact purpose of the proposal. author  TomT0m / talk page 07:48, 12 December 2018 (UTC)
Thanks @TomT0m:, I tried to find the conversation about the property proposal which describes how the property can be used but I can't find it, any ideas? --John Cummings (talk) 08:24, 13 December 2018 (UTC)
Indeed, it’s a weird situation, the link provided in the property statements is https://www.wikidata.org/wiki/Wikidata:Property_proposal/Archive/19#P1114 . The creation is acknoledged but it seems the discussion is lost. @Zolo: Any idea what happenned ?
Thanks to wikiblame we can find the initial proposal : https://www.wikidata.org/w/index.php?title=Wikidata:Property_proposal/Generic&diff=prev&oldid=76584375&diffmode=source
(I was pinged for the creation it seems but I can’t find any trace of the interventions I had in this. I’m pretty sure I participated though, weird. Old times.) author  TomT0m / talk page 12:16, 13 December 2018 (UTC)

Sry, TomT0m taxon (Q16521) and population (Q2625603) are not about the same concept. I assume the question is about how many individual (Q795052) we have on earth at a certain point in time, John Cummings? --Succu (talk) 20:38, 13 December 2018 (UTC)

Succu I know, but John points on a taxon item, not a population one. The property he points to has quantity datatype, so it’s not supposed to link taxon items to populations of this taxon items. So you’re correct about the meaning of the property. I pointed John to a property that is supposed to store exactly that. author  TomT0m / talk page 20:44, 13 December 2018 (UTC)
Sorry, but as far as I'm aware there are several disjoint populations of African bush elephant (Q36557) out there... --Succu (talk) 20:51, 13 December 2018 (UTC)
I had the same question but it did not seem that it was what John wanted. Anyway that’s interesting, do we have a model for this ? author  TomT0m / talk page 20:57, 13 December 2018 (UTC)

Hi @TomT0m: and @Succu:, maybe it would be helpful to give you the source rather than me trying to explain what I'm trying to do, I want to add the fact there are 415,000 African elephants. Thanks, --John Cummings (talk) 09:26, 14 December 2018 (UTC)

@John Cummings: as suggested by TomT0m and Yair rand : (if needed, with qualifier and reference) is probably the best way to go. For an example, see Gavialis gangeticus (Q17134). Cheers, VIGNERON (talk) 16:23, 14 December 2018 (UTC)
Thanks @VIGNERON:, it looks like Gavialis gangeticus (Q17134) displays the same information as both population and quantity, I guess the population statements should be removed? My issue with using quantity as a phrase is that population is the correct terminology for number of individuals of a species. --John Cummings (talk) 18:01, 14 December 2018 (UTC)
@John Cummings: The english label of the property has changed, see the discussion on Property_talk:P1114. You might want to reopen the topic of the english label, but the current naming should not be a blocker. author  TomT0m / talk page 11:33, 15 December 2018 (UTC)

Are Amazonia (Q2841453) and Amazon rainforest (Q177567) the same thing[edit]

Hi all

I think that Amazonia (Q2841453) and Amazon rainforest (Q177567) are possibly the same thing? Could someone have a look?

Thanks

--John Cummings (talk) 17:34, 11 December 2018 (UTC)

In french, one is about a (geographical region), the other about the forest that is located on that region. If the second disappears, the first will still be there, so I think they are different entities. author  TomT0m / talk page 17:55, 11 December 2018 (UTC)
Thanks @TomT0m:, --John Cummings (talk) 18:09, 11 December 2018 (UTC)

Using qualifier on identifiers for type of record[edit]

In the US National Archives catalog, the same identifier set is used for all types of records, including both authority records and descriptions of archival holdings. This is nice, because the number is the only thing a resource you need to know to refer to it. However, I'd like to be able to distinguish the record types on Wikidata, so that we can perform queries on only a single type of record—e.g., see all items with NARA authority IDs. Does anyone have opinions on the best way to do this? The only idea I have is to add it as a qualifier of the identifier, like the example I have done at Q59296199#identifiers. However, I have not seen qualifiers used for external identifier statements before, and I am not sure if object has role (P3831) is the best choice for it, so I was wanted to see what others thought of that before applying it to thousands of items. Thanks for any help! Dominic (talk) 22:00, 11 December 2018 (UTC)

Could we use type of reference (P3865) for this, with a constraint change to allow it to be used as a qualifier for an identifier? - PKM (talk) 20:15, 12 December 2018 (UTC)

Bug fix - references are now accessible for non-logged in users on protected items[edit]

Hello all,

Usually I don't announce every small bug fixes that we make, but I thought this one could be interesting for you to know - if you hear someone complaining about this bug, you can tell them it's fixed now :)

Before the fix, on protected or semi-protected items, non-logged in users could not display the references. The references section was folded, the link "X references" was grey and non-clickable, meaning that readers without an account couldn't display or access the references.

This bug is now fixed, non-logged in users can access the references on all items. You can test it quickly by opening Douglas Adams (Q42) in a private session of your browser. If you encounter any issue, feel free to add a comment in the ticket or ping me.

Cheers, Lea Lacroix (WMDE) (talk) 13:47, 13 December 2018 (UTC)


Model item (Property:P5869)[edit]

Just wondering, which type of entities should this property be on?

It was originally proposed for items (see samples on Wikidata:Property_proposal/Model_item), but can now also be found on a property (Property:P747#P5869).

On properties, to some extent it duplicates:

Supposedly, the still proposed inverse property for model item (P5869) couldn't work when using it on properties (Wikidata:Property proposal/Model item for).

Also, if selection is mainly done by the number of statements on an item, the query links on the property talk page are better than a static link to one item. --- Jura 14:12, 13 December 2018 (UTC)

The recent court judgement for Q48928408[edit]

The High Court of Hong Kong has made a judgement on 13 December 2018. According to paragraph 64 of the Judgement, Judge Anderson Chow said:

It is not in dispute that the West Kowloon Station Mainland Port Area (Q48928408) falls with the territory of Hong Kong under the “Order of the State Council of the People’s Republic of China No 221” dated 1 July 1997, which was promulgated in accordance with the “Decision of the National People’s Congress on the Establishment of the Hong Kong Special Administrative Region” adopted at the Third Session of the Seventh National People’s Congress on 4 April 1990.

That means the correct statement is West Kowloon Station Mainland Port Area (Q48928408)located in the administrative territorial entity (P131)  Hong Kong (Q8646)!!!! 223.17.64.233 14:31, 13 December 2018 (UTC)

Actually, the document you linked says "Mainland Port Area", not "Kowloon Mainland Port Area". Anyway, I do not know anything of this. --Ogoorcs (talk) 05:16, 14 December 2018 (UTC)
Maybe you should spend some time to read the Judgement. 223.17.64.233 12:40, 14 December 2018 (UTC)

Use of P642[edit]

In your opinion, is this use of of (P642) a good thing? Visite fortuitement prolongée (talk) 21:18, 13 December 2018 (UTC)

  • Seems reasonable to me. - Jmabel (talk) 23:24, 13 December 2018 (UTC)
  • There is maybe a better property but of (P642) seems good enough. Cheers, VIGNERON (talk) 11:42, 14 December 2018 (UTC)

Where does Wikidata's data come from?[edit]

Sincere apologies if this is answered somewhere, I haven't found anything.

I know the general answer is "lots of places;" individual and institutional donors via dataset imports. But I'm curious about Wikidata's relationship to Wikipedia. Is Wikidata also populated by Wikipedia scrapers?

The reason I ask is that I've been trying to build a database of events (as in moments or spans in history) by querying for any entities with any of the following props: point in time (P585),start time (P580),date of birth (P569),inception (P571). (I assume there are others that would be applicable too but this is what I have so far)

So then I get some thousands of results that vary radically in quality (not surprising). I can filter those down to, say, entities that have titles (quite a few do not, apparently) and some other conditions.

I end up with a bunch of event entities, but I've noticed some surprising/baffling trends in this data. Other than the assorted bits and pieces you'd expect, there are HUNDREDS upon hundreds of entities like this:

All-Ireland Senior Hurling Championship 1969 1965 Scottish League Cup Final 1964 Malaysia Open Badminton Championships – Men's doubles

And hundreds upon hundreds of entities like this:

Los Angeles mayoral election, 1969 Honduran Constituent Assembly election, 1965 Tuscan regional election, 1970

From what I can tell, these two categories of sporting events and elections outnumber all the other data by a large margin. So I assume these patterns are due to data dumps from, say, a sporting history database and, say, an election database.

I've never seen any kind of provenance field for an entity, so I'm not sure if the source of the data is recorded anywhere. But ultimately, I'm just trying to build a massive world event database and hopefully contribute back to Wikidata. It's looking like I will need to write my own Wikipedia scraper(s) to get more data than is currently in Wikidata—does that sound right?  – The preceding unsigned comment was added by 0x73DB07 (talk • contribs).

Your pattern assumption is wrong. All of the titles you picked out are wikidata items created to mirror English wikipedia articles - e.g. All-Ireland Senior Hurling Championship 1969 (Q4728160), 1965 Scottish League Cup Final (Q4484641) - the item records provide a link to the wikipedia article. Some statements on both items have references specifying that the data was imported from En.Wikipedia. In general a) the vast majority of wikipedia articles have corresponding wikidata items but b) varying and generally limited amounts of statements on items have been derived from the wikipedia article. There are tools (e.g. this, these) designed to scrape information from wikipedia records and apply to wikidata items. As far too many wikidata items lack core statements such as dates, the probability is more information can be scraped from the linked wikipedia articles, though clearly this is a non-trivial task. --Tagishsimon (talk) 02:35, 14 December 2018 (UTC)
  • "Hundreds of items" means in Wikidata, that there are relatively few items of a given class. If you click on "Random item" a few times you will see that Wikidata has a lot more items for places and scientific articles then it has for events,
While unfortunately not all claims on Wikidata have provenance information a lot of them do. You find it under the references of an claim. If you for example look at those items referenced by Tagishsimon you will find that the reference contain information that certain claims come from Wikipedia. It's also useful to look at the history of an article to see how the information got into the article if you are interested in that question. ChristianKl❫ 11:17, 14 December 2018 (UTC)

Warning about Wikidata property example (P1855)[edit]

Lucas Werkmeister (WMDE) – work account, mainly for development discussions
Jarekt - mostly interested in properties related to Commons
MisterSynergy
John Samuel
Sannita
Yair rand
Jon Harald Søby
Pasleim
Jura
PKM
ChristianKl
Sjoerddebruin
Salgo60
Fralambert
Manu1400
Was a bee
Malore
Ivanhercaz
Peter F. Patel-Schneider
Pizza1016
Ogoorcs
CennoxX
Pictogram voting comment.svg Notified participants of WikiProject property constraints

Currently, Wikidata property example is mandatory for every property but it is suitable only for properties whose datatype is "item". For properties with a different datatype there are Wikidata property example for properties, Wikidata property example for lexemes, Wikidata property example for forms and Wikidata property example for senses.

How to replace the current constraint with a better one?--Malore (talk) 03:03, 14 December 2018 (UTC)

symmetric constraint (Q21510862) works only for properties with an "item" datatype[edit]

Lucas Werkmeister (WMDE) – work account, mainly for development discussions
Jarekt - mostly interested in properties related to Commons
MisterSynergy
John Samuel
Sannita
Yair rand
Jon Harald Søby
Pasleim
Jura
PKM
ChristianKl
Sjoerddebruin
Salgo60
Fralambert
Manu1400
Was a bee
Malore
Ivanhercaz
Peter F. Patel-Schneider
Pizza1016
Ogoorcs
CennoxX
Pictogram voting comment.svg Notified participants of WikiProject property constraints

I think it should be possible to use the constraint also for properties with datatypes other than "item"--Malore (talk) 03:38, 14 December 2018 (UTC)

@Malore: do you have an example? I don't see how it can work with something else than items. Cheers, VIGNERON (talk) 09:38, 14 December 2018 (UTC)
@VIGNERON: it should work with other entities that can have statements, such as properties, lexemes, senses and forms. @Malore: I added such a constraint to Sandbox-Lexeme (P5188) and am now seeing constraint violations on Lexeme:L123#P5188, so it seems to be working as far as I can tell? (The documentation may need to be updated in places to reflect this, as JakobVoss did in this edit.) --Lucas Werkmeister (WMDE) (talk) 11:30, 14 December 2018 (UTC)
Oh yes, obviously, Danke Lucas Werkmeister (WMDE)! Cheers, VIGNERON (talk) 11:32, 14 December 2018 (UTC)
@Lucas Werkmeister (WMDE): It works but in the property talk it says constraint “symmetric constraint (Q21510862)” declaration error: “the constraint is not applicable to datatype “Lexeme””.--Malore (talk) 15:21, 14 December 2018 (UTC)
@Malore: okay, that’s a problem with Module:Constraints (specifically cItemOrPropertyonlyConstraints should probably be something like cEntityOnlyConstraints, with a corresponding change in verifyPropertyDatatype). --Lucas Werkmeister (WMDE) (talk) 16:15, 14 December 2018 (UTC)

Adding edit summaries[edit]

I've edited English Wikipedia for years and am used to adding an edit summary to each edit I make there. Here on Wikidata I haven't found a way to do that when editing the contents of an item. Am I missing a trick somewhere? DeFacto (talk) 09:30, 14 December 2018 (UTC)

It's not possible on Wikidata. I think this design design decision might be due to Wikidata being designed to be multilingual. ChristianKl❫ 11:03, 14 December 2018 (UTC)
AFAIK, it's only possible to give an edit summary when adding new data if you use the API, rather than the web interface. Most of the time there shouldn't really be a need for a summary though, as it's fairly obvious what you're doing (especially if you've provided a reference). However, if you want to correct existing data, one way to do that is to undo the change that created the current data, before adding a new value. The undo allows you to leave a summary message, and also notifies the person who originally made that addition. This is sometimes seen as a quite aggressive thing to do if the change is minimal or additive (e.g. you're changing a year-precision date to day-precision), but if you believe the original data is actively wrong, then it's generally seen as a good way of both providing more context, and letting the original editor know about it. --Oravrattas (talk) 11:09, 14 December 2018 (UTC)
Please do not add custom edit summaries. Wikidata is not only edited by English users. Wikidata is adding automated summaries which are translatable. If you add custom summaries they are not translated and you penalize people from other language regions. 130.92.162.106 12:56, 14 December 2018 (UTC)
I would like to see "destructive" edits have an edit summary. That way we know why someone is changing data already in existence. Lazypub (talk)

Car series[edit]

As we probably know, car manufacturers have a tendency to re-use car model names for successive (often very different) models in the same sector. For example we have Jaguar XJ (X308) (Q1678516), Jaguar XJ (X350) (Q1678522) and Jaguar XJ (X351) (Q1678526) which are successive generations of the Jaguar XJ. Note too that the different generations are stated as being instances of automobile model (Q3231690) (which they clearly are) but should they also be instances of Jaguar XJ series (Q1142339), which itself should be an instance of something like "series of car model", rather than of automobile model (Q3231690) too? Any advice or opinion on how this should be structured would be appreciated. DeFacto (talk) 10:01, 14 December 2018 (UTC)

They are not instances of Jaguar XJ series (Q1142339) but subclasses of it. ChristianKl❫ 11:05, 14 December 2018 (UTC)
Ah, ok, and should they remain as an instances of automobile model (Q3231690) too? And what should Jaguar XJ series (Q1142339) be - it is currently also an instance of automobile model (Q3231690). DeFacto (talk) 11:20, 14 December 2018 (UTC)
It's apparently the automobile equivalent to mobile phone series (Q20488450). Ghouston (talk) 23:45, 14 December 2018 (UTC)

ORCID scheduled outage[edit]

The ORCID API and website will be offline for planned maintenance, tomorrow for about 8 hours, from 2pm UTC; see; [1].

In particular, I've asked Magnus to take the relevant part of SourceMD offline during the outage. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:50, 14 December 2018 (UTC)

Let's get serious about norms of behavior[edit]

Hi folks.

I think Wikidata needs clearer, explicit norms of behavior. And I think it needs them now, before it's a huge problem.

Please take a moment to read my draft meta-proposal and use the talk page to share your thoughts. I'd like participation to be as broad as possible, to ensure we get the best outcome for our community. Thanks! Asaf Bartov (talk) 17:27, 14 December 2018 (UTC)

Les Républicains[edit]

Should Union for a Popular Movement (Q173152) and The Republicans (Q20012759) be merged? 46.46.199.130 20:41, 14 December 2018 (UTC)

It was not only renamed Les Républicains, but it is the succesor of Union pour un mouvement populaire. Considering that, I think they shouldn't be merged. Esteban16 (talk) 21:25, 14 December 2018 (UTC)
replaces (P1365) / replaced by (P1366) ? - Jmabel (talk) 22:20, 14 December 2018 (UTC)
They can't be merged as long as they have separate articles on frwiki. Ghouston (talk) 23:32, 14 December 2018 (UTC)
No it shouldn't and can't be merged. Cheers, VIGNERON (talk) 08:43, 15 December 2018 (UTC)

Would love some input on Q59739910 "critically acclaimed"[edit]

We are constantly being told that a book, film, record, this, that or other thing is "critically acclaimed". I would love to try to quantify that a little. Love some input from you guys (is this already covered somehow, perhaps in another language?) and if anyone feels the calling - why not create the opposite "critically panned/panned by critics"😁. Moebeus (talk) 21:10, 14 December 2018 (UTC)

It seems questionably subjective. There are quantifications available like Rotten Tomatoes ratings, but I suppose they are constantly changing. Still, adding a Rotten Tomatoes rating with a point in time seems better than just saying "critically acclaimed". Ghouston (talk) 23:39, 14 December 2018 (UTC)
@Ghouston: While a critic's opinion is absolutely 100% subjective, the fact that the critic rates something highly is not. The point of "critically acclaimed" would be to tag music albums, books, films, etc. that end up on End-of-year lists, "100 Best Movies"-lists, etc. Moebeus (talk) 16:32, 15 December 2018 (UTC)
  • To add Q59739910, I suppose you'd identify some reference. Such a reference would be a list like Q7711726 where you added this. If done, one want might as well add Q7711726 directly and let users decide which critics they want to follow. Some people are fond of one award or the other and add these. --- Jura 16:41, 15 December 2018 (UTC)