Wikidata:Project chat/Archive/2015/01

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.

Is Snow White the fictional human the subject of Snow White the fairy tale?

Are there any guidelines how to fill out fictional characters? See Snow White (Q2739228) . Thanks, Jane023 (talk) 13:30, 30 December 2014 (UTC)

present in work (P1441) should be used for things like that. Sjoerd de Bruin (talk) 13:36, 30 December 2014 (UTC)
Thanks! I have another question - is she a female human or not? Jane023 (talk) 15:30, 30 December 2014 (UTC)
female fictional human. Kaldari (talk) 21:02, 30 December 2014 (UTC)
OK got it now. Jane023 (talk) 10:24, 31 December 2014 (UTC)
You'd need an item for the non-Disney version of the character. Here it is: Q18695716. --- Jura 04:13, 31 December 2014 (UTC)
Interesting - I never thought of these as being separate entitites! Thx Jane023 (talk) 10:24, 31 December 2014 (UTC)
At least at Wikipedia, they have to be, otherwise we would have a Disney character appear in Grimm's tales. ;) --- Jura 10:34, 31 December 2014 (UTC)
Well, they don't have to, we just have it this way. We can have a more general item like "Snow White character in creative works", but this is for another discussion (which doesn't seem to come to some conclusion soon). Or we can have more specific items like Disney's Snow White character in movies, Disney's Snow White character in TV shows, Disney's Snow White character in Disneylands, Disney's Snow White character in video games...--Shlomo (talk) 09:36, 1 January 2015 (UTC)
With "have to", I was referring to the absence of an article about the character "Snow White" and Wikipedia's article en:Snow White (Disney) (and the similar in the other languages). --- Jura 09:40, 1 January 2015 (UTC)
We do have articles about the character "Snow White" on Bulgarian, Czech and Latin Wikipedias (at least, maybe more...) They are connected to the fairy-tale item Snow White (Q11831), which is not correct, but understandable due to the interwiki issue. But there surely can be an item for the character here as well.--Shlomo (talk) 10:11, 1 January 2015 (UTC)

Bestiaire d'amour

Please unify Bestiaire d'Amour (Q18384900), Bestiaire d'Amour (Q13125996) and Bestiario d'amore (Q3639061) --Sailko (talk) 18:15, 1 January 2015 (UTC)

✓ Done --Pasleim (talk) 18:20, 1 January 2015 (UTC)

Palm Sunday

There is a Czech Wikisource category Květná neděle (Q17145192) and I wanted to add Commons:Category:Palm Sunday to it. However, this category is already attached to Q42236 and so my attempt failed. Can cs.wikisource category Květná neděle be added to Q42236, althought there are only pages, not categories? Or is there any other way how to attach the cs.wikisource and Commons categories? Jan.Kamenicek (talk) 02:11, 31 December 2014 (UTC)

There is some debate about this. Q422236 seems the better option. --- Jura 10:31, 31 December 2014 (UTC)
But I noticed that e. g. both Q8967576 and Q1085 are both linked to Commons:Category:Prague. So is it possible to link both Q17145192 and Q42236 to Commons:Category:Palm Sunday? Jan.Kamenicek (talk) 01:37, 1 January 2015 (UTC)
No, you have to choose. --- Jura 08:30, 1 January 2015 (UTC)
OK. I am just curious, why is it possible with Q8967576 and Q1085. Jan.Kamenicek (talk) 15:31, 1 January 2015 (UTC)
Prague (Q1085) is linked with the gallery c:Praha. Category:Prague (Q8967576) is linked with the category c:Category:Prague --Pasleim (talk) 15:40, 1 January 2015 (UTC)
Well, I can see, that Prague (Q1085) is linked both to Commons category and Commons gallery. Jan.Kamenicek (talk) 18:26, 1 January 2015 (UTC)
There is a difference between site links and properties, Jan.Kamenicek. Sjoerd de Bruin (talk) 18:36, 1 January 2015 (UTC)
Now I see, thanks a lot. Jan.Kamenicek (talk) 01:12, 2 January 2015 (UTC)

SIPA

Do we not have a property for Sistema de Informação para o Património Arquitectónico (SIPA) identifier for monuments in Portugal? There are a number of these monuments identified with a template in Commons, with categories and Institution templates. I couldn't find any discussion related to this in Wikidata. - PKM (talk) 19:53, 1 January 2015 (UTC)

It looks like we don't have it. So you or someone else should then propose it at Wikidata:Property proposal. --Stryn (talk) 20:19, 1 January 2015 (UTC)
@PKM: I remember from Wiki Loves Monuments that Portugal has quite a complicated structure. See the template at Commons for details. You might want to discus it first with the people of Wikidata:WikiProject Cultural heritage before proposing properties.
I just noticed that the monuments database doesn't contain Portugal anymore (used to be there). I guess it got lost in the list namespace move..... Multichill (talk) 16:12, 2 January 2015 (UTC)

vandalism

For a while I took it upon myself to monitor RecentChanges for vandalism. It was hard, but doable, and I usually found about one to two hundred cases of vandalism every day.

The incremental update for December 27th showed 104 undos, reverts, or restorations. Probably there were another hundred that the community, if there is such a thing, had missed.

There were 30 users who did reversion actions on that day. Only two users did more than ten. The rest of you should be ashamed of yourselves. Wikidata is the greatest project of our generation, and you are letting it rot. --Haplology (talk) 11:54, 28 December 2014 (UTC)

I agree that there's a lot of test/vandalism edits, but it's almost impossible to find all of them. At least it takes a lot of time (I know since I reverted a lot of those many months ago). Nobody should be ashamed of not reverting vandalism. Wikidata has many other things to do as well. But work that you do is very appreciated. --Stryn (talk) 12:04, 28 December 2014 (UTC)
Do we have a page that lists and explains the ways people can help fight vandalism? I am happy to spread that through the weekly summary and our social media accounts. Maybe that'll make a few more people aware of how they can help. --Lydia Pintscher (WMDE) (talk) 14:59, 28 December 2014 (UTC)
@Lydia Pintscher (WMDE):: We do... Matěj Suchánek (talk) 15:17, 28 December 2014 (UTC)
Thanks! That's good. Could mention the constraint reports as well? And recent changes tags? Either way I'll add it to the next weekly summary and in a few days push it to social media. --Lydia Pintscher (WMDE) (talk) 15:20, 28 December 2014 (UTC)
Please, Lydia, don't forget also this new project page about AbuseFilter. Matěj Suchánek (talk) 15:22, 28 December 2014 (UTC)
Do you mean Wikidata:Abuse filter? Will add. Thanks! --Lydia Pintscher (WMDE) (talk) 15:24, 28 December 2014 (UTC)
Haplology, I made 16 reverts yesterday (as well as 16 again today, coincidentally) as far as I can tell. Having said that, I do believe your posting is a little harsh. Everyone has their own interests, these change with time too. If I choose to work on another project in a couple of days' time, does that make me a bad person? Of course it doesn't. We have less than 200 users with the ability to rollback, half of those being Wikidata:Rollbackers, the other half admins. Admins have plenty they could be getting on with, and rollbackers are voluntary users giving up their free time for the project too. Given that the site garners a lot of editing traffic, second to possibly only the English Wikipedia, I don't think we're doing such a bad job altogether. The more that Wikidata grows, the more users will vandalise, the more rollbackers there will be. It's a lot easier to find and destroy vandalism on sight than it is to think of something stupid to add, believe me. And there will always be sneaky vandalism that's hard to sniff out. There's no point worrying about that too much either. So all in all, I wouldn't worry about what other people, assuming good faith, are doing. I'm very happy with the work you have done to combat this problem and wouldn't want to see you stop or think that nobody is helping, but as Stryn pointed out, Wikidata has an awful lot of things to be taken care of, and they're not more (or less!) important than this. We are all doing our bit to make this the "greatest project of our generation", (your words, my sentiment)! But everyone has to find their niche. A wiki is a collaboration, after all. Jared Preston (talk) 22:01, 28 December 2014 (UTC)

Would it ever be possible for Wikidata to have our own version of ClueBot? This would at least help reduce the work of admins and rollbackers. Del♉sion23 (talk) 22:24, 28 December 2014 (UTC)

I think we really need a better way to fight vandalism than Special:AbuseLog. I don't like looking at that page because one can't see, which cases were already taken car of. We need a more interactive page, where one can mark cases as resolved. If we can somehow gamify it, that would help too. --Tobias1984 (talk) 10:54, 29 December 2014 (UTC)

If you have no automated tools verifying information you will not be able to detect all vandalism, see : #Data model - consistency checks - detecting false information. MeldeDichAnUmEineAndereMenüspracheAuszuwählen (talk) 17:03, 3 January 2015 (UTC)

Request for some help

Hello all, what a great service! I am a newbie here but an editor on EN wikipedia for over a year. I am trying to get a Wikidata property for anatomy templates. I don't know how to insert one. I'd be very grateful for some help.

My specific question is:

How do I add this? Gratefully, Tom. --LT910001 (talk) 05:29, 2 January 2015 (UTC)

@LT910001: Sorry but I don't understand what you want to do. First there is no item (no entry) about Template:Meninges. Then Terminologia Anatomica 98 is not a property or a descriptor of the template. Finally Wikidata doesn't aim to generate navigation template. Snipre (talk) 11:40, 2 January 2015 (UTC)
@LT910001: Maybe this helps: Currently a wiki-page can only access information from its connected data-page. If the template needs to access this TA98-value it has to wait for data access from arbitrary items (See: Wikidata:Development_plan#Access_to_data_from_arbitrary_items). --Tobias1984 (talk) 12:49, 2 January 2015 (UTC)
If you want to use Terminologia Anatomica 98 ID (P1323) in an infobox like the one at en:Template:Infobox anatomy, you need to include P1323 in a similar way as it's done there.
This can work for an article that uses such an infobox and has an item at Wikidata with Terminologia Anatomica 98 ID (P1323). The one for en:Meninges is at Q268930. --- Jura 13:04, 2 January 2015 (UTC)
Thanks for your help. I think I have figured that part out now. Now to move on the the bigger questions :P. --LT910001 (talk) 21:19, 2 January 2015 (UTC)

Preventing to add more than one Badge

We should have an abuse filter to prevent users to adding more than one badge to an item for one language like this! Yamaha5 (talk) 15:55, 2 January 2015 (UTC)

Abuse filter requests are handled here: WD:AF --Tobias1984 (talk) 16:03, 2 January 2015 (UTC)
Featured on the front page and an exelent article are different and should be kept as such. 109.247.163.112 08:49, 4 January 2015 (UTC)

population (P1082)

Is it possible that this property gives you an alert if you accidentally insert two same numbers or dates? --Impériale (talk) 22:53, 3 January 2015 (UTC)

The idea is to be able to put several population numbers at different times, so I don't think this will be possible. But external checks of the items can be done by bots. Snipre (talk) 17:47, 4 January 2015 (UTC)

Wikidata weekly summary #139

Here's your quick overview of what has been happening around Wikidata over the last week.

Demonym uncertanties

I've been adding demonym (P1549) to various countries, but I can't determine from either the property itself or the wiki page on the same concept whether I should use the noun or the adjective. For most countries this isn't a problem, they're identical, but for Spain, should I use "Spanish" or "Spaniard"? (I've been working from this list) Popcorndude (talk) 01:26, 4 January 2015 (UTC)

I would personally use the term "Spanish", since Spaniard is only applicable to humans. Or, alternatively, you could just add both. George Edward C – Talk – Contributions 09:02, 4 January 2015 (UTC)
The English description says the "singular masculine adjective" is preferred. --- Jura 09:10, 4 January 2015 (UTC)
Thanks, I didn't see that. Popcorndude (talk) 20:18, 4 January 2015 (UTC)

Links to Wikipedia

A quick question on how Wikidata handles Wikipedia links if the Wikipedia article disappears. Is there anything in place to update Wikidata? What happens if...

1) the article turns into a redirect. 2) the article is deleted due to lack of nobility. Periglio (talk) 15:04, 4 January 2015 (UTC)

  1. Nothing.
  2. It is a new feature that the sitelink is also removed.
Matěj Suchánek (talk) 15:21, 4 January 2015 (UTC)
So I can add my pet cat to Wikidata and it will never be deleted? Periglio (talk) 16:15, 4 January 2015 (UTC)
Wikidata has its own rules for keeping or deleting items. Matěj Suchánek (talk) 16:34, 4 January 2015 (UTC)

Statement with qualifier "applies to part" (P:P518) "Russian Wikipedia"

To cover content that is only combined in Russian, Q1450800 includes statements with the qualifier "applies to part". If the content is different in one language, wouldn't that be a sign that the page of that language shouldn't be linked from the same Wikidata item? If we start adding "applies to part" for every Wikipedia site, items could have just any meaning. --- Jura 22:20, 29 December 2014 (UTC)

I'd prefer to have ru:Франсуа in q1166840 instead but you don't like it. So I've chosen the way with qualifiers. Anyone is encouraged to tell the better way how to mark that the ru-page is the only page on ru-wiki about this name. --Infovarius (talk) 23:32, 29 December 2014 (UTC)
A clear sign that ru:Франсуа is linked to the wrong Wikidata item. /ℇsquilo 11:05, 5 January 2015 (UTC)

Data model - consistency checks - detecting false information

1) Are there any tools that check the consistency of claims in Wikidata?

Examples could be:

  • checking on global properties, e.g. checking type-token differences
    • anything that is a class should not have a location and not have a time
    • anything that has a location or a time should be an instance
  • checking location and time consistency, e.g. the location should have existed during the time any item was located there

2) Are there any tools that verify that the given source of a claim really claims what is said in Wikidata?

If this is fully implemented, any false or unsourced information could be whiped out from Wikidata.

Do the Wikidata bosses care about such high-level accuracy? MeldeDichAnUmEineAndereMenüspracheAuszuwählen (talk) 16:59, 3 January 2015 (UTC)

There are constraints which are doing this kind of work. Look at the talk page of any properties to see the use of these constraints (see this one for example with the corresponding report of unconsistencies.
For the second question WD is like WP: no possibility check if the data imported is really coming from the added reference. How can you create a tool to check if a book or a moivie is really displaying a certain information ? A large part of the references in WD will be books or other hard copies so no tools can do the checking work.
For online data, the only possibility is to run regularly some checking processes comparing wikidata and the original sources. But this has to be perform by some external bots. Snipre (talk) 17:55, 4 January 2015 (UTC)
We actually have a team of students work on this now. Sneak peak at https://twitter.com/wikidata/status/539891914922008577. This will be integrated nicely in the Wikidata UI. --Lydia Pintscher (WMDE) (talk) 17:19, 5 January 2015 (UTC)

Wikidata London Meetup 2

Just a reminder for anyone in or around London, we are having our second Wikidata London Meetup, it would be great to meet more Wikidatans. One item on the agenda will be the Wikidata:WikiProject Wikidata for research as well as anything else people feel is relevant. Also check the list for organising Wikidata activities in the UK/British Isles. Fabian Tompsett (WMUK) (talk) 14:32, 5 January 2015 (UTC)

Salvatorkirche

Can a German speaker advise whether Salvatorkirche should be translated as Church of Saint Salvator or Church of Our Saviour in English? Thanks! - PKM (talk) 04:51, 6 January 2015 (UTC)

Church of Our Saviour, unless it is specifically concecrated to Saint Salvador. (I am not a native speaker though).--Ymblanter (talk) 19:24, 6 January 2015 (UTC)
@Ymblanter: thanks! - PKM (talk) 20:07, 6 January 2015 (UTC)

Connect date of birth to birth

What statement should be added to date of birth (P569) to indicate that it represents a date of parturition (Q3335327)? We have many date of ... and place of ... properties which could use this. Petr Matas 11:53, 6 January 2015 (UTC)

How about facet of (P1269), which I have just found? Petr Matas 12:45, 6 January 2015 (UTC)

(SOLVED) Help me: Created new persons and can't add first statement

Hi. I am new to wikidata - last night. I have been working through some en wikipedia articles I have edited and adding/checking Authority controls. I added VIAF and ISNI statements within wikidata and the Authority control template within wikidata. I also found and fixed a few obvious errors. All good.

I also identified a some persons present in wikipedia, with VIAF numbers but not in wikipedia. I have added:

  • John William Springthorpe (Q18715199)
  • Charles Robbins (Q18715132)
  • William Adams Brodribb (Q18715993)

I can't work out how to add the first statement. When I select add under "Statements", then click in the property field box I don't get a drop down box of items (as I do if statements are already present). What am I doing wrong?

Billingd (talk) 00:17, 7 January 2015 (UTC)

Worked it out myself. I need to type instance of for property and human for value. Billingd (talk) 00:33, 7 January 2015 (UTC)

Bug when adding links to Wikinews

Tracked in Phabricator
Task T85754

For example: when I try to add n:ru:Категория:Крымские татары to Crimean Tatars (Q117458) interface gives the error: "The specified article could not be found on the corresponding site. The external client site 'ruwikinews' did not provide page information for page 'Категория:Крымские татары'". Fix it please. --sasha (krassotkin) 09:46, 6 December 2014 (UTC)

Did you try to add the link to Q117458#sitelinks-wikinews? Also, ruwikinews page seems to be a category, so I don't think that it belongs on that item. --Stryn (talk) 14:11, 6 December 2014 (UTC)

Colleagues, how things are going with this? We have to stop all work:(. --sasha (krassotkin) 10:34, 9 December 2014 (UTC)

In addition. Now I could not add regular article n:ru:Беларусь восстановила таможни на границе с Россией (from mainspace) to Q18636241. As idea. Maybe this is due to enable HTTPS by default for all users in Russian Wikinews as the first project? See: n:ru:Русские Викиновости перешли на HTTPS, Gerrit Commit; also: meta:HTTPS/Beta program. --sasha (krassotkin) 16:11, 14 December 2014 (UTC)

I think this is a bug, but not sure. I also get errors on adding ruwikinews (no errors on enwikinews, kowikinews) sitelinks. — Revi 04:25, 20 December 2014 (UTC)
I too have have this error. It turned up on the first item I created (Multiple fatalities in bin lorry crash in Glasgow, Scotland (Q18668596)) and I was quite worried I'd done something wrong. :P Blood Red Sandman (talk) 08:04, 23 December 2014 (UTC)
So only Wikinews category<->Wikipedia article is correct? Blood Red Sandman (talk) 08:25, 23 December 2014 (UTC)

What to do if there are more than one wiki news articles in the same language? Usain Bolt crowned 100m world champion (Q17957557) and Q17595144 go in Usain Bolt (Q1189) but there is already an frwikinews article. Usain Bolt breaks 200m and 100m sprint records at championships (Q17562449) has an enwikinews and a plwikinews article. Usain Bolt sets new world record in 100m sprint (Q17683152) has articles that can not be included. 24.60.49.130 07:55, 23 December 2014 (UTC)

I am new to Wikidata and feeling my way around but it seems to me your approach is not correct and those articles cannot be added. In the enwikinews case (which is where I usually edit) I know we have a category for Usain Bolt (Q1189) which I have linked in. Blood Red Sandman (talk) 08:04, 23 December 2014 (UTC)

Now I again can edit sections of q-items, but can not add pages of Russian Wikinews with error: "The specified article could not be found on the corresponding site" and details "The external client site 'ruwikinews' did not provide page information for page...". It took more than a month as we can not use Wikidata. --sasha (krassotkin) 08:44, 4 January 2015 (UTC)

This is indeed cause by the https only change for ruwikinews. I've just uploaded a patch to a script that we use to generate data about sites so that we can change things over to https. Progress for that (and for actually regenerating that data) is being tracked in T85754. I hope to get that fixed tomorrow. - Hoo man (talk) 10:49, 4 January 2015 (UTC)

Is access to sources an issue for Wikidata contributors and verifiability?

Hi all,

I know this page (as well as the listservs, IRC channel, and talk pages) has hosted quite a number of discussions about how to best support the work of adding more sources to Wikidata, with ideas ranging from the development of new tools to the renegotiation of current ideas and practices of what should (and should not) be considered an authoritative source.

With more recent conversations about the Freebase donation and the design of the Primary sources tool, I thought it might also be a good time to get some perspective on what aspects of sourcing - besides ease and convenience - contributors are having difficulty with. In particular, I'd be interested to know whether people feel like they might benefit from infrastructure or networks similar to Wikipedia's Resource Exchange or the Wikipedia Library?

For those unfamiliar with these initiatives, the Resource Exchange is a WikiProject that allows Wikipedians to request sources like articles and book chapters from other Wikipedians who already have access to them, while the Wikipedia Library coordinates free access to resources like paywalled databases for Wikipedians. Both initiatives help ensure that lack of access to sources is not an obstacle to the creation of verifiable content. It may be possible to explore options of whether something like the Wikipedia Library could also serve Wikidata contributors, however I am not sure whether or not access to sources is seen as an issue for our community. What does everyone else think? Is access to authoritative sources a challenge for you when adding data or not?

Feel free to comment here, or add your own ideas to the Wikidata:Verifiability#Finding_sources section I have been working on.

Cheers. -Thepwnco (talk) 21:26, 4 January 2015 (UTC)

Sorry but your coment is far beyond the principle currently in effect in Wikidata: to think what kind of source can be used, it has first to be accepted the idea that sources are necessary, and this, this is not yet well accepted: after the large importations of data from wikipedia without any sources, we can assist to a new large data import from Freebase again without any source. Snipre (talk) 22:08, 4 January 2015 (UTC)
@Snipre: while I agree that adding sources to Wikidata may not be seen as necessary or at least not a priority by many contributors, my understanding was that we are hoping one day to be in a place where it is common practice; providing helpful resources and information about how to find and add sources seemed to me a good step for encouraging such behaviour. thanks. -Thepwnco (talk) 23:46, 6 January 2015 (UTC)
Yes, adding real, reliable sources is the intent, but at this stage it does not happen a lot. Most likely, Wikidata will not be anything like mature until this has become commonplace, but there is no garantee we will ever get there. At the moment it is all about importing things and connecting dots. - Brya (talk) 06:39, 7 January 2015 (UTC)
The big holdup for me is that a printed reference requires a work, edition, publisher, and author(s), none of which may exist when I want to cite something. Until we have a tool or gadget that allows us to enter an ISBN + page number and have everything else automatically imported behind the scenes, referencing will be difficult to focus on. - PKM (talk) 16:38, 7 January 2015 (UTC)

Wikinews links

(Sorry for my bad english)

Hi !

After a discussion in August I see this Wikidata:Wikinews/Development so Wikinews categories can be linked to Wikipedia articles ? I think this is a good thing but I don't know if it's allowed.

Best, --Mattho69 (talk) 19:26, 7 January 2015 (UTC)

Sure, it's allowed, so feel free to add Wikinews pages also! :) Just remember to take account of Wikidata:Notability. --Stryn (talk) 19:31, 7 January 2015 (UTC)
@Stryn: There are still users saying "categories to categories!" because of whom I haven't started connecting so far. So is this really sure? Matěj Suchánek (talk) 19:40, 7 January 2015 (UTC)
Oh, I read Mattho69's message wrongly, sorry. I have no idea how should Wikinews' categories connected. I'm not really an expert with Wikinews' things. --Stryn (talk) 19:44, 7 January 2015 (UTC)

Next office hour

Hey folks :)

I'll be doing another office hour for Wikidata on IRC next Friday at 18:00 UTC. It'll happen in #wikimedia-office on Freenode IRC. Everyone is welcome for discussions and question answering. Logs will be posted afterwards. For your timezone see here. Hope to see many of you there.

Cheers --Lydia Pintscher (WMDE) (talk) 19:26, 7 January 2015 (UTC)

Again? Awesome!
+1. Matěj Suchánek (talk) 19:41, 7 January 2015 (UTC)
Hehe yeah I figured there is a lot to discuss at the moment. --Lydia Pintscher (WMDE) (talk) 19:57, 7 January 2015 (UTC)

References import: Template:Cite web

Hello, I have some issue during references import from Template:Cite web (Q5637226). Per parameter table:

template parameter property issues
url reference URL (P854)
title (language is known) title (P1476) The property is not marked for usage in reference section. Is it correct usage?
title (language is unknown) P357 (P357) The property is marked as deprecated. How to specify title if title language is unknown? In progress: [[1]]
author author (P50) The property is not marked for usage in reference section.
date publication date (P577) The property is not marked for usage in reference section.
publisher published in (P1433) or publisher (P123) The property is not marked for usage in reference section.
accessdate retrieved (P813)
archiveurl archive URL (P1065)
archivedate ? Is there appropriate property? no need because the date of archive is provided when you follow the link archiveurl

Additional question: need this table be extended? or these properties are enough? — Ivan A. Krestinin (talk) 21:12, 7 January 2015 (UTC)

Yay for importing references! :) --Lydia Pintscher (WMDE) (talk) 21:14, 7 January 2015 (UTC)
See Help:Sources for the minimal set of properties used in references. Snipre (talk) 22:05, 7 January 2015 (UTC)
Help:Sources is too minimalistic as I see. And it does not reflect compatibility questions of Wikipedia sourcing standards/traditions and Wikidata sourcing. — Ivan A. Krestinin (talk) 22:49, 7 January 2015 (UTC)

You could add original language of film or TV show (P364) for language and title (P1476) for trans-title. See also en:Template:Cite web#Examples Foreign language and translated title. --JulesWinnfield-hu (talk) 00:11, 8 January 2015 (UTC)

COI and editing

I was just checking our policies, and I could not find anything regarding Conflict of Interests while editing. In Wikipedia, the rules are quite clear - but I don't think simply taking the same guidelines makes sense. So, am I allowed to edit the items about good friends, myself, my projects, or my employer, etc.? For now, I assume yes - as said, I could not find a policy otherwise - but I left a message on the talk page of the most obvious cases of possible COI.

In Wikipedia I can see that a strong stance against editing such articles makes sense. Formulation, length of sentences, the position and prominence of arguments -- all these fine details can be easily influenced by editing text. But on Wikidata? It is just facts. As long as I follow the other policies - verifiability, notability, etc. - it should be fine.

On the other hand, removing statements that I might not like but that are verifiable, that should not be OK.

Also, imagine if one day say a movie publisher decides to add a lot of data about their movies in Wikdiata and keep it up to date. Would that be OK if they do so? It obviously would not be OK if they were taking over the maintenance of the Wikipedia articles for the movies. But what about the Wikidata items? If they keep the cast list, the lengths of movies, etc. correct?

I am afraid we need a bit of a policy here. I am pretty sure we already had this discussion, but I cannot remember the outcomes right now. Any thoughts? --Denny (talk) 17:13, 23 December 2014 (UTC)

@Denny: This sounds like we need one of the dreaded multilingual RfCs. --Tobias1984 (talk) 17:17, 23 December 2014 (UTC)
@Denny:, I totally agree with you. As long as people are adding data that meets Wikidata's referencing standards, I think it would be fine for a person or company to edit their own page here. (I also agree they shouldn't delete true things just because they don't like them.) I fear I agree with @Tobias1984: that an RfC would be in order. Sweet kate (talk) 21:36, 23 December 2014 (UTC)
In Wikipedia, there are PoV issues. Here, generally, there are no PoV issues, and I do not see much harm from CoI editing (I actually once edited an item about me, adding an ID or smth which was missing). I think as far as the edits do not have problems with verifiability (like changing the birth year or smth counter to RS) they should be allowed. Creating of non-notable items should not be allowed.--Ymblanter (talk) 10:29, 24 December 2014 (UTC)
Yes, we'd want a local policy to explicitly allow this. Assuming the ToU change in June affects Wikidata as well, we need to develop an alternative paid contribution disclosure policy for Wikidata, otherwise the global banning of undisclosed paid contribution seems to have to apply (at least in theory). Commons developed theirs in response to the change. whym (talk) 11:28, 24 December 2014 (UTC)
I agree with everything that has been said so far. Petr Matas 15:24, 24 December 2014 (UTC)
@Tobias1984: @Sweet kate: @Ymblanter: @Whym: @Petr Matas:
I tried a first draft. Feel free to proceed from here however you like. --Denny (talk) 21:48, 30 December 2014 (UTC)

If we don't have a CoI policy, then the deletion of an otherwise-notable item on the grounds of CoI would surely be an abuse of admin privileges? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:06, 2 January 2015 (UTC)

  • I agree with @Sweet kate: and @Ymblanter: in thinking that COI editing at this time is not a problem. Wikidata right now is less attractive in comparison to other Wikimedia projects as a target for being corrupted by COI problems. Until problems are observed, and my intuition that problems may not be observed for months or longer, I would be comfortable with people editing about themselves, their organizations, and fields in which they have professional bias. The purpose of having COI rules is to deter bad behavior, but a side effect of these rules is cooling good contributing, and I wish to avoid making rules until the seriousness of the problem and the likely loss of benefits of having a strict policy is qualified.
I see no reason to raise this as a project-wide RfC because no serious problem has been identified, and I think that this discussion can wait until there is a test case before the issue is taken too seriously. Before a policy is developed, I might expect to see a few test cases. Wikidata is an active project in terms of numbers of edits but there is not a large community here which really understands Wikidata, and the entire project is particularly inaccessible to outsiders. Without enough people understanding what kind of culture is developing here and without more practical experience in handling COI, I wish to avoid pre-emptively making rules to deter people from contributing.
I consider Denny to be a sort of community role model at this time. As a step toward resolving this, I think it would be useful for Denny to a precedent by disclosing what he did somewhere, and whatever he did for disclosure can be a current best practice. Blue Rasberry (talk) 14:48, 7 January 2015 (UTC)
Thank you for your trust, @Bluerasberry:, but I am not sure I consider myself to a great community role model due to a number of circumstances. So my disclosure is that I made edits to and am watching the items about myself, about my friends and some acquaintances, my current and former employers, my alma mater, my place of origin, etc. I put a note on three talk pages of those items, where it is most abundantly clear, and also linked some pages on my user page. I give editing these items much more consideration, and refrain from things I consider possibly questionable. Basically, I try to follow the rules that I drafted.
The reason why I would like to have rules for this before something happens is that it seems much easier to come to a rational decision on such a policy before bad things happen. Once they happened, overreaction and singlemindedness often may reign. But I will not push for such a policy, and it is up to the community as a whole to recognize whether a policy or a guideline or whatever is necessary or not. --17:22, 8 January 2015 (UTC)

Scaling Wikidata: success means making the pie bigger

Hey folks :)

Happy new year everyone. It is surely going to be an exciting one for Wikidata. Over the last weeks I've been thinking a lot about the year ahead of us. One thing is clear to me: It will be about successfully scaling Wikidata and keeping all the amazing things we have achieved in the process.

I've written down my thoughts on the subject in a blog post to kick off some thinking and discussions: http://blog.wikimedia.de/2015/01/03/scaling-wikidata-success-means-making-the-pie-bigger/


Cheers --Lydia Pintscher (WMDE) (talk) 14:49, 3 January 2015 (UTC)

Instead of many eyes you could use logic. MeldeDichAnUmEineAndereMenüspracheAuszuwählen (talk) 17:06, 3 January 2015 (UTC)

We are using it. And we will be using it more. But it will by far not be enough. There are errors you just can't find this way. --Lydia Pintscher (WMDE) (talk) 17:18, 5 January 2015 (UTC)
Well, you know what they say about logic: "Relying on logic is a sure way of going wrong with confidence". Used in moderation logic is a useful tool, but on the whole logic is the biggest enemy of information.
        Nevertheless, I also disagree with the many eyes. It seems to be a law of Wikipedia that pages that are popular (drawing many eyes) are likely to have many errors, which are put in again and again...
        What I would especially agree with is the An understanding that we don’t have to have it all. There seems to be too much import of redundant links (not adding information, but just links), and not as much in-depth data as would be desirable. - Brya (talk) 05:35, 6 January 2015 (UTC)
Heh interesting observation about high traffic pages. I have the exact opposite experience. Has some research gone into this topic? --Lydia Pintscher (WMDE) (talk) 08:08, 6 January 2015 (UTC)
Research would be difficult, and would require an objective standard. Remember the proud claim that vandalism is reverted in so-and-so-many minutes. That sounded great until it became apparent that this only applied to vandalism found and marked as vandalism, disregarding vandalism not found (and sitting there for perhaps years) and vandalism not marked as such upon reversal. However, it did count reversions of correct edits, marked as vandalism. All in all a hugely-skewed number, a self-fulfilling mechanism.
        And obviously, there are huge differences per field of interest. In a page on a popular television series, or a piece of software, errors are much more likely to be corrected than for a page on a topic where it would be necessary to use not only an actual book (gasp), but a selection of literature. But really silly errors can sit for years in highly trafficked pages (even in the first sentence). - Brya (talk) 11:49, 6 January 2015 (UTC)
Do you have any examples of silly errors sitting in the first sentence or paragraph on a highly read, yet unprotected Wikipedia article for a long time? --Denny (talk) 16:23, 6 January 2015 (UTC)
The longest running I know of has had it for four-and-half years now, added by an IP who could not get his spelling right. The page is getting some three to four thousand views a month. - Brya (talk) 18:18, 6 January 2015 (UTC)
Just curious, which page is that? --Denny (talk) 20:15, 6 January 2015 (UTC)
There is no point in getting into that, but it is a real case, and although it is the front-runner, it is not an isolated case, but part of a consistent pattern.
        But if you want a real case, take cattle, getting something like sixty to a hundred thousand page views a month. This has the scientific nomenclature all wrong (couldn't be more wrong), and refers for details to Bovini, which does not actually go into detail, but which as it gets no more than fifteen hundred to two thousand page views a month can afford to be correct. Technical detail is in Opinion 2027, which, with only a hundred page views a month, has no problems with offering correct information. - Brya (talk) 06:30, 7 January 2015 (UTC)
Wait, is your argument that 'the more people see it, the less it will be correct'? Also, the example, for me, is a bit too arcane to really understand it, but I certainly believe you that such specialized knowledge might be misrepresented in even heavily read articles. Nevertheless, as a working assumption, I think that the many-eyes theory is probably the best one to base decisions on how to develop the Wikidata ecosystem, which is Lydia's mission here - I would not know what else would be a better working assumption. --Denny (talk) 17:31, 8 January 2015 (UTC)
It is in the nature of assumptions that they are just that: assumed, not tested against reality. Assumptions can be deadly. The problem with the many-eyes theory is what is behind those eyes (or what is not behind those eyes). There are many areas where there are wide-spread beliefs that upon testing prove to be wrong (popular myths). For example, after the Germans invaded Western Europe in 1940, there was a widespread belief that the Germans had done that by having more tanks, and tanks that were more heavily armoured and heavily armed that those of the Allies. Complete myth, but very comforting. There is a website somewhere devoted to debunking urban myths as soon as they are invented (creating fun stories is apparently popular). Big industries like to deliberately create myths ("smoking is good for you", etc). Mark Twain is reported to have once said "It ain't what you don't know that gets you into trouble. It's what you know for sure that just ain't so."
        So, it differs per area, how the many-eyes theory works out. It will work fine for the storyline of a popular television series (and who is in it) and for software, probably also for military weapons. It wil work less well, or even in reverse for other topics, depending on how "arcane" a field is, in relation to the composition of the user community. I would say that software and military hardware are pretty arcane, but there are lots of persons who are working in IT or the military, and these are drawn to Wikipedia, so relatively speaking these are 'easy' topics.
        But there are also fields where it proves that the popular myths dominate. More worrying is that although there is a NPoV-policy, in such fields only the popular myth is presented, with no room for presentation of real facts. - Brya (talk) 05:55, 9 January 2015 (UTC)

Freebase - large content donation from Google

I think the thread about Wikidata:Project_chat/Archive/2014/12#Freebase - large content donation from Google was prematurely archived, can we reopen it? Especially the issues with sources needs further thoughts, and I think we should consider using Freebase itself as the source for the statements. If someone can make better sourced statements (including but not limited to Knowledge Vault) there should be no problem marking the Freebase ones as deprecated, but until then they are as good as any other statements. In fact in this case we have a source, while a lot of our existing statements are without any source at all.

I say import all statements, but do not mark them as preferred unless they have additional sources. Jeblad (talk) 17:46, 4 January 2015 (UTC)

Well, Wikipedia is regarded as not being a reliable source. Freebase consists of stuff imported from Wikipedia, and it would be strange if the mere action of copying stuff from Wikipedia into Freebase suddenly made it more reliable. - Brya (talk) 18:16, 4 January 2015 (UTC)
We should not import a single statement without a reliable, external source. --Succu (talk) 18:31, 4 January 2015 (UTC)
The position that only reliable external sources is valid is actually wrong. If we limit our self to first order predicate logic it sort of makes sense, but in the more modern Markov logic it does not. Also note that we already have the rank field and that already marks a movement in the direction of Markov logic. (The difference is that you really want to infer the weights and not usually leave it to the users.) There is a few valid reasons not to import stuff from Freebase. First it is the database structure which does not align very well, then there is the user generated schemes in Freebase. Those are although only some man-hours away, not to say man-month. I think merging the projects are a very good idea. Jeblad (talk) 18:48, 4 January 2015 (UTC)
LOL: Markov logic network (Q6771324) - Really? Please keep in mind the suggested curation and reference tool which is the core component of the proposal. --Succu (talk) 19:07, 4 January 2015 (UTC)
Your idea is wrong: we should import Freebase data with deprecated level during the first import and not later because Freebase content doesn't match any good practice. WP want reliable data, so please be fair, say from the beginning that these data are not reliable in the sense of Wikipedia rules. If you consider that Freebase can be considered as deprecated, do it from the beginning. Snipre (talk) 22:02, 4 January 2015 (UTC)

My problem is that I don´t know what data freebase has. I have seen maybe ten or twenty articles, but freebase has millions of articles. And how do we recognize that a freebase data is already in wikidata? Or should we copy all data to wikidata and sort in wikidata the info? PS: In my view, the sources are not the main problem. --Molarus 23:39, 4 January 2015 (UTC)

Yes, that will be the problem. A likely scenario (already seen too many times) is that here at Wikidata some errors (imported from a Wikipedia) have laboriously been removed and that those who are going to import stuff from Freebase will find "ha, this is missing, let's import it" and put the error back in. A never-ending cycle to make sure that all errors existing anywhere will be gathered in Wikidata. - Brya (talk) 04:06, 5 January 2015 (UTC)
I have not looked at all that many Freebase pages, but in what I have looked at there was nothing even remotely worth importing. - Brya (talk) 04:23, 5 January 2015 (UTC)
Google has kind of offered upload of data from Freebase, and I'm also not sure if that will be a good thing for us. If it was up to me, Google would just keep Freebase.com online for anyone in readonly mode, and we just keep doing what we're doing. Wikidata exists for only about a year or so, we'll become bigger (and better) then Freebase anyhow. Maybe imports can be helpful in some areas, but a big brutal dump of all data? We better believe in our own strenght. Edoderoo (talk) 22:37, 7 January 2015 (UTC)
@Edoderoo:, we definitively do not want to make a huge brutal dump and upload it. We will release the data in batches, where each triple can be individually assessed. Also, even though the raw numbers might suggest otherwise, Freebase is not that huge compared to Wikidata - it is more a matter of the counting methodology. But also, we do not want to force the data here: we won't upload a single triple ourselves. For me, the most important part is for Wikidata to remain a healthy and active community, and I do realize that having too much to swallow at once could be a problem, which is why we are very careful in what we are doing, and keeping the whole project in a conversation and collaborative. --Denny (talk) 16:30, 8 January 2015 (UTC)
Maybe the best way to go is to pre-import all the data in a wikimedia-environment, so we can easily move the interesting parts into Wikidata, either by bot, or with a tool. In such a way there is always a Wikimedia user accountable, the same way as we now can mass-upload properties with AutoList or QuickStatements. Edoderoo (talk) 20:30, 8 January 2015 (UTC)
I agree. That's exactly the idea behind the primary sources tool. --Denny (talk) 22:36, 8 January 2015 (UTC)

Wrong property (bot request)

276 instances of commune of France (Q484170) are using taxon range map image (P181) instead of locator map image (P242). Can a bot correct this? Sjoerd de Bruin (talk) 12:09, 6 January 2015 (UTC)

you can do it yourself if you have time. Goto http://tools.wmflabs.org/wikidata-todo/tabernacle.html, claim[181] and claim[31:484170)] , in propertylist put: P181. Then show and download the results. Edit the this textfile with the correct new property P242 as second column, add " " to the third column and paste inside http://tools.wmflabs.org/wikidata-todo/quick_statements.php . Then remove the old claims with autolist -P181. 139.63.61.23 12:17, 6 January 2015 (UTC)
Okay, running. Thanks! Sjoerd de Bruin (talk) 19:46, 9 January 2015 (UTC)

How are language fallbacks working for you?

Hey folks :)

Language fallbacks are live for a bit now and I'd like to get some feedback. Are they working for you? Anything where they're horribly wrong? Anything where they could be better? I'd especially like feedback from people who use Wikidata in a language other than English.

Everything I am already aware of is linked as "blocked by" tickets of phabricator:T76216.

Cheers --Lydia Pintscher (WMDE) (talk) 11:48, 7 January 2015 (UTC)

It's fine that Qids in statements partly disappeared. Matěj Suchánek (talk) 19:33, 7 January 2015 (UTC)
@Lydia Pintscher (WMDE): language fallbacks does not works ideally. It should fallback from any languages to any languages, not just follow the fallback chain. For example, in historic site (Q1081138), Q15077340 should show a label, even if it is in German. A label in other languages is better than Qid, as Qid tell us nothing; Whereas names in other languages are often readable, especially personal names.--GZWDer (talk) 05:44, 8 January 2015 (UTC)
Maybe there should be provided some flag in labels to mark one of them as final language fallback to avoid English as default. Or there could be provided an order of labels. The order might be generated automatically, that is any newly created label is added to the end of the list. If language fallbacks mechanism do not find its predefined parent language then scan the list to look for any language that uses similar script. That is missing English label could be replaced by any label written in latin script i.e. French or Italian, missing Russian label might use Bulgarian etc. The order of labels might help in case an item describes a person or geographical location, which might be created on wiki, where the item comes from. In that case the final language fallback flag assigned automatically on newly created item seems to be an easy solution. Paweł Ziemian (talk) 16:15, 8 January 2015 (UTC)
The list of assistant languages from user's Special:Preferences#mw-prefsection-editing could be used as well. I agree that fallback from any language to any language should be possible. Petr Matas 21:09, 8 January 2015 (UTC)
Also label fallback in Special:RecentChanges, Special:Contributions etc. is better. It will always show a label, even in foreign languages.--GZWDer (talk) 02:35, 9 January 2015 (UTC)
@Lydia Pintscher (WMDE): It doesn't seem to work when adding an item. For example, I want to add the statement creator (P170): Juan Ávalos García-Taborda (Q946078). This item has one label in English but not in French (the language in which in use Wikidata). I add the property and then type "Juan de Ávalos", but the item is not found, I have to manually add "Q946078". — Ayack (talk) 08:37, 8 January 2015 (UTC)
Yeah that is one of the open tickets still. --Lydia Pintscher (WMDE) (talk) 08:41, 8 January 2015 (UTC)
I would prefer a language fallback to taxon name (P225) in stead of a foreign language if this property is available. Lymantria (talk) 15:53, 9 January 2015 (UTC)
Yes, this probably is close to what I was asking for. - Brya (talk) 05:18, 10 January 2015 (UTC)

For antivandals: Mark as patrolled directly from Recent changes

If you use Recent unpatrolled changes, you may find handy being able to mark a change as patrolled directly from the list of changes. I have written a script, which adds the [mark as patrolled] links there. To use it, add importScript('User:Petr Matas/Mark as patrolled.js'); into your common.js. Please test and comment. Petr Matas 02:17, 10 January 2015 (UTC)

Merging help

I think Maza Township (Q10792028) and Maza (Q1914566) are the same place, but I'm not certain. Could someone help me verify this? Thanks! Popcorndude (talk) 23:52, 10 January 2015 (UTC)

Actually, they are two places. The city of Maza was dis-incorporated in 2002 and governance reverted to Maza Township (which is a redlink in EN wiki). I'll disambiguate them. - PKM (talk) 01:32, 11 January 2015 (UTC)
Thanks. Popcorndude (talk) 03:12, 11 January 2015 (UTC)

New Proposal for Bonnie and Clyde

I made a new proposal for solving Bonnie and Clyde. Not sure this already came up, so feedback is more than welcome. --S.K. (talk) 13:52, 11 January 2015 (UTC)

Wikidata weekly summary #140

Here's your quick overview of what has been happening around Wikidata over the last week.

Administered by?

Do we have a property to indicate that a museum or collection is "administered by" a foundation, diocese, public agency or other organization? - PKM (talk) 20:52, 11 January 2015 (UTC)

Maybe maintained by (P126)? Sjoerd de Bruin (talk) 21:03, 11 January 2015 (UTC)
Couldn't you use owned by (P127), occupant (P466)?--Oursana (talk) 02:14, 12 January 2015 (UTC)
There is also operator (P137). —Wylve (talk) 18:28, 12 January 2015 (UTC)
operator (P137) looks like the closest match. Thanks, all. - PKM (talk) 20:33, 12 January 2015 (UTC)

Order By

Hello,

I already asked my question here but wanted to be sure. Is there any way to order a query in Autolist by page title ?

Thanks :)

Scoopfinder (talk) 19:23, 13 January 2015 (UTC)

I would say no, but it's probably the best to ask the developer of the tool @Magnus Manske: --Pasleim (talk) 19:39, 13 January 2015 (UTC)
Page titles are loaded one-by-one afterwards. So, no, at least not easily. --Magnus Manske (talk) 22:17, 13 January 2015 (UTC)

Does Wikidata information appear on the Google Knowledge Graph?

I love adding information onto the Google Knowledge Graph and have usually used Freebase to do this. I recently heard that Freebase are quitting or something and have directed users to use Wikidata. Will information still show up on Google search results with data inputted on this site? (I am new to this software).

yes, see [2] --Pasleim (talk) 19:46, 13 January 2015 (UTC)

'1' in front of summaries

Why do most edit summaries (even old ones) suddenly have a ‘1’ before them? I don’t see anything related to that in the Status updates… —DSGalaktos (talk) 22:45, 13 January 2015 (UTC)

It's a bug. We're working on it right now. Sorry for the short uglyness. --Lydia Pintscher (WMDE) (talk) 22:47, 13 January 2015 (UTC)
Okay, no problem. Thanks for the info! —DSGalaktos (talk) 22:52, 13 January 2015 (UTC)
It is fixed now. --Lydia Pintscher (WMDE) (talk) 00:29, 14 January 2015 (UTC)
Is someone fiddling around with the page styling? Right now, for me, when viewing a Q item the wikilinks list is on the right-side of the page, whereas it was at the bottom. Danrok (talk) 02:55, 14 January 2015 (UTC)
It's a new design. --Stryn (talk) 06:06, 14 January 2015 (UTC)

Preventing bots from creating WD items via blacklists

Hello, I noticed time and again that some bots create WD items for categories from dewikisource which serve no purpose outside of the project itself. In my opinion, we don't need specific items for these categories. For example: s:de:Kategorie:RE:Autor:Maurice Besnier lists articles in the Pauly-Wissowa encyclopedia written by Maurice Besnier. This category is only needed on dewikisource.

My question is: How can we prevent bots from creating WD items for this type of category? I would suggest that administrators should be given the right to create blacklists for items on their respective projects and/or flag items there for prevention.

Please tell me your opinions on this, or thoughts on how to implement it. Thanks! Jonathan Groß (talk) 14:17, 12 January 2015 (UTC)

I'm woefully unfamiliar with Wikisource internal workings; do the projects work on translating source texts? If so, perhaps the categories are useful cross-project. Blood Red Sandman (talk) 01:10, 13 January 2015 (UTC)
Yes, many times the links are useful. But there in most categories WD items just don't make any sense. Like every subcategory from s:de:Kategorie:Projekt for example. Jonathan Groß (talk) 07:34, 13 January 2015 (UTC)
I agree with Jonathan Groß,
on fr.ws, we also have tens of thousands of individual dictionary or encyclopedia articles, that are grouped in categories... those individual article should not be imported in wikidata... no wikilinks, this is just a part of navigation through the dictionary/encyclopedia... :/ --Hsarrazin (talk) 08:04, 14 January 2015 (UTC)

Patents

Just wondering...
What are people's thoughts about patents being within the scope of Wikidata? Should individual patents be able to be have Q items? One of the particular complexities in Patents is that there is no such thing as a "global" patent - there are simply national (and sometimes regional e.g. European Union) patents that can be simultaneously applied for. It is possible for a patent to be applied for in 20 countries, approved in 10 and rejected in 10, and not applied for at all in all the other countries. As far as I now there's no global register of this and it's very difficult to cross-reference the different databases to find the 'same' patent in different countries. I see that Wikidata could be a useful authority-control mechanism for patents in different jurisdictions, but I'd like to know if the community thinks that Patents are indeed within the scope of the project in the first place. Wittylama (talk) 13:22, 13 January 2015 (UTC)

Patents as items in the sense of individual objects, no, but for patents used as references we have to decide about a formal way to describe them in order to be able to use them as sources. Snipre (talk) 15:19, 13 January 2015 (UTC)
@Wittylama: If you want more details, you can start a section in Help:Sources. Snipre (talk) 12:53, 14 January 2015 (UTC)

Can’t edit empty item

I’m not getting any “Edit” links on Microsoft Messaging Passing Interface (Q6840167). I suspect that this is a bug, and that it’s due to the item having no statements; however, since I’m unable to find another empty item, I can’t confirm that. —DSGalaktos (talk) 13:29, 14 January 2015 (UTC)

Have you tried to purge (?action=purge) the page? Sjoerd de Bruin (talk) 13:30, 14 January 2015 (UTC)
Hey, that worked! Thanks! —DSGalaktos (talk) 13:34, 14 January 2015 (UTC)

How to hide reverted edits from Recent changes?

I am trying to monitor some Recent unpatrolled changes. I mark some of them as patrolled (so they disappear) and I revert some of them. How do I hide the reverted changes? I don't think it is a good idea to mark them as patrolled as well, because I guess that that contributes to the originator's autoconfirmed status, doesn't it? Petr Matas 00:32, 9 January 2015 (UTC)

@Jared Preston: You seem to be an active vandalism fighter. What is your method? How much time do you spend on each edit and on reverting a harmful one? Is there any tool which shows whose edits are reverted often? Petr Matas 12:10, 9 January 2015 (UTC)

Another question: What prevents vandals from acquiring the autoconfirmed status? (Autoconfirmed users' edits are marked as patrolled automatically, right?) Petr Matas 12:29, 9 January 2015 (UTC)

Well, to answer a few questions all I can really say is that it takes as much time as it takes. Some changes are hard to oversee, some are suspicious, others (in a foreign language I don't command, for example) are less obvious, but all changes I revert are first opened in a new tab so I can see what I'm doing. What I have noticed though, however, is a "reverted" edit is then removed from the list of recent changes, whereas a simple "undo" to a change leaves the edit still unpatrolled upon refreshing of Special:RecentChanges. As an admin, I can see if that edit is the last revision to a page because of the [rollback] link next to the edit summary. Since I've noticed that you too are into fighting vandalism, I'd suggest WD:RB. I've had years of experience on Wikipedia as a rollbacker before coming here. Jared Preston (talk) 12:48, 9 January 2015 (UTC)
@Petr Matas:, you should just mark them as patrolled even if they were vandalism. It will not have any effect on autopatrol status. Jon Harald Søby (talk) 18:54, 15 January 2015 (UTC)
Thanks! Is it what Rollback, which hides them as well, does? Petr Matas 07:31, 16 January 2015 (UTC)

What to do when one Wikipedia has two pages on the same topic?

I was working on something today, and discovered that the Tatar Wikipedia has two separate pages for "Category:Family": tt:Төркем:Гаилә (currently at Category:Kinship and descent (Q6520267)) and tt:Төркем:Ğäilä (currently at Category:Family (Q7216461)). The name's the same really; it's just one's in Cyrillic characters and the other's in Latin ones. This appears to be intentional, as the two pages are linked via hatnote. This was briefly discussed here in July; commenting on his talk page, User:Dobroknig, who speaks Tatar, wrote "According to internal rules ttwiki , since 2013 use of Cyrillics and a Latin alphabets is permited, but duplicates of articles are forbidden, but all articles written till 2013 remain, even if they are written in a Latin and cyrillics alphabets." So it looks like tt.wp is at least not creating duplicates anymore, but there must still be quite a few that exist and won't be going away anytime soon.

Other than that discussion in July, I can't find anything else on what to do in a case like this. Do we put one article on the item with all the links, and create a new item with no other links for the other? Do we just not link the other article? In both of those cases, how do we decide which article gets the short end of the stick? Or do we ask the development team to allow two articles per item for tt.wp and any other wikis with practices like this? To be honest, the last idea there makes the most sense to me, but I doubt the devs would like the idea too much. So, yeah, ideas welcome. — PinkAmpers&(Je vous invite à me parler) 00:25, 14 January 2015 (UTC)

Making this sort of exception for some languages does not sound like a good idea: the reasons why we do no want 2 sitelinks in en are also valid in tt, even though there are complicating factors in there.
I still think the best solution would be is the one I proposed in Wikidata:Project_chat/Archive/2014/07#Proposing item merge and merging: add a statement leave one item mostly blank with just the sitelink and a statement saying that it is a Wikimedia duplicated page (Q17362920) of another item. -Zolo (talk) 08:42, 14 January 2015 (UTC)
@Zolo: That solution seems good enough for one-off cases, but maybe we can come up with something more tailor-fit to cases like these, where the duplication is intentional on the part of the wiki in question? It sounds like there's a whole lot of tt.wp articles we need to deal with. And I also remembered that the Ladino Wikipedia has all its articles in both Latin and Hebrew characters. What if we made a property like "alternate-alphabet duplicate" or something for these? And then the status of primary link would go to whichever article appears more thorough (or, if that's unclear, whichever is older). — PinkAmpers&(Je vous invite à me parler) 22:27, 14 January 2015 (UTC)
Eventually all of those wikis should be moving to the LanguageConverter, which will make it a moot issue. Kaldari (talk) 22:56, 14 January 2015 (UTC)
@Kaldari: That's good to hear. :) Still, I'd like to be able to do something in the meantime. By any chance, do you have a list of wikis that use multiple alphabets but don't have LanguageConverter? I'd like to know how large of an issue this might be. — PinkAmpers&(Je vous invite à me parler) 03:46, 15 January 2015 (UTC)
@PinkAmpersand: "alternate-alphabet duplicate" would be a new item, not a new property, but yes, if the case is really massive and unlikely to be handled quickly, a more specific item may be a good idea. --Zolo (talk) 23:19, 14 January 2015 (UTC)
@Zolo: Why should it be an item, not a property? It'd be a property to link the two items, I mean. Like there'd be the item with the Cyrillic-alphabet article and the one with the Latin-alphabet one, and both would link to the other with the "alternative-alphabet duplicate" property. — PinkAmpers&(Je vous invite à me parler) 03:43, 15 January 2015 (UTC)
Can said to be the same as (P460) be of any help? --Kolja21 (talk) 05:43, 15 January 2015 (UTC)
Yes the solution currently used is "P460: the other item" + "instance of: duplicate page". We could conflate them in a single "is duplicate of" property but having a "instance of" property in the item makes it more consistent with the rest of Wikidata, so the 2 statemetns (or perhaps P31 statement + P460 qualifier) sounds like a better solution to me. --Zolo (talk) 08:22, 15 January 2015 (UTC)

Why are valid items deleted?

Major Cineplex Ratchayothin (Q6738040) was deleted with the comment "Empty / unlinked item: the page has been deleted from wikipedia". Only thanks to the Google cache I can see what that item was about - a big cineplex in Bangkok. It is correct that the Wikipedia page was deleted, however that was in May 2014 as being not notable enough - but that was on the English Wikipedia, for the Thai WP it may be more notable. And for a travel guide about Bangkok even more. Besides, the item was NOT empty, I did spend time to fill statements into it, to proof the notability with them. I only noticed that my work was in vain because I added the item to my watchlist, as I had cases before there I was sure I had added an item before but it disappeared, and a deletion log which only shows item numbers but no labels is zero help. Yes, we need to do housekeeping and delete junk items, but items with statements IMHO should not be deleted in such a quick way - no discussion, no notification to the user(s) who worked on them. This is very discouraging, I am trying to add data here to make items useful, but if that work is at risk of being deleted without even having the chance to say "wait, stop", and if I hadn't watched the page even without the chance of asking to undelete it, why should I spend my time on it at first? Ahoerstemeier (talk) 10:29, 16 January 2015 (UTC)

Restored, please see Wikidata:Administrators'_noticeboard#Wrong_deletion ·addshore· talk to me! 12:12, 16 January 2015 (UTC)

Municipalities of Estonia, splitting a category object

I'm thinking about splitting Q9022297 in two, one for categories covering all municipalities in Estonia, one for categories covering only rural municipalities (vallad). I'd appreciate some help with this. First, is the split likely to be controversial?

The reason for the proposed split is that some Wikipedia versions have common categories the two types of municipalities, some have completely separate category trees for cities and rural municipalities. ETWP has et:Kategooria:Eesti omavalitsused for all municipalities and also some other self-governing entities (no Wikidata link) and et:Kategooria:Eesti vallad (connected to Q9022297) for the rural municipalities. However, most things connected to Q9022297 seem to deal with all municipalities.

Second, how is the split implemented in practice? Should I start a new object from et:Kategooria:Eesti omavalitsused and move most things from Q9022297 there, restricting Q9022297 to rural municipalities; or should I replace et:Kategooria:Eesti vallad with et:Kategooria:Eesti omavalitsused at Q9022297 and then start a new object from et:Kategooria:Eesti vallad for the rural municipalities? How much am I expected to add apart from the Wikipedia sitelinks?

Best regards, Essin (talk) 19:30, 9 January 2015 (UTC)

"is the split likely to be controversial?" I think this question is best asked on Wikipedia. Danrok (talk) 03:06, 13 January 2015 (UTC)
"is the split likely to be controversial?" If the split means our WDitems more accurately reflect the WPcategories then it is not likely to be controversial. If the split involves a change to the WPcategories then ask on the WP affected. Filceolaire (talk) 19:58, 13 January 2015 (UTC)
I was hoping to avoid 20 or so parallel discussions. I don't intend to change any WP categorization other than sv:Kategori:Estlands kommuner, at my "home" wiki. Currently, it's not very clear whether that category is supposed to cover all or only rural municipalities. Mostly, I think a split would mean more accurate reflection, although interwiki between versions that only have separated categories and those that only have common categories would be lost. In some cases, the separation (or lack of it) is not very explicit, but then maybe a short pointer to Talk:Q9022297 would be sufficient to get clarification. //Essin (talk) 22:05, 16 January 2015 (UTC)

Notability discussions

FYI Talk:Q15304738 and Talk:Q18780030 --Pasleim (talk) 01:15, 16 January 2015 (UTC)

Hoi, these discussion assume that the bigger projects have a say about notability in Wikidata. They do not. ANY project that has an article on a subject raises notability as to be sufficient. Thanks, GerardM (talk) 07:26, 16 January 2015 (UTC)
The bigger projects do have a say about notability in Wikidata (namely, they can establish it). All projects do. But this argument is void here anyway, because both items have no sitelinks from any of the smaller Wikimedia projects anyway. --Denny (talk) 01:35, 17 January 2015 (UTC)

P47

At shares border with (P47) mentioned "countries or administrative subdivisions, of equal level,..." so is P47 at European Union (Q458) incorrect?

If we want to say Luhansk Oblast (Q171965) or Strait of Kerch (Q187890) or Sea of Japan (Q27092) or European Union (Q458) has border with Russia (Q159) how can we do?Yamaha5 (talk) 21:23, 16 January 2015 (UTC)

Maybe this rule should be canceled and preffered ranks should take place, i.e. for Czech Republic (Q213) only Slovakia (Q214), Poland (Q36), Germany (Q183) and Austria (Q40) would be preffered. Matěj Suchánek (talk) 11:14, 17 January 2015 (UTC)
But does organizations have borders? -- Innocent bystander (talk) 13:36, 17 January 2015 (UTC)

How to upload multi-column statistics tables to WikiData?

How would I go about uploading a multi-column table of statistics to WikiData for use on a particular Wikipedia article? My main confusion is that wikidata seems to be structured so that you can only upload a single list of properties, like in the following example:

[List Name (Ex. "List of Crime Rates in 1973")]=>
[Property (Ex. "Murder")]=>[Value]
[Property (Ex. "Theft")]=>[Value]
[Property (Ex. "Assault")]=>[Value]

Which outputs the following way in a query:

List of Crime Rates in 1973:
Murder 5.2
Theft 7.6
Assault 11.5

Where as a multi-column statistics table is in fact a "list of lists containing property-value pairs" (also known as a "multi-dimensional array"), like the following example:

[List Name (Ex. "US Crime Rates by Year")]=>
[Row (Sub-list) Name (Ex. "1973")]=>
[Property (Ex. "Murder")]=>[Value]
[Property (Ex. "Theft")]=>[Value]
[Property (Ex. "Assault")]=>[Value]
[Row (Sub-list) Name (Ex. "1974")]=>
[Property (Ex. "Murder")]=>[Value]
[Property (Ex. "Theft")]=>[Value]
[Property (Ex. "Assault")]=>[Value]
[Row (Sub-list) Name (Ex. "1975")]=>
[Property (Ex. "Murder")]=>[Value]
[Property (Ex. "Theft")]=>[Value]
[Property (Ex. "Assault")]=>[Value]

Which outputs the following way in a query:

US Crime Rates by Year:
Year Murder Theft Assault
1973 5.2 7.6 11.5
1974 5.3 7.8 11.4
1975 5.5 7.7 11.2

Can I upload a statistics table so that I can query it and dump the result into a table that looks like the original table? Or is wikidata not capable of this level of functionality? --Chrisbat92024 (talk) 05:44, 17 January 2015 (UTC)

The inverse
property Murder: Value
qualifier point in time (P585): 1973
property Theft: Value
qualifier point in time (P585): 1973
Snipre (talk) 10:55, 17 January 2015 (UTC)

Wikidata weekly summary #141

Here's your quick overview of what has been happening around Wikidata over the last week.

Wikinews category with mutliple coverage

On en.wn, there's a Category:Donetsk that covers both the city of Donetsk and the oblast of Donetsk. On Wikidata, there are items Donetsk (Q43070) and Donetsk Oblast (Q2012050). Since the en.wn page is linked from the first of these items, it cannot be linked from the second. For Wikidata's ontology, that seems to make sense; but for Wikinews it means that readers who look at an oblast category in some language are deprived of an interwiki to the corresponding category on en.wn, unless a local interwiki directive is placed on each oblast category page.

Am I understanding this situation correctly? It seems as if this is a good example of incompatibility between Wikidata's desire to create an ontology and its (lower-priority) desire to provide useful interwikis. --Pi zero (talk) 01:06, 18 January 2015 (UTC)

Wikipedia article about event, not the person

At the moment we have (Q18341392) "Murder of Kitty Genovese" pointing to "enwiki-Murder of Kitty Genovese". There is also a (Q238128) "Kitty Genovese" with no enwiki link. Nothing wrong so far! However, I am tempted to link the (Q238128) "Kitty Genovese" to the "enwiki-Kitty Genovese" redirect page on the basis that it is the redirect page that carries the biographical information in this instance.

There are many articles where this happens. Bonnie & Clyde and the Kray twins being two notable examples. My personal project involves comparing Wikidata claims against Wikipedia categories and templates, so for me pointing to the redirect is perfect. Are there objections linking to redirect pages? Periglio (talk) 06:51, 17 January 2015 (UTC)

I have answered my own question. Wikidata overrides the redirect page and adds the resultant page - or it would have if the resulting page was not already linked. So at the moment, I just have to ignore Bonne and Clyde as individuals. Periglio (talk) 17:25, 17 January 2015 (UTC)
I think this approach can be used. Now these properties can be recognized by bots like your. In future infobox can use it to query data. — Ivan A. Krestinin (talk) 18:30, 17 January 2015 (UTC)
Is that an official parameter yet? I like it but i can not find any other mention of it? Periglio (talk) 07:59, 18 January 2015 (UTC)
No, it is just my idea. Now is good time to introduce ideas like this. — Ivan A. Krestinin (talk) 08:39, 18 January 2015 (UTC)

Extra feature request in WidaR

WiDaR is one of the most used tools on Wikidata, I believe, checking the recent changes it is one of the important parts of getting properties filled. I experienced that WiDaR can often suddenly stop for good or for less good reasons, therefor I check regularly if my browser tabs with WiDaR-activity are still up and running. Due to the 10-sec-throttle this is taking quite some time. It would therefor be helpful if WiDaR would show a time-stamp (like 10:08) of the last change made. I couldn't find a page for feature requests, or should I just drop this on the talk page of Magnus? Regards, Edoderoo (talk) 09:10, 18 January 2015 (UTC)

Feature request can go here. Sjoerd de Bruin (talk) 09:14, 18 January 2015 (UTC)

How to deal with interrupted statements?

For instance someone was a memeber of a music group, he left but after several years he came back. Matěj Suchánek (talk) 16:28, 18 January 2015 (UTC)

Just add the second time as a new statement. Sjoerd de Bruin (talk) 16:33, 18 January 2015 (UTC)

simple German Wikipedia

Hi,

I’m a teacher and I want to engage students in creating and editing Wikipedia articles as well as teach them something subject related.

Wikipedia is a great tool and a vast pool of knowledge, but especially in the German language the information is lost for my students within the complicated and technical jargon. Students use Wikipedia but they usually lack the knowledge to make proper use of the information provided in the regular articles. I think this is one of the reasons why Wikipedia has such a bad reputation among German teachers (and of cause the mindless copy and past they usually do).

The English articles are usually much easier to understand compared to the Germans but on top of that there is simple English.

What I would like to have fore my students is a simple German Wikipedia.

Is there something like that in the pipeline?

I would like to handout assignments for my students to create or to simplify an existing German article that is related to my subject. Let them present the information in front of the class, edit the article and finally release it on Wikipedia so that everybody benefits from their work.

This would create content for all students, would create new Wikipedia savvy citizens and my students would learn a lot in the process.

So if anybody has any idea how to start the simple German Wikipedia or if this project is already on the way and one could contribute please let me know.

Thanks,

Naso  – The preceding unsigned comment was added by Nasoboem (talk • contribs).

That question would be better addressed on Meta-Wiki (meta.wikimedia.org), not here. --Rschen7754 02:30, 27 December 2014 (UTC)
Hallo Naso, spannende Idee! Eine "Simple German Wikipedia" wäre eine neue Sprachversion. Die kannst du unter Incubator:Main_Page beantragen. Anschließend läuft eine Testphase, und wenn sich genügend Autoren engagieren, werden die Artikel vom Incubator in die neue, dann offizielle Sprachversion ("simple-de") verschoben. --Kolja21 (talk) 03:38, 27 December 2014 (UTC)
@Nasoboem, Kolja21: Post it here if the proposal is ready. I like the idea too. --Tobias1984 (talk) 23:25, 27 December 2014 (UTC)
As I know simple English Wikipedia is based on Basic English list consists of 200 words. On what vocabulary of German language do you plan to base "simple-de"? --Infovarius (talk) 23:26, 29 December 2014 (UTC)
@Infovarius: You wrote 200 words, but your link 2000 words or more. --Diwas (talk) 00:54, 30 December 2014 (UTC)
Such a project has been discussed a number of times over the years, both on the German Wikipedia and the Meta wiki. But, as Infovarius said above, while Simple/Basic English is an already established simplified form of English with a controlled vocabulary, no similar thing exists for the German language. A search on the German Wikipedia however found de:Portal:Medizin/Kinderleicht, a project to create articles about medicine for children, and Klexikon, a wiki for children. --95.90.51.227 22:14, 31 December 2014 (UTC)

Hallo @Nasoboem: ich glaube das Klexikon ist genau das, was Du suchst und es ist gerade letzten Monat richtig gestartet! Schau mal bei http://wiki.klexikon.de/ und dem Blog dazu http://klexikon.de/. Es ist gibt sogar heute ein Treffen, immer wöchentlich montags um 20 Uhr - per google Hangout und gleichzeitig in der Wikimedia-Geschäftstelle in Berlin-Kreuzberg. Da passt Deine Idee doch wie die Faust aufs Auge ;-) Mehr Infos auch bei de:Benutzer Diskussion:Wiktoriapark ("Mitmachen ist strengstens erwünscht") oder noch einfacher http://mitmachen.klexikon.de/ Besten Gruss --Atlasowa (talk) 14:47, 5 January 2015 (UTC)

Es gibt auch das Hurraki, ein Wörterbuch für Leichte Sprache. Dieses ist gedacht für Kinder, Menschen, die Deutsch nicht als Muttersprache haben und für lernbehinderte oder kognitv eingeschränkte Menschen, die mit langen Sätzen nicht klarkommen. Es geht dabei auch um Barrierefreiheit für diese Menschen. Es gelten dort diverse Regeln die ungefähr denen von Simple English entsprechen. Es ist aber kein Wikimediaprojekt. Zum Thema "Leichte Sprache" gibt es auch verschiedene staatlich geförderte Ansätze, ein Wiki das dem Simple-English-Wiki entspricht gibt es aber meines Wissens nicht. Siehe auch https://de.wikipedia.org/wiki/Leichte_Sprache --Giftzwerg 88 (talk) 16:53, 12 January 2015 (UTC)

In my opinion, Simple English should be moved to its own wikispace - and not be considered a language of Wikipedia, but instead be an instance of "Childrens/Juniorwiki", along with "simple German" and "simple" every other language. There is always a market for encyclopedias that are targeted for the juvenile audience[3] - and the "simple" version should not be considered a separate language, which of course it is not. Besides it would save me the trouble of having to delete "simplewiki" from my list of languages every time. I have to do that thousands of times. Wixty (talk) 22:18, 18 January 2015 (UTC)

User:Hazard-Bot/Archiving

Hello everyone, I've just gone ahead and updated the documentation on User:Hazard-Bot/Archiving, which shows how you can use Hazard-Bot (my bot) to archive discussion pages (such as the project chat and your userpage). I probably haven't proofread as much as I should have, but if you say anything that needs fixing or could be improved, please go ahead do so. If you need any help or clarification, just let me know.  Hazard SJ  04:34, 19 January 2015 (UTC)

Just two comments ref Wikimedia Commons and bot-scripted additions started in language-oriented Wikimedia-WP projects

Hi Wikimedians, two comments as precise as able missing as good EN-language knowldege related to Wikimedia Commons and bot-scripted additions started in other Wikimedia projects: Maybe an error of Win7pro/IE, but imho there may be missing a

  1. separated 'section' in the article related 'wikidata' to Commons and
  2. at least in my 'user contributions' may missing a hint that a wikidata edit was started in an language-oriented Wikimedia WP project, as personally do prefer to add Wikidata related data, otherwise it's just when i have bit opportunity to do, as usually focussed on other Wikimedia activities.

To explanin 2) So that the last edits i remarked were started in EN-WP or the Malayalam Wikipedia short time before, please seee 06:02, 19 January 2015, and some minutes before. to precises, i'm not Enlish language native speaker, missing a more 'exlicite declaration' that such edits are automatically bot-scripted by bot-xyz. No offense please, and just to hint hoping to be helpfull for the Wikidata project :-) Wishing having much fun within Wikimedia to you, kindly regards Roland zh (talk) 06:23, 19 January 2015 (UTC)

Anonymous user adding URL

Is it intentional that I can't add a reference URL without logging in, e.g. here for population? I assume passing CAPTCHA is needed, but all I get when trying to save is "An error occurred while saving. Your changes could not be completed." (no details available). 2001:7D0:88C2:F401:FDB3:C36:41F2:26A2 20:59, 17 January 2015 (UTC)

Should anonymous users be able to add reference URLs as property values? A bug? 193.40.10.180 09:29, 19 January 2015 (UTC)

Yeah there were changes to the captcha system recently that Wikidata doesn't deal with well. It results in anonymous editors not being able to add URLs. We're tracking the issue at phabricator:T86453. --Lydia Pintscher (WMDE) (talk) 10:06, 19 January 2015 (UTC)

User preferences

I noticed that there is a "Beta" tab, but all it does is take you to user preferences. Is there any chance of adding the original user interface to user preferences? https://www.wikidata.org/wiki/Special:Preferences#mw-prefsection-rendering It was sooooo much easier to use, and has the functionality that I need of being able to make a list of articles in every language without having to go through the BS of getting them from the XML and doing a bunch of sorts, edits, and more sorts to strip them out (copy and paste to notepad, copy and paste to spreadsheet, sort, copy and paste back to notepad, do four edits to remove the extra junk, copy and paste back to spreadsheet, sort, delete extra columns, to end up with everything - the same thing I would have had with one cut and paste with the old interface). Wixty (talk) 20:58, 18 January 2015 (UTC)

What is the relationship between the original user interface and beta features? It seems like you don't understand what the beta feature is on Wikimedia projects. Sjoerd de Bruin (talk) 07:54, 19 January 2015 (UTC)
Feel free to explain it, but in software development, the beta phase is the last stage towards product release, a version that might still have some problems, but it is being released to a limited number of users so they can run through it and help wring out any issues. It is an opportunity to try out something that is coming down the pike before it gets here. Cautious users would not touch a beta release with a ten foot pole though, because it could have dire and unforeseen consequences. So "beta = the next version". What I want in user preferences is not "the next version", but "a working version" - the original version. So the relationship between beta and original version is beta = next step, original = previous step. Both being "a different version". Now if Beta is used differently here, how is it being used? Wixty (talk) 18:18, 19 January 2015 (UTC)

New articles on wikipedias

Hello. Is there a bot creating new items for new articles on wikipedias or do they have to be created manually?--Cattus (talk) 17:44, 12 January 2015 (UTC)

Hi, there is no bot creating new items for all wikipedias and wikimedia's projects, but there are individual bots doing that tasks for individual projects (usually using own rules for new articles, usually there is a delay between new article and new item to prevent crating items about non-notable articles, that are deleted soon). For some small projects it may happen that none of bots is doing that tasks. --Jklamo (talk) 22:44, 12 January 2015 (UTC)

This [4] is useful for creating new items. It will make a list of all Wikipedia items which don't have a Wikidata item, and create those new items. Note that it follows the category tree, so don't choose a category with 5 million items in its sub-categories. Danrok (talk) 02:39, 13 January 2015 (UTC)

Thanks to both of you for your replies. That does seem like a very useful tool. I think I'll give it a try.--Cattus (talk) 22:03, 13 January 2015 (UTC)
I was about to ask the same question! May I also ask, does the tool just create the item, or is it sophisticated enough to copy across the basic claims? eg Birth dates? Periglio (talk) 22:23, 13 January 2015 (UTC)
It just does the bare minimum. But, once it is complete you can always copy/paste the item IDs to [5], and then add something like instance of (P31)=human (Q5) as a batch process. Which is a good start. There may be bots which will eventually fill in the blanks for certain types of item. Danrok (talk) 02:43, 14 January 2015 (UTC)
I would not like to see anyone use any sort of tool or bot like that which just creates new items for an article where there is already an item that article needs to be added to. It is not easy to merge them (the only way I have been able to do that is to delete the dummy item). I find it far more useful to leave no wikidata item instead of creating an erroneous item. There is of course no problem if there is no other wiki that has an article on that subject, and no item that it can be added to, but for smaller wikis especially, it is unlikely that there is not already an item that the article can be added to, instead of creating a duplicate item. Wixty (talk) 23:09, 19 January 2015 (UTC)

table of properties and P1629

Now that Wikidata item of this property (P1629) has been created and is a used by near 200 properties, i suggest to add a Wikidata item of this property (P1629) column at the tables of properties, in the sub-pages of Wikidata:List of properties. The table use Template:List of properties/Header, Template:List of properties/Header/text, Template:List of properties/Row, Module:LoP row. Visite fortuitement prolongée (talk) 22:08, 15 January 2015 (UTC)

I've tried to add it in a new column but that looked a bit awkward, so I added it at the bedinning of the "description" column, when to description is fetched from the property itself. Feel free to change/revert it.
It had me notice that most of the properties documented in Wikidata:List of properties have an hardcoded description. I think most of them should be removed, as it prevents automated updates of translations/descriptions. --Zolo (talk) 10:20, 19 January 2015 (UTC)
  • Thank you.
  • I am fine with insterting this data in the "description" column instead of creating a new column.
  • Yes, I noticed that too (altought i did not noticed "most of" have), and I support the removal.
Visite fortuitement prolongée (talk) 21:15, 19 January 2015 (UTC)

Add paired claim automatically

Some properties like these should be added paired claim by wikidata. now we should get query every month and add the paired by this tool.

for example when user adds father (P22) > item A to item B wikidata extension (or js tool) should add child (P40) > item B to item A automatically Yamaha5 (talk) 21:04, 16 January 2015 (UTC)

@Yamaha5: Per devs this will not be done by Wikibase, you will only get the possibility to add the inverted claim quickly to the second item. Maybe Lydia Pintscher (WMDE) can say I am correct. Matěj Suchánek (talk) 09:02, 17 January 2015 (UTC)
Right. We'll be working with the students team that is working on constraints to nudge editors to make the second edit in such cases though. --Lydia Pintscher (WMDE) (talk) 19:24, 17 January 2015 (UTC)
Lydia Pintscher (WMDE) wouldn't it be better if All statements which link to a page could be displayed on that page? that way we could start to get rid of these inverse properties since the relationship would be clear from such a list. Filceolaire (talk) 18:55, 19 January 2015 (UTC)
Yes and no. I consider redundancy a major piece in keeping our data in shape because it helps find inconsistencies that would otherwise go unnoticed. Fine line to walk and we need to see how what the students come up with will help and where we need to do more or less. --Lydia Pintscher (WMDE) (talk) 14:56, 21 January 2015 (UTC)

Which calendar was used?

Tracked in Phabricator
Task T87312

Is there a way of seeing if a date has been entered as a Gregorian or a Julian date without going into edit and clicking advanced? Periglio (talk) 21:54, 19 January 2015 (UTC)

At present, dates are always stored as Gregorian. There is also a calendar indicator that can be stored. At present, the user interface always displays the Gregorian date. It may include a Julian mark depending if it is calendar indicator is set to Julian and the date falls in a certain range. Basically, it's all fouled up, to the point that no birth or death date before 1924 in wikidata should be trusted. Jc3s5h (talk) 03:24, 20 January 2015 (UTC)
I have since discovered that the word Julian appears alongside if there is an actual date. e.g. Isaac Newton (Q935) Date of birth. If the date is only a year, nothing appears e.g. Clovis III (Q595053) Date of death (although Date of birth is flagged as Gregorian?). I can see the logic - if there is only an approximate date, Gregorian or Julian does not really matter. Periglio (talk) 05:25, 20 January 2015 (UTC)
Your incorrect edits to the data for Isaac Newton shows you do not understand how wikidata stores dates. I suggest you stick to examining diffs and ignore the fucked-up user interface. Wikidata ALWAYS stores Gregorian dates. Jc3s5h (talk) 16:32, 20 January 2015 (UTC)
It is not a question of understanding how the data is stored. I foolishly assumed that when the word Julian appeared it was displaying a Julian date! As I share a birthday with Newton, I knew it was the wrong Julian date. Periglio (talk) 22:27, 20 January 2015 (UTC)

I have created Phabricator bug T87312 for this error. Jc3s5h (talk) 09:45, 21 January 2015 (UTC)

Search defaults to Property namespace

I just did a search for "goose" (for no particular reason) and got no results. (!) This was because only the "Property" namespace was being searched. Is this the intended behavior? I thought the new search engine was supposed to search in a wiki's "content namespaces" by default (which would include the main namespace). - dcljr (talk) 21:50, 20 January 2015 (UTC)

I just tried and it worked correctly for me. Did you use the search box at the top right of this page? --Lydia Pintscher (WMDE) (talk) 21:53, 20 January 2015 (UTC)
May be you toggled the option Remember selection for future searches as I did once? After this all my searches went to the property namespace. --Succu (talk) 21:56, 20 January 2015 (UTC)
Hmm. I can't think of why I might have done that (even accidentally), but I suppose that's the explanation. I'm now getting searches in the main namespace. Thanks. - dcljr (talk) 04:37, 21 January 2015 (UTC)

Question: Bot vs Human edits

If I remove a property of an item which was created by a bot, is the bot going to add this false property again? Thanks for your answer,--Mischa004 (talk) 13:20, 21 January 2015 (UTC)

Contact the user driving the bot. There are tasks, maybe there is something wrong. Greetings, Conny (talk) 13:22, 21 January 2015 (UTC).
Thanks but I just wanted to know if bots generally (and VIAFbot specifically) run for example one time a week and reproduce every mistake they make which I corrected manually (no bot is perfect). I don't ask about one specific case, just in general.--Mischa004 (talk) 13:31, 21 January 2015 (UTC)
It really depends on how the bot is written. In general, I would say a bot should not do this - one should never get in the situation of edit-warring with a bot. If this happens, you should raise it with the bot author / runner, and they should fix this. If they don't, raise it here, and we will fix it. --Denny (talk) 16:00, 21 January 2015 (UTC)
No bot is perfect, that is why is necessary to contact bot onwer/operator. Mostly bot owners are willing to fix its bots errors and modify the code of bots. Unfortunately sometimes small part of bot owners is not willing to fix its bot errors or modify its code (as sometimes they are not able to, as using "prefabricated" code and have no knowledge of code at all). --Jklamo (talk) 17:05, 21 January 2015 (UTC)
You should also contact humans (especially those using Widar-based tools) if they do something wrong, otherwise there's a chance they'll just continue making mistakes (and repeating the ones you've removed).--Mineo (talk) 18:51, 21 January 2015 (UTC)
Thank you both for your answers, that's good to understand!--Mischa004 (talk) 22:12, 21 January 2015 (UTC)

+9,501

Why my edit (adding link to lvwiki article) adds so many bytes? See history and my edit (CTRL+F: Added link to [lvwiki]: 1812. gada karš (Ziemeļamerika)). --Edgars2007 (talk) 16:11, 22 January 2015 (UTC)

It is due some interface changes that were made between August 2014 and January 2015. That was asked twice or three times before, what I found is Wikidata:Project_chat/Archive/2013/09#Small_changes_but_many_bytes. --Stryn (talk) 16:21, 22 January 2015 (UTC)
The internal format was changed twice, in September 2013 and August 2014. If the item hasn't been modified since then, any change causes some other changes inside the item.
It is tracked as phab:T38976. Matěj Suchánek (talk) 17:15, 22 January 2015 (UTC)
Thanks. --Edgars2007 (talk) 03:48, 23 January 2015 (UTC)

Wikidata metrics and targets suggestions for outreach project with UNESCO

Hi all

I'm applying for a WMF PEG grant for funding to work as Wikimedian in Residence at UNESCO, the application is available here

UNESCO have a very large amount of data that could potentially be made available for Wikidata, see UNESCO Institutue for Statistics. Usually I work with organisations to release images and contribute to Wikipedia articles and understand the metrics for those activities quite well. Would someone be able to give me some suggestions of suitable targets and metrics for adding content to Wikidata or point me towards pages that would help?

Many thanks

Mrjohncummings (talk) 01:07, 23 January 2015 (UTC)

No problem with copyright? Wikidata license is CC0 1.0 --ValterVB (talk) 07:32, 23 January 2015 (UTC)

Hi Mrjohncummings, i'm a wikipedian working on some "lists of universities in country XYZ" and i recently found a new useful database: Worldwide Database of Higher Education Institutions, Systems and Credentials (WHED) by International Association of Universities (IAU). In countries like Albania, Iraq, all of Africa etc. there a lots of new founded private/state universities, those same countries do also have problems with diploma mills. Due to different scripts (i.e. arabic), changing names, changing websites etc., universities are actually difficult to identify. I use to add german GND data, but this is not very complete. WHED is quite good, a database with 18000 Higher Education institutions worldwide. IAU is a UNESCO partner org, i believe. Linking this database with wikidata would be worth looking at - and possibly benificial for both sides. --Atlasowa (talk) 14:04, 23 January 2015 (UTC)

For example, Université de la Réunion (Saint-Denis, Réunion) Q834819:
See also http://www.uis.unesco.org/Education/Pages/international-standard-classification-of-education.aspx ISCED mappings. --Atlasowa (talk) 14:14, 23 January 2015 (UTC)

Thanks very much for the information, ValterVB hopefully we will be able to get CC0, Atlasowa I will try and get some information for you. Mrjohncummings (talk) 16:44, 23 January 2015 (UTC)

P988 and P1228

The property for Philippine Standard Geographic Code is duplicated: Philippine Standard Geographic Code (P988) and P1228 (P1228). Tagging: @Exec8: and @seav:. --Bluemask (talk) 07:25, 24 January 2015 (UTC)

@Bluemask: Wikidata:Properties for deletion is the correct place to report duplicate properties.--GZWDer (talk) 09:48, 24 January 2015 (UTC)
@GZWDer: Thanks for the reminder. :) Wikidata:Properties for deletion#Philippine Standard Geographic Code ID (P1228) --Bluemask (talk) 02:37, 25 January 2015 (UTC)

Ana Caram double entry

Hi, can somebody solve this problem, please: https://www.wikidata.org/wiki/Q454715 and https://www.wikidata.org/wiki/Q16336452 are the same person. I don't know how to do it... Thank You. --Oso (talk) 21:03, 24 January 2015 (UTC)

→ ← Merged, you can merge items easily by enabling the merge gadget on your preferences. --Stryn (talk) 21:18, 24 January 2015 (UTC)
Thank you. --Oso (talk) 21:25, 24 January 2015 (UTC)

About articles on translated books

I had a big problem with connecting wikidata of translated works. There are many articles in telugu wikipedia about translated books into telugu. Can I connect this to original book pages of english wikipedia or russian wikipedia? And original book written in other language may or maynot be prominent in my wikipedia. So, what to do? Is there any policy regarding this issues?--Pavan santhosh.s (talk) 04:30, 23 January 2015 (UTC)

Probably the best place to ask is WikiProject Books --ValterVB (talk) 07:36, 23 January 2015 (UTC)
It depends what the the Telugu article is about. If it is about the work, link it to the work. If it is about the translation or a special edition, link it to the edition. The general concept is to have one item for the work and other item(s) to the edition(s) or translation. If it is a translation, it needs to have edition or translation of (P629) that is linked to the work item. However not all users think about the difference between both and sometimes the link in one language is about the work and in another it is about an edition. If you need to use the item to cite a source with page numbers, it needs to be an item about the edition.--Giftzwerg 88 (talk) 12:07, 25 January 2015 (UTC)

Bot generated items of pieces of art

I´ve run across multible items like Portrait of Gabriel Huquier (Q18783431) or Portrait of Count Ravez (Q18822460). No sitelinks, no obivous use as a source to other items, for the most part bot generated in increasing quantities. It´s like some folks out there use bots to create items about all items in the mueseums inventory. RfD or any suggestions how to handle that?--Giftzwerg 88 (talk) 13:08, 23 January 2015 (UTC)

Portrait of Gabriel Huquier (Q18783431) has Commons image, Joconde id and link to Joconde page with image and metadata. Guess it was contributed by Wikidata:WikiProject sum of all paintings. What fault do you find with it? You don't think it's worth having an item for every painting? --Vladimir Alexiev (talk) 16:39, 23 January 2015 (UTC)
Oh sorry! Is was not aware of the links to commons. Every painting that has an image on commons gets an item, I see. Somehow I thougt the links to commons are done like the sitelinks to the articles. --Giftzwerg 88 (talk) 12:14, 25 January 2015 (UTC)

EUROVOC

Maybe we should also consider EUROVOC, the central vocabulary of the EU. 25ish languages. --Vladimir Alexiev (talk) 08:04, 25 January 2015 (UTC)

Item: administrative territorial entity of Saudi Arabia

How should sitelinks (wikipedia articles) be connected in wikidata items: by their names (and therefore by the content which they must have) or by their current content? Me and Ahonc, we have different views on this theme. There exists 2 items Q74063 (region of Saudi Arabia) and Q15623255 (administrative territorial entity of Saudi Arabia) — until yesterday item Q15623255 (administrative territorial entity of Saudi Arabia) had links to 7 wikipedia articles in different languages with similar/equivalent titles to ≈ subdivisions of ~ / administrative divions of Saudi Arabia. Ahonc moved 6 sitelinks to Q74063 (region of Saudi Arabia) because he noticed currently those articles contains informations only about the first-level administrative division of Saudi Arabia (region). I tried to move them back considering (in this case) the title as the primary criterion for connecting articles between them on wikidata, because in this case the title tells what content articles must have, and some day some user looking at article's title will add more content about the other administrative units of Saudi Arabia and then that article will become attached to an wrong wikidata item. I consider the current deficit of content of those 6 articles is temporary, and it can't (shouldn't) determine us to connect those articles to other wikidata item (region of Saudi Arabia) only because those articles right now are underdeveloped. XXN (talk) 00:05, 25 January 2015 (UTC)

Firstly, regions can be not only provinces. Secondly, reason that we should merge articles by similar items. Example: Q2391218 - some items named Istiqlol, some - Taboshar (former city name until 2012). By your suggestion there should be different items: one for Istiqlol, and other for Taboshar (because of different titles). Thirdly. if we divide by items, users may not found articles in other language: for example, if accept your changes, and divide by titles, when I read article in English, I will not know about similar article in Russian WIkipedia, but it exists! (but in different item).--Ahonc (talk) 00:31, 25 January 2015 (UTC)
It is already a frequent problem that various different articles in different languages are not linked, because one is about "Bonnie and Clyde", one about "Clyde Barrow", and one about "Bonnie Parker". Or one about a town or region by one name, others about the town or region combined in a slightly different way with another town or region. Enwiki covers the event, not the person, other wikis cover the person, for a non-notable person involved in a notable event. There is just not a one to one relationship between every language articles and every other language articles. For some subjects, I have to look up the half dozen language articles that are linked, and the other half dozen articles that are linked through a different Wikidata item - with no indication that they are really a dozen articles about the same subject. And we can not do much about it because of the "Bonnie and Clyde" problem. Wixty (talk) 05:38, 25 January 2015 (UTC)
I had just been asking here about the problem of a page covering multiple Wikidata items (but you won't find the question above, because it was just archived without having received any reply). The case I'd encountered was a Wikinews topic cat, n:en:Category:Donetsk, that covers both the city of Donestk and the oblast of Donetsk. I'd since thought of an even worse example in the en.wn topic cats, n:en:Category:Guantanamo Bay, which covers the bay itself, the US territorial site on part of the bay, the US naval base at the site, and the detention camp on the naval base.
It seems to me this "Bonnie and Clyde" problem arises from the particular way Wikidata is being used to automate generation of links on other sisters. Interwiki/sister links aren't, conceptually, about identity of topic. They're about providing to the reader a link to the most-relevant page on another project (be it another-language version of the same sister, or same-language version of a different sister; links to different-language versions of different sisters being handled on a case-by-case basis since they aren't wanted nearly so often). I admit to some puzzlement about how best to handle this. Some approaches:.
(1) Of course we used to handle this by simply specifying at each page which other pages it should link to. The drawback of that is, of course, that individual projects are likely to end up not getting linked to. It's not without advantages; the two chief ones imho being that it promotes both local control of individual projects and customization based on local understanding.
(2) Link each page to a single Wikidata item, and automatically generate interwiki/sister links to and from that page for all, and only, other pages linked to the same Wikidata item. At its best, this does do significantly better at not "forgetting" to link particular pages. Disadvantages include failure to provide useful links to pages that aren't about exactly the same thing (such as the "Bonnie and Clyde" problem), which can also be seen as a loss of customizability, and of course a loss of local control. (Note: although in theory it's still possible for a local project to override the Wikidata-generated interwikis, in practice that's unrealistic atm because it requires the same sort of support infrastructure as (1), and instead of devising a better infrastructure for (1) we're simply dismantling the old one.)
(3) Use some more flexible Wikidata arrangement to generate pages' interwiki/sister links. There's presumably some danger of this getting harder to maintain because of being more complicated, but that's just something to keep in mind while designing it. Designing this would presumably need a Wikidatan insight into the infrastructure of Wikidata; as a Wikinewsie, Wikibookian, and sometime Wikipedian, I know something about the infrastructure of those sisters but I'm not deeply enough immersed in things here to consider myself a Wikidatan.
--Pi zero (talk) 12:43, 25 January 2015 (UTC)

Beta collapsed navigation gone?

I no longer get the collapsed version of the navigation bar in the upper right corner, and it’s no longer in the Beta features list. Was that removed? —DSGalaktos (talk) 10:42, 25 January 2015 (UTC)

Yes it was removed everywhere (not just on Wikidata) unfortunately because it was causing some bad bugs and has no-one working on it to fix them. Hope it comes back soon. --Lydia Pintscher (WMDE) (talk) 10:51, 25 January 2015 (UTC)
Alright, thanks for the info. —DSGalaktos (talk) 11:39, 25 January 2015 (UTC)

Wikidata weekly summary #142

Here's your quick overview of what has been happening around Wikidata over the last week.

Wikipedias with multiple scripts

How do we handle these? (e.g. tt links in Q171428 and Q13404940). It seems some Wikipedias, like tt.wiki, create articles in two scripts. Del♉sion23 (talk) 16:43, 25 January 2015 (UTC)

 Comment: Re. the above specifically, it seems that the Tatar language is variously written in two (if not three, the third being Arabic) scripts. It Is Me Here t / c 16:56, 25 January 2015 (UTC)
this may occur frequently on articles with translitterated names (i.e. from Russian or Chinese), on any wiki... but I also found the reverse for a French person on ru once.
generally, I go on the wp site, and put a Template:Merge (Q6919004) template on both items, just to let them know "caution, this is a dupe... generally they can solve it easily with a redirect...
then I do a merge, except for the duplicate item on wp, and I add instance of (P31)Wikimedia duplicated page (Q17362920) + said to be the same as (P460) and number of duplicate item
once one of the items has been deleted (or redirected), the merge can be completed here.
this was given to me by Ash Crow… --Hsarrazin (talk) 17:32, 25 January 2015 (UTC)
@Hsarrazin: I don't think that treating the issue as a content fork will work in this case. If different speakers of the language use different scripts to write it, it just isn't sensible to force all articles in that language to arbitrarily choose between scripts, and only ever pluck for one in each case. N.B. tt: even has two Main Pages. I've notified them of this discussion over at tt:. It Is Me Here t / c 14:44, 26 January 2015 (UTC)
sorry... I did not get the fact that the language had 2 or 3 different writing alphabet, like serbo-croatian, before the partition of Yougoslavia... :/ --Hsarrazin (talk) 14:49, 26 January 2015 (UTC)
Addendum: those two tt: Main Pages are, respectively, at Wikimedia main page (Q5296) and Wikipedia:Main Page alternatives/(simple layout) (Q13202705). Note also how, at Special:Preferences, you can set the language to tt, tt-Cyrl, and tt-Latn. It Is Me Here t / c 14:53, 26 January 2015 (UTC)
ttwiki users actively create articles in both TATLAT & TATCYR, but in line with tt:Википедия:Латин-Кирил игезәкләр policy, no duplicate titles in TATLAT & TATCYR are allowed for creation after 23 Oct 2013 (using #redirect instead). Internally, heritage articles are linked using tt:Template:TwinLAT/tt:Template:TwinCyr (per previous tt:Википедия:Латин һәм Кирилл мәкаләләре арасында үзара сылтамалар policy). Disputes around efforts to eliminate TATLAT from Wikipedia seized when w:Uniform Turkic Alphabet use was formally approved by the Republic of Tatarstan parliament act of 24-12-2012 (also used by w:Voice of Turkey's TATAR service website), but number of heritage articles still await standardization in line with requirements of tt:Википедия:Тавыш бирү:Латин әлифбасы-2 policy. The issue is further aggravated by the lack of reliable transliteration engine for Tatar, but Tatar Wikipedia community reached consensus about using Cyrillic-only categorization system (work in progress). - 87.117.185.105 17:01, 26 January 2015 (UTC)

Not working for me

It looks like I've been hit by phab:T50389, too. Could someone please add w:simple:Obstetric fistula? It belongs with the article that has same name on the en.wp. WhatamIdoing (talk) 00:13, 26 January 2015 (UTC)

✓ Done Ebe123 (discuter | contributions) 00:47, 26 January 2015 (UTC)

finding redirect and target for one claim

Is there any query or tool to find items which have target and redirect for one claim?

for example: Q530237 at occupation (P106) redirect item (Q15117302) should be removed Yamaha5 (talk) 18:42, 16 January 2015 (UTC)
My bot is fixing such cases. But it waits for ~2 days possible merge revert. — Ivan A. Krestinin (talk) 19:38, 16 January 2015 (UTC)
ThanksYamaha5 (talk) 21:31, 16 January 2015 (UTC)
Ivan A. Krestinin I thought we were leaving redirects from merges so that any external application using our Qitems would see permanent Qitem numbers. Filceolaire (talk) 18:50, 19 January 2015 (UTC)
I think this is about fixing statements that have a redirect as a value to point to the redirect target instead. --Lydia Pintscher (WMDE) (talk) 14:54, 21 January 2015 (UTC)
Sorry, I misunderstood. Filceolaire (talk) 23:04, 26 January 2015 (UTC)

Updating items for languages with a non-Latin alphabet

There is a village in my country whose name officially changed recently. I have updated it in all languages with a Latin alphabet but no idea how to change it in Arabic, Russian or Chinese. I was wondering if there is a way in Wikidata to inform of such things so somebody for these languages can take care of the transliterations if they find it useful. If I remember correctly, in Wikipedia there is a kind of "Embassy" but I guess that in Wikidata, being a multilingual project, such issues must be handled differently. Thanks in advance for your comments! --Quico (talk) 11:05, 18 January 2015 (UTC)

I can change/add name in Hebrew, what is the item number? DGtal (talk) 11:09, 18 January 2015 (UTC)
Item number is Q13837 (L'Esquirol, formerly Santa Maria de Corcó). Thanks.--Quico (talk) 13:30, 18 January 2015 (UTC)
If there is an article about a location in a language, it should be easy to copy and paste that language into Wikidata for the description. I have noticed that our descriptions are very sparse - is it important to add them for every language? Somehow that seems like a duplication of effort - if there is an article for an item, that really should just be the default description for that language, I would think, without having to fill it out. Wixty (talk) 21:04, 18 January 2015 (UTC)
I Added the Hebrew (I found online the correct pronounciation so I could transcribe easily). Descriptions, in my view, are critical mainly in articles that have more than one meaning. In our case the previous name "Santa Maria de Corcó" was unique, but the renewed name L'Esquirol (the squirrel) can easily be also a name of a creative work or another place, so description is important. DGtal (talk) 08:51, 22 January 2015 (UTC)
Properties created for the communication between the editors would be a good solution, similar to the maintenance categories and templates in Wikipedia. --Molarus 09:08, 22 January 2015 (UTC)
Write on the discussion page that the item has been renamed, what is the former and the new name. Give a source. Then go to every language that you can not handle and copy the label to the aliasses (as it is the former name, it should be kept this way), then delete the label. Have the labellister tool in your user setings activated, it is a very usefull tool for this purpose. Someday there will be someone who can write a new and appropriate label in this language. The description can remain untouched. This is a harsh way, but this is the the only effective way it is done to get things fixed. Its better to delete an outdated information than to keep it and hope someone will understand the problem and correct the label some day. We must stop spreading outdated information immediately. It is not of much use to copy the name from the title of the article, because the lemma of the articles might also be outdated. There might be users that undo your changes, then try contacting on the users discussion and revert it again.--Giftzwerg 88 (talk) 19:55, 24 January 2015 (UTC)
Quico I have added a statement using official name (P1448) with qualifiers for the end date for one name and the start date for the other. Can you check I got these right? I listed these as Language:es. Is that right? Should either of these be catalan? Does this comarca have an official catalan name as well as a spanish name? Filceolaire (talk) 23:00, 26 January 2015 (UTC)
Filceolaire, almost perfect, thanks. However the official toponyms in Catalonia are always the Catalan versions so I replaced "Spanish" with "Catalan". In fact most Catalan toponyms don't even have a Spanish version, only main cities and towns have it and, of those, many were "artificially" translated to Spanish during the dictatorship. I'm going to provide a source anyway.--Quico (talk) 23:33, 26 January 2015 (UTC)

Dates of birth and death

Is there any way of extracting from Wikidata's property fields a list of names of people who were born on (e.g.) January 1, or who died on (e.g.) December 31? Bencherlite (talk) 02:12, 20 January 2015 (UTC)

The problem is that Wikidata stores only points of time like "+00000001924-01-01T00:00:00Z". You can see that day, month and year are not separate. --Molarus 03:12, 21 January 2015 (UTC)
That's unfortunate. Thanks for replying. Bencherlite (talk) 09:12, 21 January 2015 (UTC)
I do not know how to extract data searching on a property but I would like to comment that storing as a point of time does not stop you looking for a specific date. You just need to match a range of a day ie date >= "+00000001924-01-01T00:00:00Z" AND date <= +00000001924-01-01T23:59:59Z Periglio (talk) 09:26, 21 January 2015 (UTC)
What do you say about creating new property for using like "Birthday: January 1 (Q2150)"? --Infovarius (talk) 02:17, 27 January 2015 (UTC)

Threats of blocking and status of Wikinews links practice

A Wikidata admin has alleged to me that following Wikidata:Wikinews/Development#Interproject links is a violation of some sort of consensus on Wikidata, unless we first go through a lengthy RfC. This is not how I understood things to have been left, following other discussions here. I've been telling everyone Wikidatans are friendly, and am having difficulty reconciling that with being threatened with blocking if I follow what I have understood to be best practice in Wikinews linking from Wikidata.

Could someone offer a Wikidatan perspective on this situation, please? --Pi zero (talk) 18:57, 20 January 2015 (UTC)

Since no rules has been clearly accepted so far, I don't think you may be blocked. On the other side, why should we connect categories with regular articles as there had been no such system before Wikidata's launch? Matěj Suchánek (talk) 19:56, 20 January 2015 (UTC)
This is the way sister links between Wikinews and Wikipedia have always been done. As long as I've been on Wikinews, anyway (six years and change). Though there have always been idiosyncratic exceptions (e.g. when Wikinews has only one article on a topic, and no category for it), in general a Wikinews topic cat would link to the corresponding Wikipedia article, and a Wikipedia article would link to the corresponding Wikinews topic cat. --Pi zero (talk) 20:02, 20 January 2015 (UTC)
Wikisource "Author:" namespace is linked with Wikipedia main namespace. So the case is not unique. User:Krassotkin can have more comments. — Ivan A. Krestinin (talk) 21:51, 20 January 2015 (UTC)
  • Claims of administrator are not valid. This is our current consensus. While we do not make another decision, we must follow it: Wikidata:Wikinews/Development. --sasha (krassotkin) 09:28, 21 January 2015 (UTC)
    • While this is the current practice, arrived at when we only linked to Wikipedia, it can be changed. An RFC is probably the best route but it doesn't have to be lengthy (though some of them have been). The problem, as I understand it, is that wikinews follows the habit of newspaper websites of having multiple articles about a subject, each of which deals with the latest developments with a cursory summary of what went before. The problem is that for many of these subjects there will be a wikipedia category which seems to be the obvious and logical place for the wikinews category to link to. You need to explain why we shouldn't follow this logic in the case of wikinews. Note that in every case the WD Category item should have a statement linking it to the corresponding wd item and vice versa so there may be a way to use template magic to create a link as you want. I don't understand quite what you want nor do I know enough template magic to advise you there. Filceolaire (talk) 22:41, 26 January 2015 (UTC)
      • To clarify: The consensus referred to is to associate Wikinews topic cats with the Wikidata items for Wikipedia articles; while this is long-established practice on Wikinews, the acknowledgement of it here on Wikidata post-dates linking from Wikidata to Wikinews. As also remarked above, this is not the only case where Wikidata items for Wikipedia articles are associated with non-mainspace pages on other sisters. The reason for this Wikinews practice is not merely that Wikinews may have mutliple articles about a particuar story, but that it may have lots and lots of articles on a topic (say, France), not necessarily about the same story, which go in a Wikinews topic category which corresponds to Wikipedia article France. --Pi zero (talk) 00:49, 27 January 2015 (UTC)

Comment property

Is there a property where we can add a comment intended for Wikidata editors? eg See talk page before changing this item. If not, is it worth proposing? Periglio (talk) 09:12, 21 January 2015 (UTC)

We actually also have the talk page for each item - but doubt anyone would read it before editing, and many edits are made automatically with Widar anyway. A similar idea - how about a property "do not merge with" (or call if "not same as" in analogy to said to be the same as (P460))? For the Thai administrative subdivisions I work on it happens regularly that items are merged even though they are different administrative entities, and as long as there are no Wikipedia articles for both of them in Thai they can technically be merged. Ahoerstemeier (talk) 10:36, 21 January 2015 (UTC)
That is the problem, unsurprisingly no one reads the talk page. It would be nice to add a "comment" reference to a claim. For example, referring editors to decisions made on the talk page. In time, automated software could spot that the claim has been commented and act accordingly. Periglio (talk) 14:31, 21 January 2015 (UTC)
Maybe there should be a kind of magic (but please, not a pop-up ;-)) to raise a bit more attention in case the talk-page is filled. Because a special property might help now most items have 1,2,3 properties filled, but on items where I spend 10 min. of my time, they often get 25 properties filled, several pages of my HD screen full. And then a new property "remark" will not be seen again. Edoderoo (talk) 19:16, 21 January 2015 (UTC)
We could create a comment property but it should ideally be multilingual text datatype (so users here can create translations) and we are still waiting for that datatype. In the meantime you could use related property (P1659) to link between items which might be confused. For the admin divisions you could make sure they have property statements which show they are different which should I hope discourage users from merging. If you find the same user has done more than one of these merges then it is probably worth dropping them a note on their userTalk page. Filceolaire (talk) 22:20, 26 January 2015 (UTC)

AGROVOC Controlled Vocabulary

I've spotted AGROVOC, a huge controlled vocabulary made by the FAO of "over 32,000 concepts with up to 40,000 terms in 20 languages: Arabic, Chinese, Czech, English, French, German, Hindi, Hungarian, Italian, Japanese, Korean, Lao, Persian, Polish, Portuguese, Russian, Slovak, Spanish, Thai, Turkish." (https://en.wikipedia.org/wiki/AGROVOC)

It's a "linked data set aligned to 13 other vocabularies" (https://en.wikipedia.org/wiki/AGROVOC#Linked_data) It says that it is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License for thesaurus content in English, French, Russian and Spanish. The rest seems subject to potential copyright by producing organisations (https://en.wikipedia.org/wiki/AGROVOC#Copyright_and_license) As part of the Food Project on Wikidata, I was wondering if it would be relevant:

  • create a relevant property proposal (AGROVOC id)
  • add the vocabulary to Mix N'Match to link it with Wikidata.

--Teolemon (talk) 21:10, 23 January 2015 (UTC)

 Support 32k is not so huge, and AGROVOC is quite specialized, but why not --Vladimir Alexiev (talk) 08:04, 25 January 2015 (UTC)

indeed, I'm very interested by the food part of it, but Eurovoc might have good stuff too given the amount of food legislation in Europe --Teolemon (talk) 12:35, 25 January 2015 (UTC)
We can create this property. Propose it on the Wikidata:Property proposal/Authority control page.
As wikidata is CC0 we cannot import the content but we can (for instance) compare our data with this vocabulary and highlight discrepancies. Filceolaire (talk) 22:02, 26 January 2015 (UTC)

Class Instance Analysis

We at Ontotext (Q7095072) have taken a count of all direct type ("instance of") statements as of Jan 2015: wikidata-instance-count-201501.txt: thanks to Atanas Popov.

Here is some analysis, all cited files are available on Gist --Vladimir Alexiev (talk) 08:20, 25 January 2015 (UTC)

Counting

  • Total instance statements
$ perl -ne 'm{^(.*)\|} and $n+=$1; END {print $n}' wikidata-instance-count-201501.txt
  • 12331093

Wikidata has 13M items. As expected, not all items have a type. TODO: how many items have several instance statements?

Total Classes

With at least 1 instance:

  • 17875

Highest-cardinality Classes

2662626|Q5|human
2409399|Q4167836|Wikimedia category page
1895952|Q16521|taxon
800880|Q4167410|Wikimedia disambiguation page
588520|Q13100073|village-level division in China
186119|Q13406463|Wikimedia list article
183324|Q11266439|Wikimedia template
164317|Q486972|human settlement
154125|Q482994|album
145863|Q532|village
140820|Q11424|film

Medium-cardinality Classes

Classes with at least 5 instances:

  • 6510 (36%)

Having >=5 instances is no guarantee that all they are used appropriately, since the question what are appropriate classes in Wikipedia is far from settled. Eg can some ship be instance of "Deutschland-class battleship", or should it be instance of "ship", while "Deutschland-class battleship" is assigned to eg "ship type"?

Specific classes

  • There are 2.6M (2662626) human. This is fairly well focused, in that it collects a large proportion of all humans. There are a few exceptions, eg "minister", "table tennis player", "chess composer": these should be used as "occupation" while "instance of" should be "human".
    • Note: Wikidata uses "human" and not "person" for people. "Person" is anything that can bear a personality, eg deity, artificial agent, etc
  • There are 5k families: 4569 noble family, 635 family, 465 Dutch noble family, 95 Belgian noble family, 35 clan
  • There are 40k+20k names: 40038 family name, 10320 given name, 5569 male given name, 4828 female given name.
    • Due to the good efforts of the WikiProject "Wikidata names", these items provide valuable information on names themselves, eg variations, male/female correspondences, etc.
    • This can probably be used for disambiguation or for generating language-specific name variants, but we have not investigated this topic
  • There are some 22k literary characters: 11993 fictional character, 6963 fictional human, 2589 mythical character, 357 group of fictional characters, 159 fictional organization
  • There are at some 215k organisations (not counting governments, city councils, etc). These are spread across a wide list of classes:
    • 55k businesses: 47149 company, 2653 business, 2321 transport company, 885 public company, 718 corporation, 152 motorcycle manufacturer, 95 joint-stock company, 80 holding company
    • 66k creative organizations 42179 band, 17904 radio station, 6187 newspaper, 1540 film production company, 843 theatre company, 22 theatre troupe
    • 31k sports clubs: 26200 association football club, 5376 sports club, 184 American football club, 169 golf club, 154 country club
    • 30k educational institutions: 16611 high school, 6396 school, 6321 university, 1062 Engineering College, 771 college, 301 research institute
    • 20k non-profit organisations: 8929 organization, 7026 political party, 2853 association, 1052 nonprofit organization, 307 international organization, 246 charitable organization, 226 Esperanto organization, 144 political organization, 73 non-governmental organization
    • 13k GLAM: 438 art gallery, 83 art gallery; 882 library, 199 national library, 114 public library, 60 library, 28 Carnegie library, 27 academic library, 16 municipal library; 108 archive, 26 cantonal archives, 24 municipal archive; 6516 museum, 2176 art museum, 873 military museum, 569 museum ship, 513 historic house museum, 181 maritime museum, 151 mus?e de France, 119 aviation museum, 80 natural history museum, 68 science museum, 57 open-air museum, 48 railway museum, 37 local museum, 37 children's museum
  • Some 500k Creative Works: 154125 album, 140820 film, 59242 single, 51765 book, 31623 painting, 23055 scientific journal, 20032 song, 26789 video game, 18338 television program, 14838 short film, 13461 television series, 13098 silent film, 11876 periodical literature, 11297 episode, 6739 literary work, 6627 television season, 3488 sculpture, 2374 manuscript
  • Some 110k heritage sites and monuments: 64806 Rijksmonument, 21076 Iranian National Heritage, 19696 scheduled monument, 1370 natural monument, 1150 World Heritage Site. This is expected to grow sharply for other countries as well.
  • There are 801k instances of "Wikimedia disambiguation page". TODO: are these already mapped to alt labels or not?

Class Defects

Low-cardinality Classes

There is a very long tail of items that are inappropriately used as classes:

  • Eg Indian Rhinoceros, Trumbull's Declaration of Independence, stud, meatloaf...
  • Eg "minister", "table tennis player", "chess composer" should become "occupation", and the instances moved to "human"

Classes with less than 5 instances:

  • 11365

All these should likely be assimilated into a bigger class.

Classes With no Label

$ grep -E '\|$' wikidata-instance-count-201501.txt | tee wikidata-instClass-anonymous.txt | wc -l

wikidata-instclass-anonymous.txt: 1408, eg 

218|Q12600360|
194|Q18116579|
167|Q366301|
156|Q14858002|
  • There's also a class with the weird label "Q"
Q217629|Q

Classes With Duplicate Labels

$ cut -d '|' -f 3 wikidata-instance-count-201501.txt|sort|uniq -d|tee wikidata-instClass-duplicate.txt |wc -l

wikidata-instclass-duplicate.txt: 265. In principle items could have duplicate labels, but I don't think that is acceptable for a class!

Do you mean some specific language or all set of labels? --Infovarius (talk) 03:45, 27 January 2015 (UTC)

Classes With a Colon in the Label

$ grep -E ':' wikidata-instance-count-201501.txt | tee wikidata-instClass-colon.txt | wc -l

wikidata-instclass-colon.txt: 86. Obviously a category or template should not be used as a class. Eg 

224|Q4913949|Help:Infobox
36|Q6512805|Category:FIS Alpine World Ski Championships
31|Q4656638|Wikipedia:Categorization of people
15|Q4168710|Help:Template
9|Q8264688|Category:Musical duos
5|Q7145070|Category:Herbicides
5|Q4427|Category:1888
5|Q6266825|Category:Horse races
4|Q6540697|Wikipedia:Books
4|Q5616006|Template:Infobox television

Add more calendars?

Tracked in Phabricator
Task T87312

I think that it is sometimes relevant to the subject of an article that a date be included in a non-standard calendar. E.g. Saint Bashnouna (Q7401025) is a Coptic Christian martyr whose saint day is given as a Coptic as well as a Gregorian calendar date on en.wp. Perhaps such calendars could also be imported to Wikidata and added to entries where appropriate? It Is Me Here t / c 14:22, 25 January 2015 (UTC)

I suggest that maybe we should learn to use the two calendars we already have, fix the user interface for these, and correct all the false data in the database, before we add another calendar. In this case, dates before the adoption of the Gregorian calendar on 15 October 1582 are usually stated in the Julian calendar. More likely than not, the death date, 19 May 1164, stated in the English Wikipeda (which was imported into Wikidata) is a Julian calendar date. But the date stored in Wikidata, 19 May 1164, is a Gregorian calendar date. Furthermore, the English Wikipedia fails to cite a reliable source for either the Julian(?) date or the Coptic date.
Another consideration is how to provide references for dates from the various calendars. Should each calendar date have it's own references, or should all the references apply to all the calendar dates? In the case of calendars where the conversion is debatable, such as the Roman Republic, I think separate references are needed for each calendar. Jc3s5h (talk) 17:41, 25 January 2015 (UTC)
The calendar conversion support in Emacs indicates that the Gregorian date 26 May 1164 corresponds to the Julian date 19 May 1164 and the Coptic date Bashans 24, 880, which further supports the premiss that the Wikidata date is wrong. Jc3s5h (talk) 18:55, 25 January 2015 (UTC)


OK, I've changed it to Julian per your advice. But this is another argument for allowing original dates: if the software can auto-convert from the Coptic etc. calendar (where we are sure of what the date is), such Julian/Gregorian conversion errors will become easier to spot. It Is Me Here t / c 14:32, 26 January 2015 (UTC)
I have undone your change for the reasons explained in the thread above, #Which calendar was used?. In short, you stored 1164-05-19 as the death date. You can see this by looking at the diff for your change. But the date must always be stored as a Gregorian date. Jc3s5h (talk) 17:08, 26 January 2015 (UTC)
Let me address your sentence fragment "if the software can auto-convert from the Coptic etc. calendar". Once upon a time the software could convert between Julian and Gregorian. I wasn't active back then, but I understand that if, in the user interface, you had entered 19 May 1164 and marked that it was Julian, it would have been stored as 26 May 1164 and the calendarmodel value would have been set (in effect) to Julian. Then when it was displayed in the user interface the conversion would have been done in reverse. But then the actions performed by the various modules was rearranged, with the result that the ability to do conversions was completely removed. So even if the calendarmodel value is set to http://www.wikidata.org/entity/Q1985727 and as a result the word "Julian" is displayed in the user interface, the date is not converted and is displayed as 26 May 1164. Jc3s5h (talk) 19:50, 26 January 2015 (UTC)
Jc3s5h. Do you know if their is a bug registered for this? Filceolaire (talk) 21:43, 26 January 2015 (UTC)
Yes, T87312. I am, in this edit, marking this thread as related to that bug. Jc3s5h (talk) 22:24, 26 January 2015 (UTC)

Wikipedia using Wikidata as source

Where can I get a list of Wikipedia articles which includes data from Wikidata (not sitelinks, languages or other metadata)? I'm especially interested infoboxes which are used in Wikipedia but have their source in Wikidata. --213.221.236.164 19:36, 26 January 2015 (UTC)

Category:Templates using data from Wikidata (Q11985372) lists some templates for each linked project. Matěj Suchánek (talk) 19:45, 26 January 2015 (UTC)

More date problems

Tracked in Phabricator
Task T66084

Just discovered that the output from a date varies depending on the number of characters in a year. I was expecting the year to always be 11 digits. It caused me a problem as I was doing a string position to extract the date. I feel I should tell someone, but I do not know who! Periglio (talk) 19:43, 26 January 2015 (UTC)

  • +000000019-10-10T00:00:00Z
  • +0000000899-10-26T00:00:00Z
  • +00000001999-02-24T00:00:00Z
Tracked on Phabricator at T66084 ·addshore· talk to me! 21:25, 26 January 2015 (UTC)

GUI issues

I have to say that the way interwiki links are oriented now makes it difficult (for me at least) to add links. The keyboard tag (which gives us the option to use the typical keyboard for our language or the pc-keyboard) hovers above the auto-completion tag for instance. I had to change zooming in order to click it. I think that the new interwiki tables are small and that overall the previous GUI arrangement was more usable. -- Spiros790 (talk) 10:12, 15 January 2015 (UTC)

Putting them side by side makes them very difficult for me. Please put them back in line, one above another. Wixty (talk) 19:24, 15 January 2015 (UTC)
In fact the page is virtually useless in the current format. It was better with the "wiki" after each language also. This is a horrible disaster. Wixty (talk) 19:30, 15 January 2015 (UTC)
This is what I use Wikidata for. To make a list of each article in every language. What I was doing is copy and pasting from Wikidata but I can not do that now. for Q3247786 This is what I get:
en Kotaro Nakao ja 中尾幸太郎 nl Kotaro Nakao simple Kotaro Nakao tr Kotaro Nakao zh 中尾幸太郎
These need to be tab separated each on a separate line, like this, and with no other column next to them, so that I can use copy and paste to copy them into a text editor:
English	enwiki	Kotaro Nakao
Japanese	jawiki	中尾幸太郎
Dutch	nlwiki	Kotio
Simple English	simplewiki	Kotaro Nakao
Turkish	trwiki	Kotaro Nakao
Chinese	zhwiki	中尾幸太郎
This is extremely important to me and saves me thousands of hours of time. Note, I delete the language column (English etc.) and it can be in native language like it used to be or in user preference language as shown above, and I only use the other two columns, enwiki and the name of the article. Note that I need them to be tab separated so they can be entered into a spreadsheet. I need the "wiki" there so that I can visually confirm that I have the language code. Wixty (talk) 19:50, 15 January 2015 (UTC)
And in trying to list them for this example I missed two of them even though there are only six - and just think what is going to happen when there are two hundred for an article??? Wixty (talk) 19:56, 15 January 2015 (UTC)
@Wixty: There will propably be some gadgets to append the strings to the language code.
AFAIK the boxes have their own classes, so you can use your CSS to change their look. Matěj Suchánek (talk) 20:00, 15 January 2015 (UTC)
Also moving the name of the language to a mouseover instead of in its own column is just plain stupid. I am accustomed to the fact that "Finnish" is "Soumi", and "Russian" is "Русский" and scan down the list to quickly find them. Having them be as a mouseover is about as useful as putting them on the Moon. Whoever came up with that clown idea needs to be fired. Here is the fact. Yes I need a workaround, as it literally would cost me thousands of hours to try to use the current GUI - and as already demonstrated highly prone to error. I do not know css or gadgets or any other programming but if someone could come up with a work around for me personally to use it will be extremely valuable. But the point is, if I am using Wikidata for this, then tens of thousands of other readers are as well, and it really needs to be put back to the previous inline, tab separated format. Immediately if not sooner. Wixty (talk) 20:09, 15 January 2015 (UTC)
To give an idea of the magnitude of the problem, right now I need the Wikidata lists of about 20 to 30 country specific articles in all of the wikis that cover those articles. That is an enormous task with the current format, and takes seconds with the old format. And tomorrow I will need another set of 20 or 30 articles in all of the languages they are in. And the next day another set. And all of those sets today with the old format. Wixty (talk) 20:18, 15 January 2015 (UTC)
In the meantime I do have one work-around - list page source, and use this:
"sitelinks\":{\"jawiki\":{\"site\":\"jawiki\",\"title\":\"\\u4e2d\\u5c3e\\u5e78\\u592a\\u90ce\",\"badges\":[]},\"simplewiki\":{\"site\":\"simplewiki\",\"title\":\"Kotaro Nakao\",\"badges\":[]},\"enwiki\":{\"site\":\"enwiki\",\"title\":\"Kotaro Nakao\",\"badges\":[]},\"trwiki\":{\"site\":\"trwiki\",\"title\":\"Kotaro Nakao\",\"badges\":[]},\"zhwiki\":{\"site\":\"zhwiki\",\"title\":\"\\u4e2d\\u5c3e\\u5e78\\u592a\\u90ce\",\"badges\":[]},\"nlwiki\":{\"site\":\"nlwiki\",\"title\":\"Kotaro Nakao\",\"badges\":[]}}}"
Pretty messy, but at least I can use global search and replace to create the tabs that I need. So with this work around, maybe half an hour for what was taking me a few seconds. Wixty (talk) 20:27, 15 January 2015 (UTC)
No that won't work - it has u4e2d instead of a chinese character. Wixty (talk) 20:29, 15 January 2015 (UTC)
Any suggestions? Wixty (talk) 20:31, 15 January 2015 (UTC)
Can be useful? https://www.wikidata.org/w/api.php?action=wbgetentities&format=xml&props=sitelinks&ids=Q42? --ValterVB (talk) 22:44, 15 January 2015 (UTC)
You can also try with format=txt and you can query more items with one request: ids=Q42|Q89 --ValterVB (talk) 22:48, 15 January 2015 (UTC)
Thanks for trying. It looked promising but as I try to use it in practice it does not work. txt is useless because it puts them on multiple lines:
                                   [site] => arwiki
                                   [title] => دوغلاس آدمز
                                   [badges] => Array
and if I try to use json instead of txt I get an even more useless {"site":"eswiki","title":"Douglas Adams","badges":[]}
This is pretty clunky but it works.
<sitelink site="arzwiki" title="دوجلاس ادامز">
<badges/>
</sitelink>
<sitelink site="azwikiquote" title="Duqlas Noel Adams">
<badges/>
</sitelink>
<sitelink site="barwiki" title="Douglas Adams">
<badges/>
which I can then sort to get rid of all the lines that I am not using, and then do a global replace to change " title=" to a tab. Whoops what's that "azwikiquote" doing in there? But please please please just put them in three columns, language, language code (enwiki - with "wiki") and the article title, and make sure that when you copy and paste the columns are tab separated.
It looked promising but when I try to use it I am having problems, perhaps because of different ways of implementing end of line characters. So far I am still stuck. Wixty (talk) 01:17, 16 January 2015 (UTC)
Good news. I finally got [6] to work by doing some more global replacements. But it is a horrendously round-about way to do something that used to be very simple. Wixty (talk) 01:41, 16 January 2015 (UTC)
Doing more than one at a time is interesting but since I have to do a sort to get rid of the extra lines it would just jumble them together and not be useful. Wixty (talk) 02:10, 16 January 2015 (UTC)

To copy a list of interwikis? For this I'd use importScript( 'User:Ricordisamoa/InterwikiList.js' ); in my common.js --Infovarius (talk) 03:56, 16 January 2015 (UTC)

How do I do that? I am not programmer savvy. But it is not the list that I need, but the language association with each item. Wixty (talk) 04:34, 16 January 2015 (UTC)
What does the script do though? It says "'old' interwiki links". I want the list from Wikidata. Wixty (talk) 05:10, 16 January 2015 (UTC)
What I end up with is a tab separated list like this:
en	Kotaro Nakao
ja	中尾幸太郎
nl	Kotaro Nakao
simple	Kotaro Nakao
tr	Kotaro Nakao
zh	中尾幸太郎
That is what I end up with, but I need "wiki" added to each language to confirm that I have the right item, which I then delete. Wixty (talk) 05:16, 16 January 2015 (UTC)
To get it to work, I am doing a global replace of " title=" with tab, wiki with tab wiki tab, and deleting all the other extraneous stuff. Then I sort by the column after "wiki" which puts all the "wikiquote" etc names by themself, then resort the remaining items and delete the two blank columns. A lot of work but it does give me the list I need. Wixty (talk) 05:20, 16 January 2015 (UTC)
I am saving a little bit of time because https://www.wikidata.org/wiki/Q30 takes 90 seconds to load, but https://www.wikidata.org/w/api.php?action=wbgetentities&format=xml&props=sitelinks&ids=Q30 only 2 seconds, or less. Wixty (talk) 05:56, 16 January 2015 (UTC)
I did a suggested layout. No surprise that it looks a lot like the original layout.
Suggested Wikidata layout
Even if the FA/GA symbols were not in a separate column, it is important that they are at the beginning of the file name, not at the end, so that they are lined up, and not stagger spaced, to make them more easily identifiable. Wixty (talk) 21:22, 16 January 2015 (UTC)
Wixty, I like your layout. I think it would be way more usable than the existing. -- Spiros790 (talk) 22:11, 18 January 2015 (UTC)
@Wixty: Just an experiment, if you have a gmail account, try with this Google sheet. I hope you can open the sheet because I'm not expert with Google sheet and is my first experiment with JavaScript (at least I think is Java script ). When you open the sheet you must see a new menu called Wikidata. In cell B1 you must input the item (ex. Q42 for one item or Q42|Q11920|Q1234 for more item), then click on menu Wikidata - >Sitelink by item after some second you have all the sitelink. Probably is possible to optimize the code, but I'm not so expert. If you have problem or you want change something ping me. --ValterVB (talk) 18:53, 22 January 2015 (UTC)
The output looks good, but I am not sure I feel comfortable with this screen:
Wikidata would like to:
More info	View and manage your spreadsheets in Google Drive	Click for more information
More info	Connect to an external service
I just want to be able to copy and paste them from Wikidata, just like I always could. It is silly to have to go through an external application for something as simple as this. Wixty (talk) 08:21, 24 January 2015 (UTC)
By the way, even though I ultimately throw it away, the first column - the language column - was very helpful, so that I was sure that something had not gone awry - it was very convenient to see Nederlands next to nlwiki, Deutsch next to dewiki, just to make sure that I had the right item. I have had columns get misaligned, and having the language is a convenient double check. Wixty (talk) 08:33, 24 January 2015 (UTC)
Actually the spreadsheet does not save me any time because it still includes wikivoyage, wikinews etc. The XML is a lot faster and easier to use. It is really incredible that it loads 50 to 100 times faster than Wikidata, yet both have exactly the same data. Why do we have a GUI that is so incredibly slow? XML and Wikidata for "United States". What is it even doing all that time? Wixty (talk) 01:44, 25 January 2015 (UTC)
@Wixty: I have modified the script, now there are two menu, Wikilink by item extract only wiki link, and now is faster. For the problem with the GUI speed, there are two point: 1) XML query extract only sitelink without labels, descriptions and claims, 2) GUI must load a lot of script and must formatting the page. --ValterVB (talk) 10:04, 25 January 2015 (UTC)
Thanks. 06:57, 26 January 2015 (UTC)
I have run into a rather odd problem now though, that I can not copy and paste Q1400, and several others, like Q61, and Q1384, directly from one spreadsheet to another (I only tried Q1400 multiple times). They crash Open Office every time. I had to put the items into notepad first, and then enter each as tab separated text. Wixty (talk) 04:08, 28 January 2015 (UTC)

WikiProject Authority Control?

-> see now: Wikidata:WikiProject Authority control

--Atlasowa (talk) 11:18, 27 January 2015 (UTC)


The best ? project

The best article projects exist in Wikipedia. We do not have something like it currently. I think this is important part of wikilife and we need to organize this project.

Key question for us is: that entities can be named the best? Candidates:

  1. Single item This is small piece of information. Too small time is needed to fill all possible values. How to define completeness criteria? New properties creation procedure delays property creation for many months. New properties will make the job incomplete.
  2. Group of items How to define the group? Same issue with completeness criteria as for single item.
  3. Property This is job required some notable time. Completeness criteria can be defined for many properties (but not for all).

The best property project is looked better than others. The best property, what is it? Criteria:

  1. documented — the property documentation is complete and clear;
  2. useful — the property is used in other projects;
  3. complete — the property is filled for all possible items;
  4. verifiable — references present;
  5. validated — validation procedures present and work;
  6. valid — low or zero error rate;
  7. actual — update/synchronization procedure presents and works continuously.

Comments? Ideas? — Ivan A. Krestinin (talk) 19:39, 27 January 2015 (UTC)

Date format

Hi. At the Danish version of Wikidata, the date format i wrong. In Danish, we use at dot after the number. Fx 4. december 2014 for 04.12.2014. Who is able to change that? --Laketown (talk) 20:56, 27 January 2015 (UTC)

Yep. I think it's phab:T63958. --Stryn (talk) 21:42, 27 January 2015 (UTC)
Can someone include Danish, or a link to this post? That references only pt as far as I can see. There are also half a dozen languages that use funny numbers instead of arabic for the year. You can see them if you for example scroll through the pages for 2014. Or 2010. Wixty (talk) 05:07, 28 January 2015 (UTC)
I edited the task on Phabricator instead, so no need to create a new task for every languages. --Stryn (talk) 19:03, 28 January 2015 (UTC)
Also Chinese, Japanese, Korean and several Chinese dialects. The correct format is:
  1. y年m月d日 (for zh, ja). example is 123年4月5日.
  2. y년 m월 d일 (for ko). example is 123년 4월 5일. Note there're spaces between year, month and day.
  3. y年m月d号 (for gan-hans, wuu).
  4. y年m月d號 (for gan-hant, yue).
  5. y nî m go̍eh d ji̍t (for nan).
  6. y n. m ng. d h. (for cdo).
  7. y-ngièn m-ngie̍t d-ngi (for hak).

--GZWDer (talk) 05:21, 29 January 2015 (UTC)

Protect Edgar Allan Poe

Edgar Allan Poe (Q16867) (Edgar Allan Poe, in case the label changes again) is getting flooded with spam. Can someone protect it please? —DSGalaktos (talk) 22:55, 27 January 2015 (UTC)

✓ Done. Sorry for not noticing this last night. Jared Preston (talk) 07:32, 28 January 2015 (UTC)
School kids. Wixty (talk) 08:18, 28 January 2015 (UTC)
Thanks! —DSGalaktos (talk) 09:02, 28 January 2015 (UTC)

Data import policy

Hi, I want to import a cc0 database (about 250.000 streets in the Netherlands) with sufficient claims (P31,P131,P276,P625). What is the policy, I can just start my own import project or should I aks permission somewhere? Michiel1972 (talk) 08:09, 28 January 2015 (UTC)

Yes, please file a request at Wikidata:Requests for permissions/Bot.--Ymblanter (talk) 19:11, 28 January 2015 (UTC)
Ah thanks. I made a request. Still, it is a a littlebit strange any user can use http://tools.wmflabs.org/wikidata-todo/quick_statements.php to import a database as new items without doing a request. But when using a bot-account permission is needed. Michiel1972 (talk) 23:02, 28 January 2015 (UTC)

Mahavira Hall / Main Hall (Japanese Buddhism)

The article Mahavira Hall on the English Wikipedia links to wrong articles in other languages. Tried to fix it at least German/English. In German it should be, Haupthalle (Buddhistischer Tempel Japan); in French Bâtiment principal (Bouddhisme japonais), in Japanse 本堂 etc. etc. Actually the article Main Hall (Japanese Buddhism) should link to all those articles as it is about a Japanese building type not a Chinese one. --Catflap08 (talk) 16:13, 28 January 2015 (UTC)

The en, and nl article are about chinese temple, with references, you should interact there. nl says Mahavirahal is chinese and japanese is: Gouden hal" (金堂) en "Hoofdhal" (本堂), main hall.--Oursana (talk) 17:24, 28 January 2015 (UTC)

To be honest I am not versed enough on how to insert the links properly – just noticed the way the articles are linked is wrong. The English article on "Main Hall (Japanese Buddhism)" should link to the German “Haupthalle (Buddhistischer Tempel Japan)” and French “Bâtiment principal (Bouddhisme japonais)”. The Japanese alias would be “Kondō”. In the en article on "Mahavira Hall" the most that there could be is a “See also”-note to “Main Hall (Japanese Buddhism)”. --Catflap08 (talk) 18:12, 28 January 2015 (UTC)

Just saw changes were done! Thanks a million who ever you were BOT or human :-) --Catflap08 (talk) 19:05, 28 January 2015 (UTC)
Just try again next time. There need to be changes in your preferences: choose move, also merge is helpful. And with move (a green bottom) you can move the site link directly to the right item. Thxs for your hint.--Oursana (talk) 19:39, 28 January 2015 (UTC)

wdq.wmflabs.org – 502 Bad Gateway

What happened to wdq.wmflabs.org? Seems to be down. --Jobu0101 (talk) 17:57, 28 January 2015 (UTC)

Toollabs had to be updated and restarted because of a security update. Wikidata Query didn't come back up. Tracking at phabricator:T87765. --Lydia Pintscher (WMDE) (talk) 18:14, 28 January 2015 (UTC)
Now it is back online but not up to date (see next section). --Jobu0101 (talk) 20:01, 28 January 2015 (UTC)

Is there a height data tag?

Tracked in Phabricator
Task T77977

I wanted to add the height of a person, Atsamaz Bekoyev to be exact, but I could not find a height tag. Do you know of one? I will gladly make one if there is none. DinoD123 (talk) 03:35, 29 January 2015 (UTC)

@DinoD123: This is phab:T77977.--GZWDer (talk) 05:12, 29 January 2015 (UTC)

Wikipedia app and Wikidata vandalism

de:Adventskalender with prominent vandalism from Wikidata item Christmas calendar (Q152862), which remained unnoticed by editors for almost two months

As I just learned from a discussion in dewiki, Wikimedia's official Wikipedia app for Android displays the description of the according Wikidata item just below the title of any Wikipedia article for some time now. This probably gives them much more attention than they had before, unfortunately combined with serious troubles even for more experienced users to find out how to edit these descriptions (the app doesn't allow it at all, the browser version doesn't show it by default and requires you to find out yourself that the description is coming from Wikidata). If this feature of the app is kept, we should try to take way more care about these descriptions. They should look nice and correct (while currently many contain misspellings and strange language), and of course free of vandalism (while it's by far the most common type of vandalism that appears on Wikidata, with hundreds of edits per day, many of them going unnoticed) and edit accidents (languages as descriptions, wrong language, names as descriptions, etc.). Thank you for your attention. --YMS (talk) 17:41, 29 January 2015 (UTC)

anarchism (Q6199) as a statement for political alignment (P1387)

I'm trying to import of anarchist from .en category, and was informed that for political alignment (P1387) values should be one included in the political spectrum (Q210918) - in my opinion that is not possible for Außerparlamentarische Opposition (Q531996) I would like to discuss the use anarchism (Q6199) as a statement for political alignment (P1387). I think it is not feasable to put anarchists (and some others) in a left / right schema. --FischX (talk) 08:21, 30 January 2015 (UTC)

political alignment (P1387) is indeed quite restrictive in its values. Perhaps it's time to expand the domain of political ideology (P1142) to include persons. —Wylve (talk) 16:33, 30 January 2015 (UTC)

GLAM-WIKI 2015 conference, April 9-12 2015, The Netherlands

Wikimedia Nederland welcomes interested Wikimedians and GLAM enthusiasts to join us at the GLAM-WIKI 2015 conference, from 9 - 12 April 2015 in The Hague, The Netherlands. The call for proposals and application for scholarships are now open!
Ter-burg (talk) 14:59, 30 January 2015 (UTC)

Why is this one disabled for editing

I go to this page Q4935149 and I get no edit boxes or links? It was not the Bobby Hughes I was looking for, but while I was on the page I was going to fill in the blank labels and description. Periglio (talk) 21:39, 28 January 2015 (UTC)

Hmm. Looks okay here. Did you try to purge your cache (add &action=purge to the end of url and press enter)? --Stryn (talk) 22:06, 28 January 2015 (UTC)
You'd have to add ?action=purge to the end of the URL to purge, and strangely enough, I had to do this too for the item in question. Jared Preston (talk) 23:53, 28 January 2015 (UTC)
I forgot I posted this message! I just tried again (without purge) and its fine for me now. I'll try and remember the purge command if it happens again. Thanks Periglio (talk) 23:09, 30 January 2015 (UTC)

Perennial issue: False friends. Another case: Spiritism / Spiritualism

Maybe some of you remember the "pamphlet" mess - now, I hope, happily sorted out with Q190399 and Q18536349. Well, I have come across another one. German Wikipedia's de:Spiritualismus (Theologie) is, through Q829348 ("belief that one can communicate with the spirits of the dead") linked with en:Spiritualism - which is wrong, as the German article isn't at all about communicating with the spirits of the dead, but deals with a particular theological point of view in Christianity (denying church doctrines and asserting direct inspiration by the Holy Spirit for every human). The "belief that one can communicate with the spirits of the dead" is known as Spiritismus in German. That article is interwiki-linked through Q15241213 (no description) with various language versions. But I haven't yet found a Wikidata item for "spiritualism" in the theological sense - is there one? There's also Q2636432 (no description) for en:Spiritualism (philosophy), "the notion ... that there is an immaterial reality that cannot be perceived by the senses". But that's a different thing, again. And another "spiritualism": Q7578143 - lacking a description as well, interwikilinked only to English ("a dualist metaphysical belief that the world is made up of at least two fundamental substances, matter and spirit"), Turkish, and Urdu Wikipedia articles. So, it would be nice if someone were willing to clear this up - the items need a description, and maybe a new item for the concept described in de:Spiritualismus (Theologie) is needed, if there is indeed no other Wikipedia having an article about theological spiritualism. That there's no English interwiki for de:Spiritismus is something that should be fixed as well, probably through Q829348 (but there's also Q214255 for Kardec's spiritism). Gestumblindi (talk) 21:13, 31 January 2015 (UTC)

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat/Archive/2015/01&oldid=1102156887"