Wikidata:Project chat/Archive/2019/10

From Wikidata
Jump to navigation Jump to search

This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion.

Contents

Set rank via API

I want to set the rank of statements via the API. But I do not find an explanation how to do this in the documentation. Can someone help me or has an code example? --GPSLeo (talk) 13:53, 29 September 2019 (UTC)

You can retrieve a claim with wbgetclaims in JSON format, modify it locally, and change it then with wbsetclaim by sending the modified claim serialization. —MisterSynergy (talk) 19:57, 29 September 2019 (UTC)
I think it would be better to use action=wbgetentities&props=claims|info, so that the response also includes a lastrevid; then you can pass this revision ID as the baserevid of action=wbsetclaim, so that the API can detect and possibly patch changes if there were any other edits between the two requests. --Lucas Werkmeister (WMDE) (talk) 12:45, 30 September 2019 (UTC)
Ah, thanks. But that also means that it always needs two edits. One for adding the statement and one for setting the rank for the statement. --GPSLeo (talk) 13:50, 30 September 2019 (UTC)
No, why should that be necessary? If you‘re creating a new statement, you can create it with the correct rank right away. (You can create such a statement manually, e. g. on Wikidata Sandbox (Q4115189), and use the browser dev tools to see what network request is used.) --Lucas Werkmeister (WMDE) (talk) 22:34, 30 September 2019 (UTC)
It would be incredibly useful if somebody could patch Wikidata-CLI to do this, or create a similar tool with a similar interface that can do this. At the moment there doesn't seem to be any tool one can use to alter the rank of a set of statements. Jheald (talk) 23:20, 30 September 2019 (UTC)
+1. We have thousands of items with old population (P1082) numbers in "preferred" rank, because while it's easy for everybody to add new numbers through QuickStatements, it's impossible to change ranks if you don't write your own bot. Ayack (talk) 08:15, 1 October 2019 (UTC)
Pywikibot has a "setRank" function that is pretty easy to use; with PAWS, it is also easy to execute such a bot script. I'd say that ~1 screen of code is probably enough to fix this problem. --MisterSynergy (talk) 08:48, 1 October 2019 (UTC)
@MisterSynergy: Is there some documentation about the "setRank" function? I've found nothing in Manual:Pywikibot. Thanks. Ayack (talk) 09:01, 1 October 2019 (UTC)
Here you can see that it is existing, but no useful documentation is available. I have used it in the past, however, so maybe you can understand e.g. this pywikibot script (look for "getRank" and "changeRank" functions in that particular case; "setRank" should be very similar). --MisterSynergy (talk) 09:10, 1 October 2019 (UTC)
Thanks, I'll have a look at it. Ayack (talk) 09:14, 1 October 2019 (UTC)

Wikidata WikiProject India Newsletter is live!

WikiProject India bars.svg

Hi,

We are very happy to announce that we have started a quarterly newsletter featuring news, updates and plans related to Wikidata activities in India. You can find our first issue here. If you do not want to receive this kind of notification further, you can remove your username from here.

Regards,
Bodhisattwa (CIS-A2K) (talk) Sent through  --MediaWiki message delivery (talk) 07:21, 1 October 2019 (UTC)
(on behalf of WikiProject India.)

Alphabets, writing systems, scripts, typefaces

Is anyone aware of a robust ontology for alphabets, writing systems, scripts, typefaces, and the like, other than/in addition to Getty's AAT? As the linked identifiers on blackletter (Q213686) show, Wikidata (and most Wikipedia articles) have a tendency to conflate a handwritten medieval script and the typefaces based on that script. I believe these need to be disentangled into separate hierarchies.

Because of examples like "blackletter" I think it may be prudent to keep blackletter (Q213686) as "class of medieval scripts and typefaces based on them" with <has parts of the class> [blackletter-script] and [black-letter typeface]. AAT has guide layer in its hierarchy "<script and type forms>" and perhaps this points a way toward a compromise between what's in Wikipedias and Commons and a strict hierarchy that separates handwriting from typefaces. Thoughts? - PKM (talk) 23:32, 1 October 2019 (UTC)

Q7703332: Czech folk music groups VS Czech contemporary folk music groups

Could somebody, especially native English speakers, help to solve the problem with this item? ŠJů (talkcontribslogs) changed its Commons interwiki link from "Category:Folk musical groups from Czechia" to "Category:Contemporary folk music groups from Czechia". I changed it back and talked with ŠJů. After arguing with me, he/she reverted my edit and also changed its English label to "Category:Czech contemporary folk music groups" in spite of its category in the English Wikipedia.

As I understand it, his/her rationale is its Czech category is translated as "Category:Czech contemporary folk music groups" in English, therefore he/she believes that its English and Commons categories should be called like that as well. I doubt it can be called a solution, it looks like ŠJů imposes his/her personal views without a thorough discussion with the English-speaking community.

Also, I ask Wikidata administrators to confirm his/her actions or revert it back. Thanks in advance.--Russian Rocky (talk) 00:47, 29 September 2019 (UTC)

As it was clearly explained in the previous discussion, the problem is that English Wikipedia in its category tree en:Category:Folk music groups by nationality mixes together the Anglo-Saxon meaning ot "folk music" (= folk music (Q235858) - national music, folklore music) with the European meaning (= contemporary folk music (Q43343), contemporary folk music). As I can see, Russian Wikipedia uses also the European terms (Russian article ru:Фолк-музыка is about contemporary folk, ru:Народная музыка about (traditional) folk), i.e. the core should be understandable even for Russian Rocky, if he was able to perceive facts and arguments from the discussion. As explained in the previous discussion, Category:Czech folk music groups (Q7703332) was created, intended and used just for categories which collect music groups of folk style in the European meaning, i. e. for the Czech en:Category:Contemporary folk music groups. For the traditional-folk groups, we have the item Category:Folk groups from Czechia (Q9232134), which has not special "musical" subcategory yet, because most of the groups combine music, dance and other folklore traditions together. However, Category:Czech folk music groups (Q7703332) falls clearly under Category:Contemporary folk music groups (Q8462878), not (directly) under Category:Folk music groups (Q55958741). The fact that en:wiki misses Category:Contemporary folk music groups by country (Q61983681) within Category:Contemporary folk music groups (Q8462878) and doesn't distinguish this "by country" structure from Category:Folk music groups by nationality (Q8462879) is a problem of en:wiki only. Wikidata an Commons have this problem clearly solved in principle (even though some individual interwiki links or national labels can be incorrect or mistaken). --ŠJů (talk) 14:12, 29 September 2019 (UTC)
ŠJů, the problem is that your renaming practices don't make much sense and based only on your whim. If everybody starts to rename Wikidata categories despite their names in the English Wikipedia, it'll turn Wikidata into a mess.
The category is called en:Category:Czech folk music groups, period. Nominate it and prove your point in the English Wikipedia, rather than forcibly renaming it in Wikidata on weak excuses. That's the message.--Russian Rocky (talk) 11:06, 30 September 2019 (UTC)
Please try to express your reaction to discussed facts, arguments and problems, if you are able understand what's going on. --ŠJů (talk) 12:11, 30 September 2019 (UTC)
You're unable to properly explain why you rename categories in Wikidata despite their names in the English Wikipedia. Your "explanation" isn't really an explanation, but a weak excuse. It emboldens other people to rename English categories in Wikidata proceeding only from their own views. That's the problem, but you turn a blind eye to the dangerous precedent.--Russian Rocky (talk) 12:43, 30 September 2019 (UTC)
It seems to me that ŠJů explanation makes sense. Words frequently have multiple meanings and at Wikidata we care about linking items together that mean the same thing instead of items that are named the same. ChristianKl❫ 15:03, 30 September 2019 (UTC)
@ChristianKl: it seems you didn't get my point. I'm not against linking the Czech and English categories together. The problem is ŠJů arbitrarily renames English categories despite their names in the English Wikipedia. As you can notice, the category is called Category:Czech folk music groups in the English Wikipedia, but ŠJů renamed it to Category:Czech contemporary folk music groups in Wikidata. It's confusing and makes searching difficult.--Russian Rocky (talk) 17:20, 30 September 2019 (UTC)
There's nothing arbitary about finding names on Wikidata that are intend to reflect what a category is about instead of names that match external names 1-to-1. ChristianKl❫ 07:04, 1 October 2019 (UTC)
@ChristianKl: Help:Label says the opposite: "But note that for other client wiki namespaces like "Category" and "Template" the labels should be identical to sitelinks (the disambiguation parts shouldn't be removed), as these types of pages usually have only one common type of descriptions and there may occur API errors of non-unique pair consisting of label + description when trying to set descriptions to other items.".--Russian Rocky (talk) 11:27, 1 October 2019 (UTC)
I think "should" should generally read for policy in the RFC 2119 sense e.g. "there may exist valid reasons in particular circumstances to ignore a particular item, but the full implications must be understood and carefully weighed before choosing a different course". I do think that ŠJů argued why there are valid reasons in this case not to have the same name. The core intent of this sentence is also about not removing disambiguation, it's not about not adding further disambiguation. ChristianKl❫ 11:40, 1 October 2019 (UTC)
Every guidline in the English Wikipedia uses "should" to begin with, but people still follow them. I don't think Wikidata is kind of different.
Sorry, but I'm not convinced with your vague answer. It looks like your own interpretation rather than a reflection of consensus. I don't think ŠJů's reasons are more valid than possible API errors and complicated searching affecting everyone. I thought you could explain it better in Help talk:Label#Category and template labels, but it's your subjective interpretation again which doesn't explain when we should or shouldn't change category labels. Others are welcome there as well, share you opinion please.--Russian Rocky (talk) 16:10, 1 October 2019 (UTC)

Seems to me that if we intend to model categories, and if there is an English-language name for the category, used on en-wiki, we should follow it for our English-language name. The only exception I can think of is if Commons has a distinct name for the same concept, and we think the chose better. en-wiki and Commons both have processes to discuss category names if people think they are wrong. One person changing it here unilaterally because they disagree with the consensus on the sites that have a process seems like a poor plan. What's to stop the next person coming along and changing it yet again? - Jmabel (talk) 05:33, 2 October 2019 (UTC)

Find WD redirects

Does somebody knows how to find the items that have been redirected after a merge ?. Thanks, Amadalvarez (talk) 16:04, 1 October 2019 (UTC)

Special:WhatLinksHere for instance, or ?redirect owl:sameAs ?item in SPARQL… —MisterSynergy (talk) 18:17, 1 October 2019 (UTC)
Thanks, MisterSynergy. SPARQL is what I looked for. Amadalvarez (talk) 04:43, 2 October 2019 (UTC)

Merge Q817393 and Q6939850

Because these items overlap, they require specialized merge. Monniasza (talk) 19:35, 24 September 2019 (UTC)

@Monniasza: they can't be merge, not until en:Black people and en:Murzyn is merged. And it seems to be two different concepts (the second being a subclass of the first I guess?). Cheers, VIGNERON (talk) 20:24, 24 September 2019 (UTC)
Both sitelinks are for the term "Murzyn", not for an article about black people. Searching for a link to "Black people", I found pl:Afroamerykanie (African American) links to pl:Czarna Afryka (Sub-Saharan Africa), which links to pl:Czarna odmiana człowieka, which is linked to negroid (Q338460) here. Peter James (talk) 20:42, 24 September 2019 (UTC)
@Peter James, VIGNERON, Monniasza: Please also consider merging both Commons categories, otherwise this merge request can't work. --Liuxinyu970226 (talk) 23:48, 1 October 2019 (UTC)
They shouldn't be merged, items linked to Q6939850 are about the word "Murzyn" - the Commons category is a subcategory of Category:Polish language (Q6916114). The English and Commons content doesn't belong in the pages connected to Q817393, so Q6939850 is still needed unless the article and category are deleted. The Polish article could be expanded and moved to Q817393 but for now should stay linked to Q6939850. Peter James (talk) 11:51, 2 October 2019 (UTC)

RfDs archival

Hi,

how often is WD:RfD archived? And what about sections marked as resolved for days?

Nomen ad hoc (talk) 20:40, 1 October 2019 (UTC).

Most sections get almost instantly archived, but they must be marked with {{Deleted}}, {{Not deleted}} ({{Not done}} doesn't get recognized by the bot) or {{Merged}}. Sjoerd de Bruin (talk) 09:13, 2 October 2019 (UTC)
@Hazard-SJ: Could we allow the bot to also recognize ({{Not done}}? ChristianKl❫ 09:59, 2 October 2019 (UTC)

Gadgets with wrong link to help...

Hello,

the link to help on KeyShortCuts gadget is wrong, since it has been archived. Could someone please change it to Wikidata_talk:Tools/Archive_1#Keyboard_shortcuts

It is a very useful tool, that I often recommend to new contributors. It would be better if the help was findable for them :)

Thanks to User:Ricordisamoa who coded it :) Hsarrazin (talk) 07:29, 2 October 2019 (UTC)

@Hsarrazin: I have updated the description. Feel free to translate it: MediaWiki:Gadget-KeyShortcuts. ‐‐1997kB (talk) 01:39, 3 October 2019 (UTC)

scope of "main subject"

How "main" should the use of the main subject (P921) property be? Should it be restricted to just one or two, for example? I am creating items for a series of historico-biographical articles which have titles like "Social Reformers: John Howard, Elizabeth Fry, William Lloyd Garrison" and each article has a broad topic like "reform" in this example or "history of science" or whatever, with the historical theme being exemplified through discussion of between one and six key persons. Should the exemplar persons be considered "main subjects" even when there are many of them? Surely if there is just one exemplar person, for instance a phase of European political history examined through the life of Klemens von Metternich, they are a main subject. Levana Taylor (talk) 20:06, 2 October 2019 (UTC)

Use as many “main subject” statements as make sense to you. I always use anything specifically mentioned in a title (unless it’s used metaphorically). There was some discussion around multiple “main subjects” last year at Wikidata:Property proposal/subject facet, but that property was not implemented, so main subject is what we have to work with. - PKM (talk) 20:33, 2 October 2019 (UTC)

VIAF records with multiple WKP entries

Dear friends, this is the second time that I noticed a VIAF record wirh multiple WKP entries.
The first time the additional record was a redirect to the "main" WKP entry.
This time I noticed VIAF:56731778 –– https://viaf.org/viaf/56731778/ containing both a "main" WKP entry Hafez (Q6240) and an additional one "No label defined" (Q10293519).
It is obvious that the WKP related Wikidata items should be combined.
questions:

who can "scan" tte VIAF database for similar cases?
by which procedure the VIAF records will be updated?
where a "log" page will be available in wikidata to reflect the status of the individual cases ("tickets handling")?
other ideas
who can fix this particular case?
people involved: Magnus Manske, Lydia Pintscher, LydiaPintscher, Alkanot4ran, ...

Best regards
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 03:49, 2 October 2019 (UTC)

In this particular case, Q10293519 is a Portugues entry that has now been redirected to the other. I've merged the item. Circeus (talk) 04:32, 2 October 2019 (UTC)
thanks a lot Circeus!
no bias — קיין אומוויסנדיקע פּרעפֿערענצן — keyn umvisndike preferentsn talk contribs 17:18, 3 October 2019 (UTC)

Auto-notifying users when their username is added to a wikidata item

Are users currently notified when Wikimedia username (P4174) is added to a wikidata item? If not, it would be good to automatically do so (first so that they can fix if it's incorrect, or add further information if correct). It would also avoid any accidental outing going unnoticed. T.Shafee(evo&evo) (talk) 02:45, 2 October 2019 (UTC)

  • No and, if it's done publicly on their page, this might compound the problem. Not sure either what we should do with users who add these (especially those knowing the person doesn't want it to be in Wikidata). One could design a filter that prevents most people but the user to add it. --- Jura 15:57, 4 October 2019 (UTC)

WD:N and Q68030853

Hello,

The (currently deleted) item Q68030853 link Jean-Jacques Netter (Q68030698) and Charles Gave (Q2959181). Does it fullfill the criteria « 3. It fulfills a structural need, for example: it is needed to make statements made in other items more useful. » in WD:N? Visite fortuitement prolongée (talk) 20:06, 3 October 2019 (UTC)

Just a comment: I'm not convinced that linking the two items would fulfill any real "need". Nomen ad hoc (talk) 20:10, 3 October 2019 (UTC).
That's why I'm asking ;-) Visite fortuitement prolongée (talk) 20:42, 3 October 2019 (UTC)
Links them how? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:18, 3 October 2019 (UTC)
Q68030853 founded by (P112) Jean-Jacques Netter (Q68030698) and Q68030853 founded by (P112) Charles Gave (Q2959181). Visite fortuitement prolongée (talk) 20:42, 3 October 2019 (UTC)
This is not how we understand "structural need"; this term usually refers to the opposite situation, i.e. the item in question is linked by at least one other item. The item Q68030853 ("Institut des libertés", a think tank) in its most recent form did indeed qualify for deletion. If you happen to have relevant sources or identifiers, we can restore it of course. —MisterSynergy (talk) 20:46, 3 October 2019 (UTC)
This is exactly how I understand "structural need"; we need it to show the (structural) relationship between two entities. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 07:59, 4 October 2019 (UTC)

Item restored per critera #2: "clearly identifiable conceptual or material entity". Feel free to delete it again if there is a consensus. Ayack (talk) 10:38, 4 October 2019 (UTC)

Unknown entries

Hey all. How can i put unknown date entry for a person? Like death date for Razzuq Faraj Razzuq (Q60913968)? there is lack of info for his death date. --Ruwaym (talk) 21:48, 3 October 2019 (UTC)

Born in 1919, maybe he is still alive? Ghouston (talk) 06:05, 4 October 2019 (UTC)
Ghouston I found a relative of him in Facebook, hope he respond me. --Ruwaym (talk) 07:45, 4 October 2019 (UTC)
It seems like the usage instruction for the two could be more clear to get people to understand it.
You only put in an unknown death date if you have a source that tells you the person is dead but not when the person is dead. Otherwise you simply leave it unfilled. Otherwise, see the guidance of unknown values on https://www.wikidata.org/wiki/Help:Statements. ChristianKl❫ 10:23, 4 October 2019 (UTC)
I think for someone born more than 120 years ago, it would be reasonable to set the date of death to unknown value, with derived from (P5191) date of birth (Q2389905). Ghouston (talk) 10:53, 4 October 2019 (UTC)
(tangential) From the description of derived from (P5191) this is not the intended usage. OTOH if that is not frowned upon I'll happily use it for references instead of based on heuristic (P887). --SCIdude (talk) 13:36, 4 October 2019 (UTC)
Yes, that was the wrong property, I was thinking of inferred from (P3452). based on heuristic (P887) could be used too, although I think you'd have to create an item for the "heuristic". Ghouston (talk) 00:21, 5 October 2019 (UTC)
  • Rather than adding P570, you could add floruit (P1317) with the last known date. Please avoid adding P570 unless you are sure about it. We already had problems with people marked as dead by BdF and a person complaining as we copied their info. --- Jura 11:31, 4 October 2019 (UTC)

Help re architect

One source for Old Warren County Courthouse (Q14693338) says "It should be noted that although William Weldon is listed as the architect in some sources, the museum itself has no precise record of this attribution. There is even a theory—again without documentation—that the courthouse was designed 'by a slave named Jackson.'" How should I express the concept that the courthouse may have been designed by a slave named Jackson? I'm not sure about creating an item for a person who may or may not have existed. Right now I have architect (P84) = enslaved person (Q12773225), with qualifiers stated as (P1932) = Jackson and sourcing circumstances (P1480) = possibly (Q30230067). Other thoughts? Calliopejen1 (talk) 17:10, 4 October 2019 (UTC)

Since the slave architect is just a theory, wouldn't it be better to set the architect to "unknown", maybe with a link to the source that describes the uncertainties? Moebeus (talk) 18:22, 4 October 2019 (UTC)
You could also add “statement is disputed by” to the attribution to Weldon, and include the quote about Jackson in the reference. - PKM (talk) 19:51, 4 October 2019 (UTC)

year precision for far future dates

On the items Bad Wolf (Q2497649) and The Parting of the Ways (Q2668975), I've tried to enter a date depicted (P2913) statement with the year 200,100 as its value (as the story takes place in that exact year in the far future). But for some reason, the value is saved with a century precision, even through I've manually switched to a year precision before saving. Testing around, it seems everything up to the year 4000 works as usual, e.g. the years 2400, 3000, 3550, etc. all automatically have the year precision and don't change when saved. After that, e.g. the years 4010, 4500, 5000 automatically switch to the less precide decade or century precisions and changing them to year precision doesn't work. Is this working as intended (for whatever reason) or a bug? --Kam Solusar (talk) 22:21, 4 October 2019 (UTC)

Odd. I entered 1.1.200100 then switched from day to year precision: that worked. --- Jura 09:14, 5 October 2019 (UTC)
Oh, right. I did that a while back on The Long Game (Q958270) by using +200000-06-00 and switching to year precision instead of just +200000-00-00 (which originally gave a precision of 100000 years). Still not sure why it's necessariy. --Kam Solusar (talk) 12:16, 5 October 2019 (UTC)

Hiero syntax

E1
D40
N28mS40t
O49

If you are interested in WikiHiero syntax (<hiero></hiero>), you might want to look into Wikidata:Property proposal/name in hiero markup. --- Jura 09:00, 5 October 2019 (UTC)

P373 etc. links to disambiguation pages or redirects

The values of Commons category (P373), Commons gallery (P935), and Commons maps category (P3722) often do not link to a Commons page or Commons category but to disambiguation pages or redirects at Commons. Cause is surely the page/category move at Commons itself made after the creation of a Wikidata dataset without sending a notification to Wikidata. I think of two solutions of this problem. Of course, Wikimedia Commons should have a bot to make the necessary changes on other Wikis, Wikidata included, similar to changes of image name moves. The second solution is to start a bot at Wikidata to substitute redirects to the page/category itself and to mark values to disambiguation pages.

This problem more seldomly occurs in the commons sitelink in the "other website" section, too. Trying the replacement of this value can produce a new error because the correct value is used in another (category) dataset. --RolandUnger (talk) 09:32, 5 October 2019 (UTC)

Help:Move

What is the procedure for moving a misplaced Category? The greek wiktionary 'Surnames' Κατηγορία:Επώνυμα is misplaced at Category:Names (Q5642451).

which practically is identical to Category:Surnames by language (Q9819961). Futhermore, why can't I place the greek category to BOTH these Categories? It fits to both. But if not possible, I need to move it at Category:Surnames (Q7045213). Thank you -sorry that I am not familiar with wikidata-. Sarri.greek (talk) 16:16, 5 October 2019 (UTC)

How many Wikipedia/Wikimedia entities lack Wikidata items?

Is there a way to visualize the Wikipedia articles (in any language) or Commons categories/galleries that don't have a corresponding Wikidata item? I know that bots generally create a Wikidata item within a few hours to days after a Wikipedia article is created (at least in English), but the lag can be up to a week or more, and who knows how many "orphan" entities there are. Are there tools to identify these and create needed items? Conversely, a way to find Wikipedia redirects that correspond to existing Wikidata items would be helpful (e.g. a musician with a Wikidata item is redirected to her band's article on Wikipedia due to notability or other reasons). I'm not a computer programmer, so advice along the lines of "design a script to do x y and z" is unhelpful. Thanks, -Animalparty (talk) 16:52, 5 October 2019 (UTC)

For example, see:

M2k~dewiki Thank you! This is very helpful. -Animalparty (talk) 17:35, 5 October 2019 (UTC)

Cheap tickets

New field of activity for Wikidata? Wikidata:Property proposal/Ticket Liquidator artist ID, etc. --- Jura 18:05, 5 October 2019 (UTC)

Add statement: phylogeny

As a phylogeneticist, I work a lot with phylogenies. I would enjoy being able to download a phylogeny of a certain genus/family/order/etc. in a raw format.

I think Wikidata is the best place to add these. Have I concluded this correctly?

  • If no, could you redirect me to a the proper Wiki project?
  • If yes, there's another question below

To avoid doing original research, I can imagine that Wikidata decides to only host published phylogenies. Do I guess right here?

  • If yes, there is a problem: the raw data and calculation of the phylogeny may already be lost, therefore on may have to result to calculate the phylogeny from afresh
  • If no, there is a problem: from one same input, there are many ways to construct a phylogeny from it. Therefore, there are more than one answers possible. Supplying a script how to create a phylogeny from an input, however, does result in reproducible answers. Would that fit Wikidata?

I hope Wikidata is the place where once I can easily obtain phylogenies from, even though there are some problems due to the nature of constructing these.

Looking forward to your reaction, Richel J.C. Bilderbeek  – The preceding unsigned comment was added by Bilderbikkel (talk • contribs) at 09:31, 1 October 2019‎ (UTC).

  • Wikidata isn't a place that hosts raw data. We rather host items for individual species/genus/family/order's and relationships between items. Primates (Q7380) would be an item for the order of the primates and Haplorrhini (Q82261) the item for the suborder haplorrhini which has a "parent taxon" relationship with the former. In this case the claim has two high authority sources. In other cases relationships might be sourced less well.
Currently, I don't think there's a Wikiproject that hosts raw datasets of the kind you are thinking about. ChristianKl❫ 12:11, 1 October 2019 (UTC)
  • Bilderbikkel Additionally, phylogenies are not static things one can download or generate from hierarchical relationships. Each one is inferred by different lines of evidence and analyses (anatomy, mitochondrial genes, nuclear genes, etc), and any single clade may have numerous alternative and contradictory phylogenies. We don't host the intricate, taxon-specific character matrices required for phylogenetic inference. There are better repositories for genetic sequence data, i.e. GenBank. -Animalparty (talk) 16:16, 5 October 2019 (UTC)
  • Bilderbikkel Here is an example of tree for ape (https://w.wiki/9YB ). SPARQL soruce code for that is here [1]. After click the excute button Execute button, you can download the data. See download help. There is another example at Twitter [2]. Currently Wikidata can not provide one-coherent phylogenetic tree, but a patchwork-tree based on many edits from various contributors. --Was a bee (talk) 20:18, 5 October 2019 (UTC)

Unable to edit Q6308220

Category:Kingdom of Hungary (Q6308220) just had a merger by &beer&love (talkcontribslogs), but now when I try to edit it, I get the error "Could not save due to an error." due to several conflicting sitelinks. It's part of a few such mergers that I've spotted just now, since en:Category:Kingdom of Hungary and commons:Category:Kingdom of Hungary have become unlinked from Wikidata even though the sitelinks still exist. Making an edit to the Wikidata item resolved the other cases, but that's not possible here. Does anyone have an idea of what's happening, or is this something for phabricator? Thanks. Mike Peel (talk) 15:31, 4 October 2019 (UTC)

Never mind, this seems to have spontaneously resolved itself. Thanks. Mike Peel (talk) 21:31, 5 October 2019 (UTC)

Happy Birthday

A belated happy birthday to Wikidata:Property proposal/(de)evolution method, which is now one year and two days old. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:58, 5 October 2019 (UTC)

Improved edit summaries for edits coming from wbeditentity API

Hello all,

We are about to make some changes to edit summaries that appear when an edit is made using wbeditentity API, which includes for example editing terms (labels, descriptions, aliases) from the new mobile termbox. The summary messages currently contain “Changed an Item” as a comment. The new summaries will include the message "Changed label, description and/or aliases in # languages", where # is the count of distinct languages that terms in them have been affected.

This change will be deployed on Wednesday, October 2nd, and the new edit summaries will start changing from that point. The old edit summaries will not be changed. More improvements will follow in the future (for example, more details regarding the terms that have been changed and the languages).

You can see more details in this ticket. If you encounter any issues, please let me know. Thanks, Lea Lacroix (WMDE) (talk) 14:06, 30 September 2019 (UTC)

Pinging Pintoch, since I think this will affect the “labels”/“tags” in the edit groups tool. (This should make it possible to classify edit groups that use wbeditentity rather than other APIs more accurately, but I’m not sure if the tool needs to be updated to take advantage of that.) --Lucas Werkmeister (talk) 14:18, 30 September 2019 (UTC)
Many thanks for the ping, I will update EditGroups accordingly. I also replied on the wikidata-tech mailing list about this. (summary: this is fantastic news and I wonder if it was considered to reuse summaries from atomic API actions in wbeditentity, if they are applicable. That would avoid having to do that client-side, as is currently done in Wikidata-Toolkit: https://github.com/Wikidata/Wikidata-Toolkit/blob/master/wdtk-wikibaseapi/src/main/java/org/wikidata/wdtk/wikibaseapi/TermStatementUpdate.java#L388). − Pintoch (talk) 14:26, 30 September 2019 (UTC)
The new edit summaries are now supported: https://tools.wmflabs.org/editgroups/?tags=wbeditentity-update-languages-and-otherPintoch (talk) 11:56, 6 October 2019 (UTC)

Clean up country / country of citizenship

A significant number of items for humans use country (P17). I think we should look for ways to clean this up. An obvious initial step would be automatically removing P17 from all entities that are instances of human (Q5), have country of citizenship (P27) defined and its value is the same as country (P17) (example query). What do you think? --MarioGom (talk) 12:00, 2 October 2019 (UTC)

  • That seems harmless, as a first step. - Jmabel (talk) 15:52, 2 October 2019 (UTC)
  • @MarioGom: Sounds good! --Marsupium (talk) 02:48, 5 October 2019 (UTC)
  • I'd recommend also making sure that the item doesn't also have another P31 before removing P17. Some items might be the result of incorrect mergers between human and non-human items, or otherwise have ambiguous referents. --Yair rand (talk) 18:49, 6 October 2019 (UTC)

property for Number of services produced ?

I try to find which property I may use to represent the "production of services" (no goods) by period of time. It could be applicable to a company (call center services) or a venue (airport), etc. Now, we have production rate (P2197) which it seems limited to material goods. I also tried to select collection or exhibition size (P1436) or the most generic quantity (P1114) but reflect "things subject has" or a "number of parts" of the subject, but I look for "things subject produce". Some examples of what I want to show are: Airport number of operations (by year, for instance), number of loaned book in a library, number of information services in a call center, etc.. Thanks, Amadalvarez (talk) 05:00, 6 October 2019 (UTC)

named as (P1810) vs. stated as (P1932)

How do i know what which one to use? --Trade (talk) 15:52, 3 October 2019 (UTC)

Use "named as" to record how the subject was named in a database or reference. Use "stated as" to record how the object (value) was stated in a database or reference. - PKM (talk) 19:09, 3 October 2019 (UTC)
I know I used "named as" where I should have used "stated as" quite a few times -- is there a way I can search to find all my edits that contain "named as"? Never mind, I figured that out, but now I'm running into 2 different problems. Levana Taylor (talk) 19:53, 3 October 2019 (UTC)

Use of "author name string" with uninformative names and unknown identities

"Stated as" isn't considered a valid qualifier for author name string (P2093). So what do you do when you know the author of an article, they're credited differently in the magazine, and they have no item? I don't necessarily want to create an item for every single author for reasons discussed in this wikisource conversation.Levana Taylor (talk) 19:53, 3 October 2019 (UTC)

I think author name string (P2093) is intended to already contain the value that's listed in the database. If there's "J. Doe" then you say author name string (P2093) "J. Doe". Do you think there are cases where you know that "J. Doe" is actually named "J. Doe" but there's still not enough information to create an item for her? ChristianKl❫ 10:24, 4 October 2019 (UTC)
I'm dealing with articles in Once a Week magazine (1860s), in which many of the articles are signed with initials or pseudonyms or unsigned. Luckily, though, an account book survives with records of payments to contributors which allows identifying quite a few contributors (info is in the Curran Index. Not all can be identified because some articles don't have a payment recorded or note an uninformative name like "E. Smith" or "Miss Brown." So we've got multiple names and identities to record: the signature in the magazine (sometimes null), the contributor's real name as recorded in the account book (sometimes null), and their full name allowing creation of an item. If their full identity is unknown, practice here is to use "author name string." So there are two problems: the fact that there are two "stated as" values to record, one in the account book and one in the magazine; and the fact that author name string doesn't currently allow a stated as qualifier. Levana Taylor (talk) 11:20, 4 October 2019 (UTC)
author name string (P2093) shouldn't have it as it's meant to be a temporary solution until one adds author (P50) possibly with stated as (P1932) as qualifier. --- Jura 11:09, 4 October 2019 (UTC)
I dunno about temporary, the identity could be forever unknown. Levana Taylor (talk) 11:20, 4 October 2019 (UTC)
An author identified with an author name string isn't unknown .. it's just that for some reason we didn't make an item for them. --- Jura 11:35, 4 October 2019 (UTC)
The tip text for the "author" property says "use 'author name string' if Wikidata item is unknown." Am I misinterpreting that? I suppose "is unknown" could be meant as saying "you think there is a Wikidata item but you don't know what it is," in which case you would use unknown value for "author" if the person is completely unknown. Levana Taylor (talk) 11:52, 4 October 2019 (UTC)
The main use of the property is to enable creation of items about papers or for references without having to go through creating items for authors. --- Jura 14:32, 4 October 2019 (UTC)
Or with named as (P1810), which seems more precise. Nomen ad hoc (talk) 17:52, 5 October 2019 (UTC).
I don't really get why that was added as qualifier. I think it's an error. It leads to stuff like [3] --- Jura 18:19, 5 October 2019 (UTC)
According to ((ping|PKM}} in the section just above this, it would indeed be incorrect to use "named as" with author name string. Levana Taylor (talk) 00:00, 6 October 2019 (UTC)

Of course author name string (P2093) shouldn't have a "stated as" as a qualified, it literally takes the author (P50) and "stated as" itself and combines them into a single text-string type property. By and large, it's intended to ease imports (and in particular automated imports) of sources by avoiding automatic creation of hundreds of author items that would them have to be manually merged or disambiguated. The end goal is to have a proper author (P50) later on. Of course it's also useful where the full author cannot be determined in any way. Circeus (talk) 21:48, 6 October 2019 (UTC) Some concrete examples will be useful.

  1. The Grass of the Field (Q69649876) is by Margaret Plues (Q21606756), recorded in the accounts as "Miss Plues," and unsigned.
  2. John Brown at Woolwich (Q69649269) is by Robert Barlow McCrea (Q62475737), not recorded in the accounts, and signed "R. B. M."
  3. How an Advertisement Got a Wife (Q69650426) is by ??, in the accounts as "Mrs. Davis," and signed "M. R. J."
  4. A Ramble in the Forest of Dean (Q69649595) is by ??, not in the accounts, and signed "B."
  5. My Uncle's Handbook (Q69649453) is by ??, not in the accounts, and unsigned.

Levana Taylor (talk) 12:15, 4 October 2019 (UTC)

  • Here is one for "B.": Q69659100. Maybe with the general information about the periodical it can be determined who uses "B." as signature. --- Jura 14:39, 4 October 2019 (UTC)
OK then, what do you when you have zero of anything to refer to the author by, as in my last example? "author unknown value stated as no value" or what? Levana Taylor (talk) 20:04, 4 October 2019 (UTC)

How to indicate unsigned

Is the right way to indicate an unsigned article "stated as no value"? Levana Taylor (talk) 19:53, 3 October 2019 (UTC)

Should there be a Uniform Resource Name (Q76497) and/or Uniform Resource Identifier (Q61694) property (similar to official website (P856) / URL (P2699)) ?

I think it would be good to have. I can recall wanting it before but I can't exactly remember why now. I will try think of some examples and add them. Maybe most generic would be to just have "Official URI" property and then we can add any valid URI? Probably it should allow multiple values. Iwan.Aucamp (talk) 06:20, 4 October 2019 (UTC)

Discussion

T.seppelt (talk) 21:00, 18 February 2016 (UTC) Vladimir Alexiev (talk) 11:59, 13 March 2017 (UTC) GerardM (talk) 15:58, 26 March 2017 (UTC) Jonathan Groß (talk) 17:52, 26 March 2017 (UTC) Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits Jneubert (talk) 13:47, 29 April 2017 (UTC) Framawiki (please notify !) (talk) BrillLyle (talk) 10:09, 10 July 2017 (UTC) Sic19 (talk) 20:42, 12 July 2017 (UTC) Wikidelo (talk) 21:15, 8 May 2018 (UTC) salgo60 Salgo60 (talk) 07:09, 10 June 2018 (UTC) ArthurPSmith (talk) 19:52, 22 August 2018 (UTC) PKM (talk) 19:40, 23 August 2018 (UTC) Ettorerizza (talk) 06:44, 8 October 2018 (UTC) Fuzheado (talk) 03:47, 19 December 2018 (UTC) Daniel Mietchen (talk) 16:30, 7 April 2019 (UTC) Eihel (talk) 15:13, 19 June 2019 (UTC) NAH Iwan.Aucamp

Pictogram voting comment.svg Notified participants of WikiProject Authority control DarTar (talk) 08:28, 19 May 2018 (UTC) Daniel Mietchen (talk) 11:24, 19 May 2018 (UTC) Maxlath (talk) 11:33, 19 May 2018 (UTC) Jumtist (talk) 11:34, 19 May 2018 (UTC) Pintoch (talk) 11:40, 19 May 2018 (UTC) JakobVoss (talk) 11:44, 19 May 2018 (UTC) PKM (talk) 20:12, 19 May 2018 (UTC) ArthurPSmith (talk) 13:47, 22 May 2018 (UTC) Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits Vladimir Alexiev (talk) 12:43, 27 November 2018 (UTC) Ivanhercaz Plume pen w.png (Talk) 11:55, 3 February 2019 (UTC) Epìdosis 11:23, 15 April 2019 (UTC) Tris T7 TT me


Pictogram voting comment.svg Notified participants of WikiProject Wikipedia Sources

Andheb (talk) ElanHR (talk) 10:05, 17 March 2019 (UTC) Jneubert (talk) 20:55, 23 July 2019 (UTC) Daniel Mietchen (talk) 20:47, 26 September 2019 (UTC) Pictogram voting comment.svg Notified participants of WikiProject DatasetsEihel (talk) 08:06, 5 October 2019 (UTC)

  • Symbol support vote.svg Support Iwan.Aucamp (talk) 06:20, 4 October 2019 (UTC)
  • Can you give a few samples? Would this be for properties or items? To duplicate URLs or other identifiers? --- Jura 11:06, 4 October 2019 (UTC)
  • Pictogram voting comment.svg Comment A URI can be a URL or an URN (as well as other usage schemes - IETF Definition). For URN, there are less than 100 namespaces. They are registered by IANA here. Unless I'm wrong, some will be able to serve on WD and some will not. An example of a URN: urn:isan:0000-0002-E60F-0000-0-0000-0000-3 for The Intouchables (Q595). The ISAN identifier on Qs is already defined by ISAN identifier (P3212). This is the same for many IANA namespaces (NID of URN): ISBN, ISSN, etc. There are Properties for ISBN, ISSN, etc. —Eihel (talk) 11:59, 4 October 2019 (UTC)
    • Maybe we could create a URN formatter property (with values like "urn:isan:$1" for the sample given). Not sure if storing values like "urn:isan:0000-0002-E60F-0000-0-0000-0000-3" would be a good idea. This although we already have similar duplication with DOIs for some properties. --- Jura 14:25, 4 October 2019 (UTC)
      • Hello @Jura1, Iwan.Aucamp:, I was also waiting for a response from the proposer on your request, because I do not understand his exact request. When he writes "I can not exactly remember why now", I am perplexed. My comment is only to make other contributors think about the usefulness. There is a permanent resource of URIs and URNs: ​​if a URL is not permanent, an organization's registration to one of these concepts makes a resource permanent. Just knowing that ISBNs or ISANs are permanent does not advance us. These 2 properties (if they exist) should be of string type as proposed by IETF (so not as you propose: urn:$1). For example, I gave an ISAN URN, but do not create a property by namespace (one for ISAN, then one for ISBN, then one for ISSN, and then ...). An answer to your question will tell me what values ​​a URI property can take. But after all that I wrote, I will wait for a response from the proposer and other opinions, but I feel redundant information unnecessary.I am sure that proposals have already been made. Cordially. —Eihel (talk) 07:52, 5 October 2019 (UTC)
  • Wikidata:Property proposal/urn formatter. --- Jura 08:23, 5 October 2019 (UTC)

Symbol oppose vote.svg Oppose unless the proposed can give some useful examples --Vladimir Alexiev (talk) 06:53, 7 October 2019 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────

Thank you for the feedback. I think you make a very good @Eihel: point and in general I would agree that it would be better to have rather have specific identifiers like ISSN and ISBN instead of the URN equivalent and in such cases a URN formatter property would be better. I think the cases that I wanted were IETF URNs:

urn:ietf:rfc:2141
urn:ietf:std:50
urn:ietf:id:ietf-urn-ietf-06
urn:ietf:mtg:41-urn

But there are similar cases

OASIS URNs:

urn:oasis:names:specification:docbook:dtd:xml:4.1.2
urn:oasis:names:tc:docbook:dtd:xml:docbook:5.0b1
urn:oasis:names:technical:memo:9502:1995
urn:oasis:member:A00024:x

ITU URNs:

urn:itu:t:rec:g.8052.1:yang:connectivity

I think though following from what @Eihel: said these cases may be better handled by specific properties (i.e. RfC ID (P892) with URN Formatter). There may be cases where this does not quite work that like tag URIs and Federated Content URNs but I think for the time being URN Formatter may be the best option. I will add some details there and see if we can get that property proposal accepted. Iwan.Aucamp (talk) 07:04, 7 October 2019 (UTC)

I think we could already do something clean before making a proposal (hence my vote against the proposal). You are not in a hurry. Do not put in production a wobbly Property, but rather something accomplished. URN is a very specific URI compared to a URL: it does not designate a location, it is unique and… soon more —Eihel (talk) 13:01, 7 October 2019 (UTC)

Graphing relationships

Do we have a tool (or a query) to show the shortest graph of connections beteen two arbitrary items, in a "six degrees of separation" manner? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 08:02, 4 October 2019 (UTC)

  • If not, the algorithm would presumably be to start from one, list out all of its connections, then recursively do the same for each of those until you encounter the other item. Possibly might optimize average case (but not worst case) starting from both ends. Could get pretty big pretty fast if they have a lot of connections but aren't close to one another. - Jmabel (talk) 16:19, 4 October 2019 (UTC)
  • WDQ has a graph view and shortest path queries (using a Blazegraph specific extension). Maybe post a more specific case on the "queries wanted" page --Vladimir Alexiev (talk) 06:57, 7 October 2019 (UTC)

Property for oral history / voice interview of a subject ?

Do we have a property for oral interview of a subject, describing their life (or some part of it) in their own voice?

It struck me that this would be quite an interesting thing to note, for items about people.

eg Ian Charles Scott (Q21454862) -> https://wickheritage.org/voices_play.php?ind=91 ; or, perhaps, any episode of Desert Island Discs (Q1200587).

I suppose one could use described by source (P1343), marking the project as an oral history project with instance of (P31) an oral history (Q558929), and then URL (P2699) for the URL of the interview. Setting object has role (P3831) = "oral history interview" could help systematically identify such sources with queries.

What do people think? If were to start systematically trying to identify and state the coverage of oral history projects around the world, how might we model the data? Jheald (talk) 20:23, 6 October 2019 (UTC)

described at URL (P973)? audio (P51)?, spoken text audio (P989)? Circeus (talk) 22:00, 6 October 2019 (UTC)
None of that, actually. audio recording of the subject's spoken voice (P990) is the one. Thierry Caro (talk) 11:39, 7 October 2019 (UTC)
Not for an intervew. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:28, 7 October 2019 (UTC)
I would create an item for the interview, and link that item to the subject using main subject (P921) (or some more precise property, did not look much into it) Jean-Fred (talk) 17:12, 7 October 2019 (UTC)

Canonicity

Are there any property that i can use to show that a creative work have been deemed non-canon by the creators? --Trade (talk) 22:49, 6 October 2019 (UTC)

Canon is with respect to a specific "creative universe" I would assume here? So I think the way to do this would be to have a statement part of (P361) "item for creative universe" with qualifier statement disputed by (P1310) the creator who states otherwise, and support that by a reference. ArthurPSmith (talk) 12:06, 7 October 2019 (UTC)

Removing double entities/tags from dutch municipalities

Is it possible to make a query that deletes the dutch municipality tag (Q2039348) from entitites that don't have a start (P580) or/and end time date set (P582)? And if yes, how would such a query look? if not, what's the best way to tackle this problem (clicking 2000 times is not really an option)? Antoni1626 (talk) 08:44, 7 October 2019 (UTC)

This is the query for the municipalities with a start date set:

SELECT ?municipalityLabel ?municipality ?begindatum ?einddatum ?CBS_gemeentecode WHERE {

SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
?municipality p:P31 ?statement.
?statement ps:P31 wd:Q2039348;
  pq:P580 ?begindatum;

optional { ?statement pq:P582 ?einddatum } optional { ?municipality wdt:P382 ?CBS_gemeentecode. } }

Query for municipalities without start and end date:

SELECT ?municipalityLabel ?municipality ?begindatum ?einddatum ?CBS_gemeentecode WHERE {

SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". }
?municipality p:P31 ?statement.
?statement ps:P31 wd:Q2039348;

minus {?statement pq:P580 ?begindatum;} minus {?statement pq:P582 ?einddatum } optional { ?municipality wdt:P382 ?CBS_gemeentecode. } }

Here is an example: https://www.wikidata.org/wiki/Q2096012 (place), no start or end date, but the other entity https://www.wikidata.org/wiki/Q40671450 (municipality) has the correct dates Antoni1626 (talk) 08:44, 7 October 2019 (UTC)

A query is a query, you cannot change the database with it. I use sed (Q305876) or, in more complicated cases, Python to generate QS from a query's CSV or TSV. --SCIdude (talk) 08:59, 7 October 2019 (UTC)
Also, please don't post what is essentially the same question to two forums [4]. --Tagishsimon (talk) 09:22, 7 October 2019 (UTC)

Hong Kong Protests

2019 Hong Kong protests (Q64509602) is a complex, sensitive, and poorly sourced item. At least 3 points that would need attention:

  1. label: it was initially called "protests against the extradition bill" or something like that. As the protests have outlived the extradition bill, most Wikipedias have renamed their article but not all labels have been changed accordingly.
  2. Causes. The item has has cause (P828)->causes of the 2019 Hong Kong protests (Q67184561) but causes of the 2019 Hong Kong protests (Q67184561) is virtually empty. Beside, we have several properties that express causation: has cause (P828), has immediate cause (P1478) and has contributing factor (P1479). I am not sure that everything should be lumped together in causes of the 2019 Hong Kong protests (Q67184561).
  3. number of injured (P1339), number of deaths (P1120).
  • The demonstration are not over, so date qualifiers are important. Which one should it be, latest date (P1326) ?
  • Can we use qualifiers to explain what exactly is included ? Who is counted as injured is not necessarily clear. Number of deaths seems clearer, but apparently all deaths are demonstrator's suicides and I feel that it should somehow be mentionned. -Zolo (talk) 13:33, 7 October 2019 (UTC)

Merging of fatigue (Q15729017) with fatigue (Q9690)

I want to merge fatigue (Q15729017) with fatigue (Q9690). I tried once but there was some opposition so I would like to clear up what people expect be done before the merge. Before I started editing for some languages fatigue (Q9690) was linking to "work fatigue" and for those languages fatigue (Q15729017) was linking to fatigue in general as opposed to work specific fatigue. I don't think either item is appropriate to be used for work specific fatigue and would rather make a new item for that. @Infovarius:. Iwan.Aucamp (talk) 07:45, 1 October 2019 (UTC)

It's more complicated then that. DeWiki has items for both and for Fatigue-Syndrom (Q2713697) and the difference is not about whether or not something is work specific fatigue.
We might rename fatigue (Q9690) into "Malaise and fatigue" to make it more clear that the item is about the clinical symptom.
If there are other Wikis that do have an article about work fatigue, feel free to create a new item for that. ChristianKl❫ 07:59, 1 October 2019 (UTC)
@ChristianKl: If fatigue (Q9690) is the symptom then is fatigue (Q15729017) more general/colloquial tiredness/fatigue? It definitely should be clearer what should link where - and fatigue (Q15729017) should not be marked as said to be the same as fatigue (Q9690) and the talk page should maybe explain how they are different. @Infovarius: Maybe you can weigh in here given you understand Russian and the two items at least have different Russian labels? Iwan.Aucamp (talk) 11:56, 2 October 2019 (UTC)
Yes, that's at least how dewiki sees the difference between the two. I agree that it can make sense to have better descriptions and maybe an explanation on the talk page. "said to be the same" however does seem appropriate to me and gets frequently used for cases like that where the meaning is very similar. ChristianKl❫ 16:01, 2 October 2019 (UTC)
If this is the case does fatigue (Q15729017) subclass of (P279) fatigue (Q9690) make sense ? It seems like the more general and colloquial concept of fatigue should not be a ssubclass of (P279) the specific symptom. If any subclass relation makes sense to me it seems it would be fatigue (Q9690) subclass of (P279) fatigue (Q15729017) though I'm not sure if there should be subclass relation between them either way. Iwan.Aucamp (talk) 11:36, 3 October 2019 (UTC)

Tobias1984
Doc James
User:Bluerasberry
Wouterstomp
Gambo7
Daniel Mietchen
Andrew Su
Peter.C
Klortho
Remember
Matthiassamwald
Projekt ANA
Andrux
Pavel Dušek
Was a bee
Alepfu
FloNight
Genewiki123
Emw
emitraka
Lschriml
Mvolz
Franciaio
User:Lucas559
User:Jtuom
Chris Mungall
ChristianKl
Gstupp
Geoide
Sintakso
علاء
Dr. Abhijeet Safai
Adert
CFCF
Jtuom
Lucas559
Drchriswilliams
Okkn
CAPTAIN RAJU
LeadSongDog
Ozzie10aaaa
Sami Mlouhi
Marsupium
Netha Hussain
Abhijeet Safai
ShelleyAdams
Fractaler
Seppi333
Shani Evenstein
Csisc
linuxo
Arash
Morgankevinj
Anandhisuresh
TiagoLubiana
ZI Jony
Viveknalgirkar
Pictogram voting comment.svg Notified participants of WikiProject Medicine

Proposals

Below is a list of changes we could make. Please indicate support with Symbol support vote.svg Support/Symbol oppose vote.svg Oppose/Symbol neutral vote.svg Neutral or add additional proposals. I will add more later. Iwan.Aucamp (talk) 11:54, 3 October 2019 (UTC)

Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support Iwan.Aucamp (talk)
Symbol support vote.svg Support to all your ideas. --Liuxinyu970226 (talk) 00:18, 8 October 2019 (UTC)

Strange book entries

Is it possible to tidy up the weird data entry seen here and track down where that all came from (there may be more, I found these on a search for 'University of Minnesota Press'):

It looks like someone did a bad batch of data entry and copied a university press catalog without pulling the data apart and entering it properly. Carcharoth (talk) 17:06, 1 October 2019 (UTC)

@Carcharoth: I checked a few; if my sample is a good one they seem to be NOT book entries, but the titles of review articles written about the books. As such we generally use the article title as the label, though maybe a more descriptive label like ("Review of xxxx") would be better, and then move the exact title to the value of a statement. These could probably be easily fixed up with a Quickstatements batch if there's a consensus to make that change. ArthurPSmith (talk) 17:12, 1 October 2019 (UTC)
Oh, I get it now. Yes, this is the weird way some book reviews are titled, putting all the pricing information in the article title. May not be much we can do about that. It does look strange though. Carcharoth (talk) 17:19, 1 October 2019 (UTC)
  • @Richard Nevell: it seems you created at least some of the items in question. ChristianKl❫ 19:08, 1 October 2019 (UTC)
    • They are indeed reviews. The naming is a bit odd, but does at least reflect how the book reviews are titled. Richard Nevell (talk) 07:12, 2 October 2019 (UTC)
      • It seems to me very unlikely that the title of those review includes a price or it's ISBN. ChristianKl❫ 07:44, 2 October 2019 (UTC)

These are reviews. Titles seem correct. Nomen ad hoc (talk) 20:18, 1 October 2019 (UTC).

  • It's useful to have items for individual books reviews, even if they occur in part of a larger section. We have a property review of (P6977) to connect them to the work they are reviewing.
As for the labels, it's perhaps worth remembering that there's no requirement for the label to match the full title of the article. The title as given can (and should) be recorded in a title (P1476) statement. But, as Jura proposes, as for the labels, it may make a lot of sense to normalise these into a form that is likely to be more convenient, more reusable, and more meaningful as a search result. Jheald (talk) 13:32, 2 October 2019 (UTC)
Maybe shortening the labels of the long list given initially to "review of"+title of book would do. --- Jura 11:16, 4 October 2019 (UTC)
Interesting... What do others think? Nomen ad hoc (talk) 13:55, 5 October 2019 (UTC).
I’d love to have a standard for shortening labels of book reviews. They drive me nuts. - PKM (talk) 23:46, 7 October 2019 (UTC)
☺ I left a note about the proposal on Help_talk:Label#Review_of. --- Jura 09:22, 8 October 2019 (UTC)
Symbol support vote.svg Support shortening to "Review of (original title)"; descriptions will need to be modified to disambiguate when there are multiple such reviews. ArthurPSmith (talk) 14:23, 8 October 2019 (UTC)

Q19188 vs. Q7262427

Well, for several years I really doubt if there are differents between both items, to indicate that why Wikidata should have separated items for both things. If Mainland China (Q19188) already doesn't include Hong Kong (Q8646)+Macau (Q14773), then why there's benefits to have Mainland China (Q7262427) and their linked articles? What's the actual relation between both items? 1. Q19188 never includes HK+MO, so Q7262427 has two Wikimedia duplicated page (Q17362920) articles, so we should consider merging zh:中国内地 and zh-yue:中國內地 back to Q19188 articles (zh:中国大陆 and zh-yue:中國大陸); 2. Q19188 never includes HK+MO, but there are reasons to hold up mergings, so Q7262427 is a Wikimedia permanent duplicate item (Q21286738) of Q19188; 3. Q19188 should (or even must likely) include HK+MO, but Q7262427 should not, so Mainland China (Q7262427) part of (P361) Mainland China (Q19188) (and do so for HK+MO), and then add reverse statements (i.e. Mainland China (Q19188) has part (P527) Mainland China (Q7262427), and Q8646+Q14773 should also be P527 values); 4. both Q19188 and Q7262427 may or may not include HK+MO, so keep the de facto said to be the same as (P460) linking each other. --Liuxinyu970226 (talk) 08:56, 6 October 2019 (UTC)

As far as I can tell, Q7262427 is a definition in chinese law. Why are you asking here anyway? This is an issue that is strictly for zh:wp to decide. Circeus (talk) 21:57, 6 October 2019 (UTC)
@Circeus: Actually, under PRC's law, "中国大陆" can include HK+MO, but under Taiwanese law, "中國大陸" can't. --Liuxinyu970226 (talk) 00:22, 8 October 2019 (UTC)
@Jyxyl9, Baomi, SC96: Wondering if they can tell us the differents. --Liuxinyu970226 (talk) 00:36, 8 October 2019 (UTC)

Both "中国大陆" and "中国内地" are translated to "Chinese mainland" by Chinese goverment, they referred to a same geographical scope, both can't include HK+MO. like one's "father" is the "husband" of his mother, but the "son" of his grandpa. "Chinese mainland" or "Mainland China" should be called as "中国内地" when it was mentioned with HK+MO in Chinese, but as "中国大陆" when with Taiwan.--Baomi (talk) 12:20, 8 October 2019 (UTC)

Wikidata weekly summary #385

P7375

Why MWNF ID (P7375) is not listed among external identifiers? Visite fortuitement prolongée (talk) 20:23, 7 October 2019 (UTC)
Because the data type is URL, and unfortunately nobody who participated in the discussion objected to the proposal of creating the property that way. ChristianKl❫ 08:11, 8 October 2019 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Apparently not, given that two of the examples are:

note the different sub-domains: "islamicart" vs. "sharinghistory". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 09:30, 8 October 2019 (UTC)

Country property for paintings/monuments

The Immaculate Conception (Q29117926) is a Spanish painting that resided in Spain for the first ~200 years of its existence and now resides in a church in France, where it is a designated historical monument. I took off the country (P17) property because I wasn't sure that France (the old value) was appropriate, though I added location (P276) = Église Saint-Gervais-Saint-Protais de Langon (Q3581683). (I also added location of final assembly (P1071) = Madrid.) heritage designation (P1435) wants me to add country (P17), and Palissy ID (P481) is complaining that it should be country (P17) = France. What is the best way to handle this? Is there an established way to treat paintings? I didn't see anything definite at the visual arts project. Calliopejen1 (talk) 17:41, 7 October 2019 (UTC)

There are a couple of options I can think of: 1) use country (P17) = France qualified with a start date and country of origin (P495) = Spain. 2) Use country (P17) = France and country (P17) = Spain with start and end dates, and set the “Spain” value as deprecated. I suppose you could even do both of these. Personally, I like the first option. IMHO, country (P17) should always be the country where the item is presently located. - PKM (talk) 19:02, 7 October 2019 (UTC)
Thanks, User:PKM. It is unknown how and when exactly it got to the French church; a priest discovered it in the sacristy in 1965 and it apparently presumably had been there for some time. But all we know is that it was definitely in Spain in 1837 and was definitely in the French church in 1965. What kind of qualifier(s) would I use to model this? Calliopejen1 (talk) 21:14, 7 October 2019 (UTC)
point in time (P585) = 1837 and 1965, respectively, with citations to the references that state this (and I'd add a quote from your source if it's available). - PKM (talk) 21:28, 7 October 2019 (UTC)
I just did something slightly different before you responded... Let me know if you have any objections! Calliopejen1 (talk) 21:29, 7 October 2019 (UTC)
@Calliopejen1, PKM: Deprecated rank must only be "used for statements that are known to include errors" (see Help:Ranking). Thus, it is not adequate in this case. (Otherwise, all seems OK) Ayack (talk) 08:19, 8 October 2019 (UTC)
@Ayack: Thanks! Calliopejen1 (talk) 15:58, 8 October 2019 (UTC)

determination method (P459) qualifier for number of words (P6570)

What are the possible values of determination method (P459) when qualifying number of words (P6570)? Levana Taylor (talk) 11:37, 8 October 2019 (UTC)

  • You can specify the method or tool used. --- Jura 12:51, 8 October 2019 (UTC)
right, but I can't find the list of methods and tools, so that was my questioin. Levana Taylor (talk) 13:08, 8 October 2019 (UTC)
Last time I used word count in Libre Office. --- Jura 13:12, 8 October 2019 (UTC)
what is the item for "word count in Libre Office"? there is none by that exact name. My point is that I can't find a valid value to enter into the statement. Levana Taylor (talk) 13:30, 8 October 2019 (UTC)
What method did you use? --- Jura 13:40, 8 October 2019 (UTC)
TextFX for Notepad++, which (by their own say-so) uses the same word segmentation rules as Microsoft Word. Levana Taylor (talk) 13:43, 8 October 2019 (UTC)
I'd create an item for TextFX. Obviously, creating an item about that actual segmentation method used would be better, but I couldn't find a good description for LibreOffice Writer's either. --- Jura 13:49, 8 October 2019 (UTC)
OK, here counted by TextFX (Q70092255). That should work. Levana Taylor (talk) 14:00, 8 October 2019 (UTC)
Nice. This ensures that it can be reproduced. --- Jura 14:08, 8 October 2019 (UTC)