Wikidata:Project chat/Archive/2014/12

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.

Several concepts treated in a single wikipedia page and interwiki links

Hi. I encountered a problem when I tried to add a very reasonable interwiki link between the french fr:diazote page (about Nitrogen gas) and the english en:Nitrogen page (about both the atomic element and the gas molecule). The problem is that in french we choose to have two separate pages for the atomic element and the chemical compound, so the fact that there is already a link between fr:Azote (page about the atomic element) and en:Nitrogen makes it impossible to add another link between fr:Diazote and en:Nitrogen#Nitrogen gas. I don't imagine we can force the english wikipedia to create two separate pages, so why don't we allow interwiki links to point to subsections of a page? Something like fr:Azote <-> en:Nitrogen and fr:Diazote <-> en:Nitrogen#Nitrogen gas PiRK (talk) 07:27, 1 December 2014 (UTC)

Same problem with Oxygen / Dioxygen, btw. --PiRK (talk) 07:32, 1 December 2014 (UTC)

@Snipre: Can you look at this? --Tobias1984 (talk) 11:03, 1 December 2014 (UTC)
In case my lengthy explanation confused you, my problem is that I can't find a way to add a link to the english wikipedia here: https://www.wikidata.org/wiki/Q2370426#sitelinks-wikipedia PiRK (talk) 14:14, 1 December 2014 (UTC)
@PiRK: Try this solution: delete the redirect of the page en:Dinitrogen to en:Nitrogen then create the link between WD and en:Dinitrogen, et finally recreate the redirectin the page Dinitrogen to nitrogen in EN:WP. Snipre (talk) 15:03, 1 December 2014 (UTC)
✓ Done for dinitrogen. Snipre (talk) 16:15, 1 December 2014 (UTC)
✓ Done for dioxygen. Snipre (talk) 16:27, 1 December 2014 (UTC)
✓ Done for dihydrogen. Snipre (talk) 17:00, 1 December 2014 (UTC)
I don't know if dichlorine, difluorine, dibromine and diodine are used in English. Snipre (talk) 17:02, 1 December 2014 (UTC)
It works, I can see the en link on fr:Diazote. Thanks for sharing the trick, Snipre! PiRK (talk) 17:06, 1 December 2014 (UTC)
✓ Done for dichlorine, difluorine, dibromine and diodine. Snipre (talk) 19:08, 1 December 2014 (UTC)

MediaWiki namespace messages

Is there a way to automatically find wikis which have created a particular page in the MediaWiki namespace? They usually have names fixed by the software so if the pagename is the same then they may be assumed to be the same page. For example, MediaWiki:Histlegend is displayed at the top of page histories. The link is blue due to a feature of the namespace, but the page has not been created at Wikidata so some of the usual tabs are omitted and the MediaWiki default is shown. A manual check shows some of the wikis which have created a customized version to replace the default: en:MediaWiki:Histlegend, fr:MediaWiki:Histlegend, es:MediaWiki:Histlegend, de:MediaWiki:Histlegend. Here is one which doesn't currently exist so that wiki should not be listed in the wanted feature: ko:MediaWiki:Histlegend. I imagine there could either be a fully automatic feature (maybe an external tool) with live updating, or a bot which periodically checks for page creations/deletions and updates a Wikidata item. Special:AllMessages shows thousands of MediaWiki messages which may or may not have been created at some wikis but it's OK if a feature is restricted to a limited number of pagenames selected by editors. MediaWiki:Histlegend would for example be useful to see which tools have been added in different wikis, such as the "External tools" line in en:MediaWiki:Histlegend. PrimeHunter (talk) 18:48, 1 December 2014 (UTC)

Two questions about position held (P39)

I've been updating a number of position held (P39) and am unclear about two things:

Follows vs. Replaces — issue with the suggested qualifiers

Best I can tell, for political offices where new people are voted in, best practice is to use replaces (P1365) and NOT follows (P155). However, in practice, most examples use follows (P155) (including Barack Obama (Q76), the example for this property, meaning that the example on the talk page does not match the item itself). I am guessing much of this usage is because when you start adding qualifiers to position held (P39), it auto-suggests follows (P155). Where do these suggestions come from? Are they manually added, or are they calculated as most common qualifiers per property?

Either way, one of two things needs to happen:

General vs. Specific — best practices

The talk page for position held (P39) says Allowed values: generic office like municipal councillor (Q15113603) or whatever. It then recommends using qualifiers like of (P642) to clarify what office was held. While that works quite nicely in visual view for individual items, I'm concerned that since I've read from others that qualifiers will not be queryable for the foreseeable future, we're not going to be able to query by position very well. We can search for people who have been a United States senator (Q13217683), but not specifically for the electoral district (P768) of Illinois (Q1204). We could find all bishop (Q29182) (thousands of people over time), but not bishops of a certain diocese. Or am I wrong?

Thus sometimes I see a very specific item created like for presidents — President of the United States (Q11696) or president of Germany (Q25223). Then we can easily query for people who have been the President of Germany. This was clearly someone's answer to the query issue, instead of using president (Q30461) of (P642) Germany (Q183). But this method feels like a very slippery slope, since suddenly we'd need to create many many specific items to represent every role of every organization.

So my question is: which is preferable here?

I was surprised to learn that Wikidata is not working towards queryable qualifiers. This limitation seems to be rather a significant one.

Thanks in advance for any insight. Sweet kate (talk) 19:22, 1 December 2014 (UTC)

Regarding the follow/replace qualifiers, the reason why follows (P155) is used in more items is that it was created long before replaces (P1365) was created here. Prior to replaces (P1365), there was no differentiation between "replacement" and "succession". No one cared to replace all uses of follows (P155) in position held (P39) after the creation of the new property. Although I didn't see the point of differentiating the two, the new property was created nonetheless and we should use it for transferals of political offices.
Regarding the specificity of position held (P39) values, I would opt for the second option. I don't see a problem with creating many new items if Wikidata were to be a comprehensive database. —Wylve (talk) 01:18, 2 December 2014 (UTC)
Agree with Wylve on the specificity. Don't care much about the previous/next qualifiers. Start date/end date seem more important. --- Jura 11:36, 2 December 2014 (UTC)

Items without P31

I got a query to have list of articles by number of interwikis without P31. it shows articles which doesn't have instance of (P31) please help to adding instance of (P31)Yamaha5 (talk) 07:24, 2 December 2014 (UTC)

The first one on that list, February (Q109), has "subclass of month" instead. Not sure why though. --- Jura 07:43, 2 December 2014 (UTC)
Yes. we have many cases which are added subclass of (P279) instead of instance of (P31). I added to year monthYamaha5 (talk) 08:28, 2 December 2014 (UTC)
Useless list. Sjoerd de Bruin (talk) 08:33, 2 December 2014 (UTC)
February 2014 (Q15710325) is a instance of month. February (Q109) is a class of all months which are called February: February 2014, February 2013, February 2012, etc.
The list could be useful if all items with a P279 claim are excluded. --Pasleim (talk) 08:44, 2 December 2014 (UTC)
✓ Done List is updated Yamaha5 (talk) 09:59, 2 December 2014 (UTC)

Languages

I apologize if this is the wrong place to ask this, but how do I edit stuff in other languages? I've got the Babel box for Spanish on my userpage, and that was fine for a while, but now there's no edit button next to the Spanish stuff. Supernerd11 (talk) 05:29, 30 November 2014 (UTC)

@Supernerd11: Well, you should be able to edit as it is now, but an alternative is to click the link at the top of the page that says "English" (next to the link for your userpage) and type/select Spanish. Ping me if you have any more questions/problems. :) --AmaryllisGardener talk 02:34, 1 December 2014 (UTC)
Thanks. It's kind of a roundabout way of doing it now, but at least I can edit in other languages again. Supernerd11 (talk) 04:08, 3 December 2014 (UTC)

Dates and precision

Probably it has been discussed but I'm not sure. There are cases where we need to put dates (for example date of birth) as circa 1240, or even c. 1241. In cases like this we cannot put the precision to decade, which is displayed as 1240s and spans 1240-1249, while it would be more accurate as 1235-1245 (or 1239-1241 depending on the sources). Is this something that is going to be managed, or we choose not to put a date at all (since 1240s is not good representation of c.1240)? -geraki talk 12:31, 1 December 2014 (UTC)

Future work. The data structure allows it, but the UI cannot deal with it yet. Actually, some work on how a UI for that should look like would be helpful… --Denny (talk) 17:05, 1 December 2014 (UTC)
see Property talk:P569#More than one date, Help:Modeling/general/time--Oursana (talk) 23:31, 1 December 2014 (UTC)
This is a problem which is quite urgent with the current drive to add items for artists and relate them to other databases -- both to facilitate the d:Wikidata:WikiProject sum of all paintings project, and in preparation for the Structured Data project on Commons. For artists it seems particularly prevalent only to have approximate dates of birth or death; but accurately representing the nature of the approximation is very important for accurate matching.
The sooner we have some agreed standards in place for representing at minimum "circa", "after" and "between", the better. Jheald (talk) 00:49, 2 December 2014 (UTC)
Maybe start a proposal and spec on how it could look like, so that the dev team has an easier time starting work on this? It's a non-trivial problem, so this could be really helpful. --Denny (talk) 17:53, 2 December 2014 (UTC)
When I asked this question elsewhere, I was directed to use sourcing circumstances (P1480). @jheald:, does this meet your needs? - PKM (talk) 21:42, 2 December 2014 (UTC)

Data access for Commons enabled + reminder for statements on properties

Hey folks :)

As announced the week before last we just enabled data access for Commons. This does not yet include access to information from arbitrary items but only the item that is directly connected to the page using the data via a sitelink. Please help with polishing the relevant pages c:Commons:Wikidata and Wikidata:Wikimedia Commons. What will come later: arbitrary access (in January/February) and storing meta data about individual files (part of the ongoing work on the structured data on Commons project)

Tomorrow we will enable statements on properties so you can store mappings to other database vocabularies for example.


Cheers --Lydia Pintscher (WMDE) (talk) 21:11, 2 December 2014 (UTC)

sneak peek time - checking against 3rd party databases

Hey folks :)

The student team working on data quality and trust is hard at work and just showed me a first demo. I wanted to share that with you as well. One part of the team is working on checking Wikidata's data against other databases. Here is a screenshot of their first demo showing checking of our data against MusicBrainz. In the end this will be nicely integrated on Wikidata as part of the constraint violation reports probably. This is already working way better than I expected it ever would. So heads off to the students. I'm sure they'll kick ass over the next months.

Cheers --Lydia Pintscher (WMDE) (talk) 21:29, 2 December 2014 (UTC)

X (Q14949238) verification

Is X (Q14949238) a hoax? Articles claim it is a type of Curculionidae (weevil), but the Wikipedia articles have irrelevant photos, and it was deleted on Wikispecies:X for having no useful content, and the alleged source http://www.catalogueoflife.org/annual-checklist/2011/search/all/key/x/match/1 only allows two letters or more now. I'm asking here only because it's the nearest English-language site where the item still exists. Should I ask this somewhere on Wikispecies instead? --Closeapple (talk) 16:47, 2 December 2014 (UTC)

I can't answer the original question, but I think I can tell that the Wikispecies article probably wasn't related to those Wikipedia articles at all. The deletion log says its content simply was "George Emil Palade", so probably only a vandal's quick deed (compare the other edits by the user of the same name). The articles in cebwiki, svwiki and warwiki are ordinary bot articles for sure not directly triggered by a vandal. (However, they all contained a link to the commons:Category:X, which is for the letter, and a gallery of random pictures from that category. This is bullshit, and I removed it in all three articles.) --YMS (talk) 17:06, 2 December 2014 (UTC)
No, it is not a hoax. It is a typo of above-average magnitude. Lsjbot has made a lot of Wikipedia pages (ceb, sv, war) based on CoL (Catalogue of Life), which is a database that is notorious for not being all that good (Lsj claims it is the best there is, which is true in that it is the only one of this kind, which is the same as saying it is the worst there is). I have long since lost count of the errors I have tracked in this output, and there is no end in sight.
         I vaguely recall having made a start in resolving this particular error, and making some progress, but it was not easy and something else came up. It is not really worth the effort (none of these Lsjbot pages really are). - Brya (talk) 18:45, 2 December 2014 (UTC)
@Byra:: So that bot is adding massive numbers of things that only use a search page (failure) for a specific species instead of the actual species ID number, so we can't verify it? That sounds like would fail Wikipedia:Verifiability (Q79951) on all Wikipedias. --Closeapple (talk) 11:36, 3 December 2014 (UTC)
I found this particular one. http://eol.org/pages/21923369 -> http://www.catalogueoflife.org/annual-checklist/2011/browse/tree/id/2460348 (wouldn't click on that link: its JavaScript is quite slow) -> http://www.catalogueoflife.org/annual-checklist/2011/details/species/id/7810593 (X cinerea, Hust. 1928) -> http://wtaxa.csic.es/search/ValidNameGenHP.aspx?level=S&Id_Name=88252 which says "Confidence level: Secondary source only." but calls it "Mecysmoderes humeralis". So I guess that X (Q14949238) is for a genus of 3 species that researchers think may be, but have not yet confirmed to be, part of Mecysmoderes (Q14923469) (which has dozens of species). --Closeapple (talk) 12:36, 3 December 2014 (UTC)
It's a database artefact. X stands for the new genus Xenysmoderes (Enzo Colonnelli: Notes on the Ceutorhynchinae tribe Mecysmoderini WAGNER, 1938 (Coleoptera, Curculionidae). 1992). --Succu (talk) 13:17, 3 December 2014 (UTC)

Admin needed for property deletion - P71 family

Property:P71 (family) has no links from item namespace [1] Andrea Shan (talk) 12:58, 3 December 2014 (UTC)

AFAICS P70 is the only one left to be replaced with parent taxon:

Andrea Shan (talk) 13:31, 3 December 2014 (UTC)

✓ Deleted Matěj Suchánek (talk) 14:06, 3 December 2014 (UTC)

Intersection properties

I've just spotted Q18542787 (being used with occupation (P106)), which appears to be a combination of "nationality: Iranian" and "occupation: footballer". I'm not sure if we have a formal position on this kind of "intersection property", but I'd have thought we would want to do this by applying both properties separately. Thoughts? Andrew Gray (talk) 19:34, 17 November 2014 (UTC)

(and @علیرضا:, who created this} Andrew Gray (talk) 19:38, 17 November 2014 (UTC)
Well, we do have "American football player"  ;) --- Jura 19:49, 17 November 2014 (UTC)
No nationality though. "American football player" refers to someone that plays American football (Q41323), regardless of nationality. ;) --AmaryllisGardener talk 20:18, 17 November 2014 (UTC)
I'm against such properties and items. Separate the nationality from the occupation. Filceolaire (talk) 01:44, 18 November 2014 (UTC)
+1 --Pasleim (talk) 09:33, 18 November 2014 (UTC)
+1 --Tobias1984 (talk) 10:36, 18 November 2014 (UTC)
+1 --Casper Tinan (talk) 13:00, 18 November 2014 (UTC)
+1 --Haplology (talk) 04:32, 19 November 2014 (UTC)

The above voters are all against certain items that are classes? What generic logic is behind that? If a Wikipedia has a page for such an item that you don't want, will the Wikipedia have to delete the page? Or will they have to manage it without easy interwiki etc? Do you mean to disallow more than one subClassOf statement on an item page? Andrea Shan (talk) 07:08, 19 November 2014 (UTC)

@Andrea Shan: In my opinion classes that are too specific, will become unusable. Think about how we could subclass of (P279) a dog: (1) hairy four-legged mammal (2) hemerophile barking mammal. How would we ever get consensus for such classes across 270 languages? And even a 100 word long label would still be missing things that all dogs have in common. --Tobias1984 (talk) 09:05, 19 November 2014 (UTC)
@Tobias1984: Can you specify "too specific" and "unusable"? As I wrote, what happens if what some users deem to be "too specific" has a page in a Wikimedia site? Andrea Shan (talk) 04:41, 1 December 2014 (UTC)

I don't see how Q18542787 passes the threshold for notability. Andrea's objection is somewhat valid if the item has a sitelink to Wikipedia but this is not the case here. Pichpich (talk) 18:38, 19 November 2014 (UTC)

How did you test for passing the threshold of notability? Does having a sitelink change that? Andrea Shan (talk) 04:41, 1 December 2014 (UTC)

This doesn't appear to be a good way of solving it. Maybe User:علیرضا can respond here? This is probably just a misunderstanding. Also pinging Amir. Multichill (talk) 19:59, 19 November 2014 (UTC)

I send him an e-mail and asked him to participate in this discussion. Amir (talk) 21:53, 19 November 2014 (UTC)
I agree, this would be a nice "Commons category", but occupations and nationalities should be separated here, except when the nationality implies a specific kind of occupation (i.e. if Iranian football was different from what most people (apart from Americans) call "football") - Is it the case here ? — Note : I added a FR label to this item, but wondered about it when I found it… had no time to submit it to Project chat at the time ;) --Hsarrazin (talk) 22:59, 20 November 2014 (UTC)
how about Cachoeira do Sul (Q166378)? In this item instance of (P31) > municipality of Brazil (Q3184121) it has also nationality. there are many the same casesYamaha5 (talk) 07:03, 27 November 2014 (UTC)
This is different. Municipalities of Brazil are not the same as municipalities of Japan or Finland. They have different laws, power, organization structure. For example in Brazil, municipalities are second-level administrative divisions, in Finland they are on the third level. A soccer player, however, is always the same independent of its nationality. --Pasleim (talk) 12:56, 27 November 2014 (UTC)
Also true: Football players of Brazil are not the same as football players of Japan or Finland. Andrea Shan (talk) 04:41, 1 December 2014 (UTC) And not to forget Q11722303 "rural municipality of Sweden or Finland (Q11722303)", which just got something unvisible merged in (07:40, 12 November 2014 Epìdosis (talk | contribs) deleted page Q15051065 (Merged with Q11722303 (merge.js))). Andrea Shan (talk) 08:07, 1 December 2014 (UTC)

@Andrew Gray, Pasleim, Tobias1984, Casper Tinan, Haplology, Pichpich: Here is an example, a subclass of "street dog" is only different by one property, namely location: "street dogs in Moscow (Q3663093)". So, if there is a specific dog Mikhail, that lived in Moscow and qualified as street dog, then, why should someone be prohibited from using instanceOf=Q3663093? Same for Bangkok, Bucharest, Chennai Special:WhatLinksHere/Q5279053. Also, the title of this section is misleading, these are intersection classes, not intersection properties. And that a subclass of another class has some extra property (here nationality or city) is what is classification is all about. Andrea Shan (talk) 08:25, 3 December 2014 (UTC) (pinging User:Emw via the ping-template didn't work, maybe there is a limit on how many users can be pinged.)

The original objection was to an item that was created when zero (0) MWF site links existed. This example, street dog in Moscow (Q3663093), has many links. This is an essential distinction lying at the crux of the matter. If you wish to continue this discussion you should provide a concrete example that corresponds in its particulars to the item that the discussion is about. --Haplology (talk) 12:03, 3 December 2014 (UTC)
Haplology: The original objection referred to "formal position" and the discussion moved on, with four users voting +1 for "I'm against such properties and items. Separate the nationality from the occupation." - that statement made no reference to whether article links exists. That's when I started participating. Now go on, the dogs, what if there is along the lines with "street dog in Moscow/Chennai/Bucharest" an item "street dog in London" but this time without an article link, would that be disallowed? It would fit in the system of the others. Regarding zero WMF site links,
  • the item has "topic's main category" and from there, there are 43 site links.
  • there are many more items that don't have sitelinks, shall they all be deleted if they are an intersection class?
Before going around and randomly deleting, I would suggest a bot maintained list of unwanted "intersection items" is created. Andrea Shan (talk) 22:57, 3 December 2014 (UTC)

"The time allocated for running scripts..." message

I see the message "The time allocated for running scripts has expired" on many pages (Wikidata:List of properties/Terms). Is this a problem with my personal load time, or a more general problem that all users are seeing on longer pages? - PKM (talk) 22:50, 2 December 2014 (UTC)

first answer: There was such a problem in last summer for many or most users on longer pages, (see this) but it seems to be solved, (not completely. see bugzilla:71169) But now for example, Wikidata:List of properties/Terms or Germany (Q183) are loading for me, (Firefox 34.0, but apart from that, no new and fast software and hardware) without problems. --Diwas (talk) 23:34, 2 December 2014 (UTC)--Diwas (talk) 00:53, 4 December 2014 (UTC)

Taxonomy clean up

I am cleaning up Special:WhatLinksHere/Property:P71, only 40 left, so it can soon be deleted.

I found Ephialtias choba (Q5382205) which currently has parent_taxon=Actea and is connected to the following articles having the shown upward classes: 

sv:Actea choba      - Actea                  - Notodontidae                    - Noctuoidea - Lepidoptera
ceb:Actea choba     - Actea                  - Notodontidae                    - Noctuoidea - Lepidoptera
war:Actea choba     - Actea                  - Notodontidae                    - Noctuoidea - Lepidoptera
nl:Actea choba      - Actea      - Dioptinae - Notodontidae                                 - Lepidoptera
en:Ephialtias choba - Ephialtias             - Notodontidae                                 - Lepidoptera
vi:Ephialtias choba - Ephialtias - Dioptinae - Notodontidae - Macrolepidoptera              - Lepidoptera

So it looks as if parent_taxon should have two values or the item would need to be split.

Andrea Shan (talk) 10:42, 3 December 2014 (UTC)

Yes, a common enough situation. There is no really good solution, although it will always help to know the taxonomy. - Brya (talk) 11:38, 3 December 2014 (UTC)

Also, I created a new genus item:

  • OLD Demonax (Q15871577), genus of Sabellidae, worms
  • NEW Demonax (Q18601001), genus in family Cerambycidae, insects

and changed some species that have a name starting with Demonax. The diffs are a disaster, see: [2] and [3] ... Showing only the label "Demonax" and not the Q-number, one can guess that the Q-number changed. Using the mouse and hovering over the labels confirms this. Andrea Shan (talk) 11:14, 3 December 2014 (UTC)

There can not be two generic names Demonax for animals. One of them must be wrong. In this case, there is no such thing as "Demonax genus of worms", so that will have to go. Looks like a significant mess. - Brya (talk) 11:38, 3 December 2014 (UTC)
@Brya: Thank you for clarification. Then one could merge the two and alter the data of the OLD item. It was User:ReimannBot that added Sabellidae [4]. The original item was made by User:SuccuBot based on NL Wikipedia, which currently has Cerambycidae. I will merge the two. Andrea Shan (talk) 13:05, 3 December 2014 (UTC)
@Brya: I reverted my merge. [5] states Sabellida. Andrea Shan (talk) 13:11, 3 December 2014 (UTC)
I removed the wrongly placed ITIS- and WORMS-ID, changed the parent taxon to the correct value und merged the items. --Succu (talk) 13:43, 3 December 2014 (UTC)
But if there are two "Demonax" why merge them? And why remove the reference to WoRMS ID 129526 from Wikidata? Also why did User:Epìdosis delete Q18601001? Now only admins can look into it. Andrea Shan (talk) 14:01, 3 December 2014 (UTC)
The correct mame of the "second genus" is Parasabella (Q3894071). --Succu (talk) 14:14, 3 December 2014 (UTC)
The WoRMS entry indeed indicates that Demonax has in the past been placed in the SABELLIDAE but also that this name may not be used. It has been replaced by Parasabella (this is not to say that Parasabella is the correct name, this is a taxonomic question, and taxonomic opinions may vary).
        This kind of thing tends not to be simple. For instance, matching the species list in WoRMS turns up Demonax brachychona (Q3887327) which is not a beetle. WoRMS doubts that it is a species at all, but nevertheless, somehow, it is in Wikidata, degrading data quality. - Brya (talk) 17:33, 3 December 2014 (UTC)
Wow Brya, I think you found the needle in the haystack. nl:Demonax lists 327 species. But Demonax (Q15871577) links to 328 species. --Succu (talk) 18:24, 3 December 2014 (UTC)
@Brya, Succu: Taking one position is what degrades Wikidata quality. If Wikidata cannot express disputes, than it is able of less than WoRMS. Andrea Shan (talk) 22:40, 3 December 2014 (UTC)
Actually, WoRMS is taking one position only, while Wikidata is not necessarily doing so. It is important to make the distinction between nomenclature and taxonomy. The fact that Demonax may not be used for a genus in the SABELLIDAE is a matter of nomenclature (the rules on names). This is so, independent of any position.
        If there are different positions on taxonomy, Wikidata can include these, with proper taxonomic references added, even if it is not all that easy. It is relatively easy to add different "parent taxa", and it is not impossible to indicate different generic placements (although this leaves much to be desired). - Brya (talk) 06:33, 4 December 2014 (UTC)

statements on properties are live (+deletion propagation)

Hey folks :)

We just enabled statements on properties. You can for example use this to:

  • describe mappings to other projects' vocabularies
  • indicate constraints for the usage of this property (currently stored in templates on the talk page of the property)
  • store information about where this property is used in other projects to make sure they can be notified when major changes are made to how the property is used
  • provide links to extensive documentation about and showcases for the usage of this property
  • store patterns for how an identifier should be expanded to form a proper URL
  • ...

Some notes:

  • You can't yet make statements linking to other properties to for example indicate that property X is the inverse of property Y. We are adding a new datatype for that. It'll come soon. The ticket for that is phabricator:T75302.
  • Constraint violation reports currently don't take these statements into account. They will continue to use the templates on the property's talk page. A team of students is currently working on overhauling the whole constraint system. It'll take them a while still though.

Unrelated but also important for you to know: If a page is deleted on Wikipedia/Commons/... that has a sitelink in an item on Wikidata that link will be removed automatically now.

Cheers --Lydia Pintscher (WMDE) (talk) 21:17, 3 December 2014 (UTC)

User:Lydia Pintscher (WMDE) - That looks great! Andrea Shan (talk) 00:11, 4 December 2014 (UTC)

Chrome bug?

Recently I haven't been able to save Statements when using Google Chrome for Mac OS. (Yosemite)

I can still save using Mozilla Firefox for Mac. --Andrewssi2 (talk) 06:59, 23 November 2014 (UTC)

Is this still an issue? If so can you open the console (F12 on my system - not sure what that translates to on Mac) and see if there are any errors? Can you add ?action=purge to the URL of the item you can't add statements to and see if that helps? Can you edit labels, descriptions, aliases and sitelinks? --Lydia Pintscher (WMDE) (talk) 19:40, 26 November 2014 (UTC)
Hi, sorry for the late reply. I do get a 404 error in the console when I open the property up for editing. The full error stack I have placed here
Using the -Purge flag didn't help.
As far as I can tell, all the edit functionality is like this. The 'save button is always grey and inactive. --Andrewssi2 (talk) 05:32, 1 December 2014 (UTC)
Just a small ping to Lydia Pintscher (WMDE) , thanks! --Andrewssi2 (talk) 15:32, 4 December 2014 (UTC)
Thanks for the ping. Missed this. Hmmm I get the 404 as well but this is unrelated. I'll investigate that as well. As for your issue: Can you go to a random item (best use the link in the sidebar here)? Does it occur there too? As no-one else seems to be experiencing this issue my guess is that somehow you are getting a bad cached version either on your system or on one of the servers. Next thing to try: do the same thing logged out and logged in. --Lydia Pintscher (WMDE) (talk) 15:39, 4 December 2014 (UTC)

The European library puts million of bibliographic metadata in CC0

Hi everyone,

I don't think the topic has been already discussed, but the European Library (which represents a lot of European libraries) has just published several millions of bibliographic metadata in CC0. I think this is a very valuable resource for Wikidata — both as data per se, and as pre-registered sources to reference other data (currently, the process of entering manually bibliographic metadata to get accurate sources is a bit painful).

They have just contacted me on twitter and would love to have part of this huge dataset reused on Wikidata.

Alexander Doria (talk) 13:11, 4 December 2014 (UTC)

ok, all script and bot creators... could you build a one-click import tool for any reference found in it ? with a pretty-pretty-pretty PLEEEAAAAASE ? --Hsarrazin (talk) 13:25, 4 December 2014 (UTC)
Sounds lovely :) ·addshore· talk to me! 15:27, 4 December 2014 (UTC)
@Alexander Doria: Move the topic to Wikidata:WikiProject Books and ask them to match the description system of this database with our properties model. Snipre (talk) 15:45, 4 December 2014 (UTC)

Statistics

I was just looking at the statistics. We passes 50 million statements this month (http://tools.wmflabs.org/wikidata-todo/stats.php), which is incredible. But can anyone explain to me why the graph "statements per item" goes up to 16 million, but Wikidata's front page says we have roughly 13 million items? Do we already have 3 million redirects? --Tobias1984 (talk) 09:06, 25 November 2014 (UTC)

And even better, if the concept would be fixed. The section "In other languages" on an item page states names, incl. aliases, in other languages than the user's interface language. These are statements too, but User:Lydia Pintscher (WMDE) does list them outside the section "Statements". But when looking at Wikidata:UI redesign input it seems Wikidata is caught by featurities - creating new things, and not fixing bugs. Andrea Shan (talk) 06:00, 26 November 2014 (UTC)
@Tobias1984: According to my counting, there are 16,604,166 items and 53,615 redirects
@Andrea Shan: The meaning of a label, description and alias is not user interface dependent and conceptually, there is a large difference between terms and statements. If you don't like the "In other language" section, you can put a single babel-box to your user page. Then only terms in this language will be shown and all others are removed. And please, avoid personal attacks. The relation between the development team and the Wikidata community is quite well. We do not have to change this. --Pasleim (talk) 08:05, 26 November 2014 (UTC)
@Pasleim: And do you maybe know why the statistics on the front page say 12,873,909? How is that number generated? --Tobias1984 (talk) 08:19, 26 November 2014 (UTC)
The frontpage shows what is in Special:Statistics. Higher numbers might be due to any of the following for example: 1) taking the Q number instead of actually counting 2) not taking deletions into account 3) counting redirects differently --Lydia Pintscher (WMDE) (talk) 08:27, 26 November 2014 (UTC)
@Magnus Manske: Sorry to be so nosey about this, but I would really like to know how these numbers are derived. I don't think it is the Q-number Lydia because those are already at 18575104. --Tobias1984 (talk) 10:54, 26 November 2014 (UTC)
Maybe can be useful «will be counted as an article in the statistics if it contains at least one wiki link». --ValterVB (talk) 18:22, 26 November 2014 (UTC)
Hmmm interesting. So you mean the low number wouldn't include items without statements/sitelinks? I had not thought of that but it is possible. I opened phabricator:T76071 to investigate further. --Lydia Pintscher (WMDE) (talk) 19:46, 26 November 2014 (UTC)
Lucie investigated it. Her conclusions are in the ticket. --Lydia Pintscher (WMDE) (talk) 17:24, 27 November 2014 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Lydia Pintscher (WMDE): Wouldn't it be good then to make 2 rows on the Statistics page: (1) Total items; (2) Items (excluding those with 0-statements). I think it would help especially for outside reporters or the press to get the numbers right. --Tobias1984 (talk) 20:48, 27 November 2014 (UTC)

Yeah that makes sense. Or at least have a clear description explaining what it actually means. I have to be honest though it's not super high on the list of things to do at the moment imho. --Lydia Pintscher (WMDE) (talk) 21:52, 28 November 2014 (UTC)

Name statements

@Pasleim: There is no conceptual difference between a name statement in the section "In other languages" and in the section "Statements". See the various name related properties. Maybe you got confused by the interface that WMDE under the reign of Lydia Pintscher (WMDE) created. What does User:Emw think? I would like all name statements be properly handled via Wikidata:Properties, so they are then included into the statement count, making the value number Tobias1984 mentioned above easier understandable. Andrea Shan (talk) 21:17, 28 November 2014 (UTC)

Andrea Shan, I don't know why you are so unhappy with me but can we please keep this is friendly space? Feel free to email me and tell me what happened that you are unhappy with me so we can sort this out. --Lydia Pintscher (WMDE) (talk) 21:49, 28 November 2014 (UTC)
I don't currently have an opinion on this. I do recommend less barbed language toward Lydia, though. Emw (talk) 00:44, 29 November 2014 (UTC)
One of the few people here, that seem to have knowledge of outside ontologies, does not have an opinion on splitting out name statements from other statements. But quick was the user with putting a named label on the language used by another user ("barbed") - and going so far as to "recommend" to that user how to contribute. Andrea Shan (talk) 07:56, 30 November 2014 (UTC)
Lydia Pintscher (WMDE) - I am first of all unhappy with the UI. Little changes could make it much better. But such little things are dismissed by you, under an account labeled with WMDE. See Wikidata:Project_chat/Archive/2014/11#Reordering_Statements_wiki-style_.28non-js.29, you claim "I don't think that would give us any meaningful sort order." - but that is only a personal "believe", no reasoning given. If the current order is random - you didn't reply when asked for the current order but Emw said it is random - then sorting by usage and/or grouping by data type would make the order predictable and it would have a meaning, first placed is the most used. A gave several more feedback at Wikidata:UI redesign input, e.g. search box width, simple change in CSS could already reduce the problems. Several statements of yours seem to indicate that you are the main decision maker for the UI, e.g. "I had in mind to show", "I am definitely not set on having it at the top" [6] , "My goal is to make it less of a backbone." [7]. Back to the original topic of that section, I created one more feedback section at the UI redesign input page: Wikidata:UI redesign input#Name statements. Andrea Shan (talk) 07:56, 30 November 2014 (UTC)
As the product manager I do indeed have the last say, yes. I think I am pretty good with giving reasons for my decisions. These decisions are hard but they need to be made to ensure Wikidata is on the right track long-term and not just fulfilling short-term needs. And I did give you reasons for the particular case you mention: It will not lead to a sort order that is useful for the user because it doesn't group properties that should be grouped and it gives too much weight to properties that should not have a lot of weight like identifiers. Others have agreed to this. I have also said that the current system isn't great. But we need to find a better way than simply ordering by number of occurrences and I and others named possible alternatives. It being simple to implement (which it isn't btw) is not the only factor we can use to make such an important decision. I am here to help us all make sure this continues to be a striving project and we are moving forward in the right direction on the technology side. This project is close to my heart like no other. We're not fighting against each other here. --Lydia Pintscher (WMDE) (talk) 10:27, 1 December 2014 (UTC)
Lydia Pintscher (WMDE) - "It will not lead to a sort order that is useful for the user" - reiterating this without giving evidence is the kind of ignorance that is discouraging. User:Emw was the only that gave an example, namely that birth_date and birth_place might be listed apart. But why is this so? Because some people thought it is clever to have features of one event distributed between different properties, and then additionally have dozens of date_of_<anyevent> properties (date of discovery/dissolution/birth/death/launch...). If this is fixed and a more generic event-property is used instead this problem goes away. You re-mention the IDs - by grouping they would at least not be listed mixed with properties of datatype=item. You say it is difficult to implement. The datatype for each property is already in the DB, why can the properties not be sorted by that? Why should that be difficult? I don't believe it is. All the IDs which are probably all "Monolingual text" - "string is defined once and reused across all languages" - would be grouped. Without any change in the DB the default output of Wikibase would be much better. Andrea Shan (talk) 13:16, 4 December 2014 (UTC)
Date of birth and date of death are one example. The next example where this will not produce useful order is family relations. Identifiers will not be grouped together because their usage varies widely. As I said: Yes we can and will improve this but not by grouping by number of occurrences. --Lydia Pintscher (WMDE) (talk) 15:29, 4 December 2014 (UTC)
Lydia Pintscher (WMDE) - You don't even get the basics. I proposed grouping by data type. And yes, then the IDs will be all in one group. Show me a counter example. @family relations - if there is one property "family relation" then they all will be together. So, dates and family relations are solved. What next reason will you make up to not do this simply change. Also, you didn't reply why something like "ORDER BY DATATYPE" would not be classified as being simple to implement? Andrea Shan (talk) 03:53, 5 December 2014 (UTC)

Editing sections on this page

Every time I click the link to edit a specific section of this page, I am presented a different section to edit. I am using the desktop version on an iPad. Is anyone else seeing this problem? - PKM (talk) 19:36, 4 December 2014 (UTC)

Wikidata:Relation between properties in RDF and in Wikidata

FYI Wikidata:Relation between properties in RDF and in Wikidata. Andrea Shan (talk) 04:19, 5 December 2014 (UTC)

Thank you. We should maybe, in the future, move these data to property themselves using statements. equivalent property (P1628) may be helpful for that. Tpt (talk)

P279 - 4000 classes removed in one day

Property:P279 dropped by ~4000 [8] On which items did that happen? Andrea Shan (talk) 03:47, 4 December 2014 (UTC)

Probably the items edited by Special:Contributions/ProteinBoxBot, which is just a guess based on investigating recentchanges using the API with a query like this: https://www.wikidata.org/w/api.php?action=query&list=recentchanges&format=json&rcstart=20141129210000&rcprop=ids|user|comment|sizes&rcshow=bot|patrolled&rclimit=100 --Haplology (talk) 05:22, 4 December 2014 (UTC)
Thanks a lot. Massive removal of subclass of (P279) subClassOf=disease. "neonatal abstinence syndrome (Q18558236)" has ICD-9 and "Disease Ontology ID". The latter in "Relationships" via a tree on "is_a" links up to [9] where it says "is a disease". So this bot made mass removal of valid claims. Could the bot owner User:Andrawaag explain when he plans to put the items back into the P279 tree for disease? Andrea Shan (talk) 06:10, 4 December 2014 (UTC)
The removal of the the >4000 claims were done after a request by Emw on WikiProject_Medicine#subclass_of_disease:_lo-fi_DO_import_problematic. A term being a subclass of a Disease is not explicitly stated in the Disease ontology, which is currently being imported. As such the initial imported claim were invalid and were removed. For details on this topic I refer to the discussion in WikiProject_Medicine Andrawaag (talk) 09:18, 4 December 2014 (UTC)
@Andrawaag, Andrea Shan readded the removed claims. --Succu (talk) 20:38, 4 December 2014 (UTC)
Succu - You made a false claim about me and didn't even notify me. I am only -still- repairing the removal from the disease tree. Did User:Emw really request removal from this tree? On which Wikidata policy this is based? Why did User:Andrawaag carry this out, what basis s/he had to remove this information? To remove false references is a different matter. Andrea Shan (talk) 04:00, 5 December 2014 (UTC)
Andrea, those 4000+ claims were problematic because their provenance was questionable, and they were redundant with a large body of existing subclass of claims. The discussion Andra linked expands on that. Please undo your re-addition of those 4000+ claims. Thanks, Emw (talk) 13:46, 5 December 2014 (UTC)
Emw, I only made additions that were not problematic and not redundant. My initial complaint was about removal from classification. All the items had no subclassOf statement anymore. Take more care with reading next time. Thanks. Andrea Shan (talk) 13:52, 5 December 2014 (UTC)

Research proposal: Wikidata for research

Dear all, I've started to work on an EU research proposal about integrating Wikidata with scholarly databases, which will be drafted via Wikidata:Wikidata for research. The background to the project is explained in this blog post, and your feedback or contributions in any form would be most appreciated. Thanks, --Daniel Mietchen (talk) 14:19, 5 December 2014 (UTC)

Standard Items

I've been searching the Help and FAQ but couldn't find an answer to this question...

Is it possible to create standard items? For example, every planet has the same properties with different values so start a new instance of planet with a framework of properties that can be filled out?

No, we don't have barebones (I think that's the name for it) . The entity suggester (properties that show up when you want to add a new statement) gives good suggestions for things to add to a item. Sjoerd de Bruin (talk) 17:49, 3 December 2014 (UTC)
Has there ever been any proposals to create something like this in a gadget or extension?
Look at Wikidata:Showcase items. Snipre (talk) 08:04, 4 December 2014 (UTC)
I think what's wanted here is the ability to use one item as a template from which to populate the statements for a new item. I would find this very helpful - I frequently want to create a new item that is part of (P361) an existing item (for example, an art gallery within a palazzo, or a named chapel in a cathedral). I'd love to be able to use a gadget to "clone" the current item except for the unique properties (so, copy location, coordinates, country, located-in-place but not commons category, commons institution, or Wikipedia links). It would be very cool if the gadget presented the list of statements that were available to be copied and the editor could check or uncheck which ones to carry over to the new item. - PKM (talk) 19:27, 5 December 2014 (UTC)

Bible instance of

Currently it says: Bible (Q1845) instance of religious text, literary work, biblical canon.

But there are many different translations. So it should be a class, i.e. use "subclass of". Andrea Shan (talk) 00:01, 4 December 2014 (UTC)

 Support Aside from different translations, there are different compilations of chapters that are still called Bibles. - PKM (talk) 19:30, 5 December 2014 (UTC)

retrieved (P813)

Tracked in Phabricator
Task T76859

Is there a way to automatically get the Day date ? it is really painful and may lead to wrong dates added to have to type the date every time, when it is generally "today" (date of the input).

As this qualifier is generally used (manually, at least), just after checking the site or url concerned, a "default" value Today could even be enough.

So, is there a simple way : magic code, control-key, anything ? Thanks --Hsarrazin (talk) 12:29, 5 December 2014 (UTC)

Currently I think there is not. I've opened a bug about it. Tpt (talk) 13:10, 5 December 2014 (UTC)
Oh yes please. - PKM (talk) 19:32, 5 December 2014 (UTC)

Need help with properties

I would like to use my bot to add as much data as it is possible to all memebers of Category:People executed on the Old Town Square 1621 (Q9537378) (see also Old Town Square execution (Q3565105)). To do it the best way, I would like to read some opinions on this.

Matěj Suchánek (talk) 16:34, 4 December 2014 (UTC)

See André Kamperveen (Q2506619) Hope it helps. Jane023 (talk) 23:42, 4 December 2014 (UTC)

User:Matěj Suchánek: Maybe all these properties should be changed. There is an event "death", that has time and place (the most precise, if Old Town Square is in Prague, then this is sufficient). Then there is a parent event "XYZ executions" of which this event was a child event. Then there are participants in the events which have a role shooter/shootee. Andrea Shan (talk) 07:04, 5 December 2014 (UTC)

Updated. Anyone? Matěj Suchánek (talk) 09:11, 6 December 2014 (UTC)

We need "Help:Badges" or "Help:Article status"

At this point of time the article status badge that was highly advertised months ago is now a descriptionless self-redirecting note, and not a particular feature of WD. It would be helpful if someone who understands the badges and how sites can utilise this feature, especially new sites, and non-Wikipedia sites can set up/use the function for their sites. Step by step instructions would be fantastic.  — billinghurst sDrewth 11:00, 6 December 2014 (UTC)

Focus

Since the last changes on the Wikidata site, whenever I press "add property" on a page, I have to click subsequentially on the field to enter the property type (P555 etc). Before the update, that field would automatically get focus, which was way better. Can this be fixed please? I prefer to work without the mouse whereever I can! Hot keys and automatic focus are pretty important to my idea. Thanks! Edoderoo (talk) 19:04, 6 December 2014 (UTC)

It's Phab:T76860. --Stryn (talk) 19:08, 6 December 2014 (UTC)

No Statements for Property:P21

For this Property it isn't possible to add a statement, for all other I seen it's possible? --77.239.38.225 21:27, 6 December 2014 (UTC)

no problem here, added subject item. Michiel1972 (talk) 21:29, 6 December 2014 (UTC)
The page needed purging and the it can only be edited by logged in users. --- Jura 21:32, 6 December 2014 (UTC)

Taxonomy data removal

Q14024729 was labeled via "Property:P71 family" as Formicidae [10]. Since P71 is labeled obsolete, I used subClassOf to express the same [11]. Later I wanted to remove the now redundant family claim.

But then user:Succu stepped in, reverted my edit [12] and removed the family claim [13]. Now "Paraparatrechina caledonica" is no longer placed under "Formicidae".

It seems he destroyed more data. See one of his ~100 "undid revision" [14] edits, which shows that information was lost be his/her edit. Andrea Shan (talk) 09:57, 30 November 2014 (UTC)

I agree that such reverts depletes the structure of Wikidata. However, Succu is not the only one who make such depletions. You do the same thing with ships. /ℇsquilo 10:21, 30 November 2014 (UTC)
ℇsquilo - "You do the same thing with ships." - care to give evidence for what looks like libel to me? Andrea Shan (talk) 10:34, 30 November 2014 (UTC)
Yes, Succu uses the agreed format. There is a widely supported decision that the key property is to be P171 (parent taxon). Subclass of is as redundant as P71. - Brya (talk) 10:18, 30 November 2014 (UTC)
P171 is only to be used for direct upper class, I thought? I not, I could use parent taxon instead of subClassOf. Andrea Shan (talk) 10:34, 30 November 2014 (UTC)
If P171 is present (genus for a species, family for a genus, etc), the rest can be deduced, so everything higher than the parent taxon is superfluous. - Brya (talk) 10:42, 30 November 2014 (UTC)
Brya - care to show how for the examples above that is possible? Q14024729 has parent_taxon Q18579881, which at time of my WiDaR edits and at time of reporting it above looked like this: taxon_name, taxon_rank, instanceOf. How can one deduce the family? Andrea Shan (talk) 10:50, 30 November 2014 (UTC)
Well, coverage of P171 is better than 85%, but that means it is still incomplete. It is possible to include a taxobox by putting "importScript( 'User:FelixReimann/taxobox.js' );" in the common.js; this gives an overview for the completeness of P171 for that page. - Brya (talk) 11:33, 30 November 2014 (UTC)
The problem is that the gadget is broken now... M, @FelixReimann:? --Infovarius (talk) 17:14, 6 December 2014 (UTC)
Infovarius the gadget works, but you have to scroll down now. --Succu (talk) 17:22, 6 December 2014 (UTC)
Well, it still works for me. - Brya (talk) 17:19, 6 December 2014 (UTC)
It is loaded from time to time, not more than in half of cases... On some pages I cannot reach it after several reloadings. --Infovarius (talk) 22:28, 8 December 2014 (UTC)
Lost information: [15], [16], [17]? Simply wrong claims. --Succu (talk) 10:24, 30 November 2014 (UTC)
Your first diff is for an item where family is still present. My claims of data loss were about those were it is not, I gave the diffs above. I spare my time to look at your other "evidence" for "simply wrong claims". Step back from such libel. Andrea Shan (talk) 10:34, 30 November 2014 (UTC)
Spare us all the time, Tamawashi. Add the correct value of parent taxon (P171) - if not present - then remove the family claim (example). --Succu (talk) 10:53, 30 November 2014 (UTC)
Spare us all the time, Dinosaur. Let users improve Wikidata replacing obsolete properties with generic properties if you are not able to fix P71 with your bot. You have no right to prohibit data improvement. Andrea Shan (talk) 10:59, 30 November 2014 (UTC)
Again making things personal, Tamawashi? --Succu (talk) 11:05, 30 November 2014 (UTC)
What do you mean by "personal", Dinosaur? Andrea Shan (talk) 11:14, 30 November 2014 (UTC)
As such things go, this does not count for much. It is not hard to put in a link to the generic name, and Succu makes a lot of such edits almost every day (so percentage wise, it is minute). What would be valuable taxonomic data would be references to solid taxonomic literature (we have embarassingly few of these). - Brya (talk) 14:41, 30 November 2014 (UTC)
Brya - since User:Succu edits taxonomic data s/he is entitled to reduce the data quality temporarily? Nothing was gained by removing the subClass statements. My edits were factually correct. Andrea Shan (talk) 02:27, 1 December 2014 (UTC)
Well, the term "data quality" does not really apply. What we are dealing with here are the basic components of a framework: a taxon item should contain P225, taxon name, P105, taxon rank and P171, parent taxon (and by popular demand P31, "instance of: taxon"). Almost invariably, these are only tentative, that is, they were copied from whatever database some user fastened on (this can be pretty low-grade). The real taxonomic data is not included ("who decided to place this species in this genus, or this genus in this family"). It really is nothing more than pencilling in a draft, hoping there will be real data later on. Deleting "subclass of" does not really remove anything that cannot be replaced with little or no effort. Adding "subclass of" does not add anything relevant, so it cannot really be "factually correct", a meaningless phrase in this context. - Brya (talk) 05:50, 1 December 2014 (UTC)

Family as parent taxon

User:Reinheitsgebot is using the family value to set the parent taxon of a specie [18]. So, can one use any item? I thought only direct parent is allowed. I changed the value [19]. Andrea Shan (talk) 04:01, 1 December 2014 (UTC)

In most cases, the recommended thing to do is to add genus as a parent taxon to a species, family as a parent taxon to a genus, order as a parent taxon to a family, etc. Again, "direct parent" is not a handy phrase, as there may be many taxonomic viewpoints about what exactly is a direct parent, and when going beyond the main ranks (species, genus, family, order, etc), these detailed ranks should be accompanied by a solid taxonomic reference. - Brya (talk) 05:57, 1 December 2014 (UTC)
Magnus bot made a mistake. So what's your point? --Succu (talk) 22:44, 1 December 2014 (UTC)

ISNI - 200 000+ items have wrong value

Spaces are wrong. There are no spaces in the ISNI. It is a 16 character ID:

No other ID maintained by TC 46/SC 9

like ISBN, ISCI, ISSN, ISRC, ISMN, ISWC has spaces, nor does this one have.

Autolist says there are 200043 items for "claim[213]". All that I have seen, except the ones that I added were wrong.

The reality is not a democracy. Reality is represented by applying the rules found in the reality. Does Wikidata want to represent the reality, or does it want to do what appeals to its editors? John B. Sullivan (talk) 12:38, 7 December 2014 (UTC)

Hi John, starting the same discussion in multiple locations is probably not a very good idea. Let's keep it on Property talk:P213#Spaces are wrong. Multichill (talk) 14:34, 7 December 2014 (UTC)
At that page only few people see what is going on. The non-reality has been re-inserted the second time. There is nothing to discuss. This anti-reality restriction in the formatting stops me from adding correct ISNI. Only short time here and already thinking, what is this all about? Is Wikidata something like a blog where anyone can edit as s/he likes? John B. Sullivan (talk) 15:49, 7 December 2014 (UTC)
This property exists since 2013-03-06 20:40. On the talk page, the first edit says "ISNIs are 16 characters, not 20.", this was made 23:49 of the same day. Today is 2014-12-07. This is a joke, isn't it? And Multichill threatening me with blocking, is a joke too, isn't it? If anyone blocks me for reverting vandalism and for telling the truth, then this tells about that person. Here I am. I stand up for the truth. John B. Sullivan (talk) 16:04, 7 December 2014 (UTC)
Yes, you're blocked. This is a wiki that anyone can edit, but if you don't behave you'll find you're edit privileges revoked. Multichill (talk) 16:10, 7 December 2014 (UTC)

Commons links: pages vs. categories

On Wikimedia Commons you can sometimes stumble upon pages which are used to describe the main category it belongs to and show a gallery of images/media from the category and subcategories (example). Most of the time these pages don't use Commons' recommended multilingual description template and as such I've not found the pages very useful and have been linking Wikidata items to categories instead. Categories are also (arguably) the main method of finding and browsing files on Commons, and I believe linking directly to relevant categories would be more beneficial than linking to pages or galleries.

I think it'd be useful to have some sort of guideline at Help:Sitelinks regarding these situations. -- Veikk0.ma (talk) 00:30, 7 December 2014 (UTC)

Nor do I find gallery pages particularly useful. I would prefer a template in most Commons categories that use the property category's main topic (P301) to display a multi-lingual header. /ℇsquilo 08:42, 7 December 2014 (UTC)
I understood that the special Commons properties Commons Institution page (P1612), Commons gallery (P935), Commons category (P373), and Commons Creator page (P1472) supersede links to Commonswiki in the sitelinks section. I do see editors making new sitelinks to Commonswiki, but I don't understand the logic. - PKM (talk) 19:20, 7 December 2014 (UTC)
Sitelinks to Commons are actually sitelinks from Commons linking to the corresponding categories on Wikipedia. The properties are, well, properties. They have other uses and do not supersede anything. /ℇsquilo 09:19, 8 December 2014 (UTC)
@Veikk0.ma, Esquilo, PKM: Rather than a template to go on categories, TheDJ (talk • contribs • logs) has created a prototype for a rather nice gadget, which you can find at c:User:TheDJ/wdcat.js, that uses WDQ to look up to see whether any items have a Commons category (P373) that points to the category in question. If there are, it displays a link from the category to Reasonator on the lower right hand side of the category introduction. This is in many ways a better solution, because 700,000 Wikidata items have Commons category (P373) set, whereas only 260,000 Commons categories have links to Wikidata category items, and of these only 87,000 have category's main topic (P301) set.
To give the gadget a try, go to Commons and add importScript('User:TheDJ/wdcat.js'); to your common.js file.
At the moment it's still only a prototype, so at the moment it's only in one language; it's a little slow to add itself to the page while it waits for WDQ; and it might put too much load on the server if rolled out to everybody by default in its current form. But it does show the sort of thing that should be possible -- without having to change Category pages at all, or to hard-code interwiki links. So give it a try, and see what you think.
As regards Commons sitelinks more generally, it is useful for a category like c:Category:Anonymous_masters or c:Category:15th-century painters from Italy to be site-linked to the corresponding category-like item on Wikidata, that in turn sitelinks it to the corresponding categories on appropriate Wikipedias.
Because we have the limitation that each item on Wikidata can only be site-linked to a single page on each wiki, the view has been taken that Commons categories should only link to category-objects, in order to have a structure that is systematic, predictable, and traversable. Site-links should not be added from Commons categories to article-like Wikidata items, and there are bots operating that remove such links. Instead, the most important thing is to have Commons category (P373) set on the relevant article-like item, to make the Commons category discoverable. This is where effort at the moment would be most valuable -- identifying the article-like items that have direct 1:1 correspondences to Commons categories, and adding Commons category (P373) to them, so the link can be found in WDQ searches, and can be presented by interfaces by Reasonator and Krotos.
At the moment there is little point in creating category-like items for Commons categories unless there are corresponding Wikipedia categories that should appear in the Commons sidebar. Yes, in future I am sure we will want to store structured information about Commons categories, but that will probably be stored systematically in category-like items on the new Commons wikibase, rather than haphazardly on category-like items here. So unless there are corresponding categories on other Wikipedia projects that should be sidebar links, making new category items here for categories that are only on Commons would appear to serve little point, and is probably just unnecessary makework.
Esquilo is right that sitelinks are not properties. But I think PKM is actually closer to where we are now. Current effort ought to be going into building up Commons Institution page (P1612), Commons gallery (P935), Commons category (P373), and Commons Creator page (P1472), and making sure the coverage for those properties is as comprehensive and complete as possible. Every Commons object that would be target of one of those properties should have a corresponding article-like item on Wikidata from which it is discoverable in that way. But there is still a lot of work needed to achieve that. The links going the other way can be presented in Creator or Institution templates, or automatically through gadgets like TheDJ's wdcat.js. But itelinks should not be added from Commons categories to article-like items, because they are not consistent with the scheme that has been adopted, and they will get removed. Jheald (talk) 10:10, 8 December 2014 (UTC)
@Jheald: thank you so much for the detailed explanation. - PKM (talk) 17:11, 8 December 2014 (UTC)

Wikidata - unsystematic unpredictable untraversable duplicative

Jheald, PKM:

  1. " the view has been taken that Commons categories should only link to category-objects, in order to have a structure that is systematic, predictable, and traversable."
  2. ...
    1. " the most important thing is to have Commons category (P373) set on the relevant article-like item "
    2. " At the moment there is little point in creating category-like items for Commons categories "

Commons category (P373) is unsystematic, unpredictable, untraversable, duplicative if there is category's main topic (P301) and topic's main category (P910). The most important thing is to delete if the goal is to have a serious database. But maybe the goal is something else? Perfektinski (talk) 10:13, 9 December 2014 (UTC)

Special:Contributions/Muhib mansour

Sometimes User:Muhib mansour removes links (by accident?) without any reason. See Q7419968 or Q7087591 for example. --Yoda1893 (talk) 09:40, 9 December 2014 (UTC)

Open Data Institute open heritage and culture datasets

Hi All

I stumbled upon a huge list of open (and maybe not so open) datasets the ODI has produced as part of a competition they are running. Thought it may be of interest.

https://docs.google.com/document/d/1hgDvgOwyhCPNtIjCf4rUUcAoFlMlWQXblfq_NcV4_6E/edit?usp=sharing

The first one that jumps out to me is No. 28 - British Museum Semantic Web Collection Online

Thanks

--Mrjohncummings (talk) 12:14, 9 December 2014 (UTC)

Experimental Wikidata feature coming soon to English Wikipedia Mobile site

Hey folks! In the next week or two we will be rolling out a new experimental feature on the English Wikipedia Mobile site that is related to Wikidata. The feature is very similar to Magnus's Wikidata Game, but will be accessible directly from Wikipedia articles and won't require a separate OAuth verification step. The feature is called WikiGrok, and you can read more about it on MediaWiki.org. During the testing phase, we will not be sending the response data to Wikidata, but will be collecting and analyzing the data to get a sense of the data quality level and what factors affect the quality (See #Pushing responses to Wikidata). We will not be sending any actual data to Wikidata until January at the earliest. Once we do start pushing data to Wikidata, it will include a WikiGrok edit tag so that any bugs can be properly upstreamed.

WikiGrok will operate based on a configurable set of 'campaign' definitions. These are similar to the game categories in Magnus's game, e.g. gender, occupation, nationality, etc. We want to run our campaign ideas by you guys and make sure that they make sense and are going to be adding useful data. Until Wikidata has support for simple and complex queries, these campaigns are going to be quite basic. For now, they will only be based on existing property claims in articles. Also, for the time being, we will only be adding claims, not removing or replacing any.

The three campaigns that we will initially be testing with are:

  • Actor campaign: Item is 'instance of human' and 'occupation actor'. Suggest 'occupation television actor' and 'occupation film actor'.
  • Album campaign: Item is 'instance of album'. Suggest 'instance of studio album' and 'instance of live album'.
  • Author campaign: Item is 'instance of human' and 'occupation writer'. Suggest 'occupation author'.

We are also interested in working with the Wikidata community to establish best practices for adding claims and proper ontology for claim properties and values. Please let us know what questions, concerns, or suggestions you have. Kaldari (talk) 21:59, 5 December 2014 (UTC)

Are the suggestions in your examples intended to supplement the existing values or to replace them? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:27, 5 December 2014 (UTC)
Supplement. If that seems like a bad idea, let me know. I know the writer/author pair seems redundant, but they are apparently considered different things in Wikidata. Kaldari (talk) 22:33, 5 December 2014 (UTC)
Will new gamers be notified about his actions reverts? — Ivan A. Krestinin (talk) 22:41, 5 December 2014 (UTC)
Yes, if a person's WikiGrok response is pushed to Wikidata and it is reverted, they will receive an Echo notification on Wikidata (and an email depending on their notification prefs). Since we don't have cross-wiki notifications yet, unfortunately, there is no way to notify them on their home wiki, though. Kaldari (talk) 19:13, 8 December 2014 (UTC)
For the Actor campaign, will the value "both" be allowed? It's certainly the correct answer in many cases (Benedict Cumberbatch, Viola Davis, even Chrisopher Walken...) Since the person in question might also be a 'stage actor' or 'voice actor', we should probably keep the 'actor' occupation as well. (Or 'child actor', which should be time-bound in most cases.) [I love the WikiGrok idea, by the way.] - PKM (talk) 18:27, 7 December 2014 (UTC)
@PKM: We will keep the 'actor' occupation as you suggest (since the more specific actor claims might not cover the full range of a person's acting career). In WikiGrok version B, we will allow people to choose more than one value at once. For example, they could choose both TV actor and film actor. In WikiGrok version A, it only allows verifying one value at a time. So a person would be randomly asked either 'Was X a television actor?' or 'Was X a film actor?'. During our testing phase, we will be testing both versions and see which one gives better results. Kaldari (talk) 19:15, 9 December 2014 (UTC)

Some questions about properties

Hi, I have not worked very much with properties, so I have some noob questions:

Question 1

Do we add properties like in Kristiina Mäki (Q17275699) where all of these are under the P31:

instance of (P31): human (Q5)
sex or gender (P21): female (Q6581072)
country of citizenship (P27): Czech Republic (Q213)

Or should we add those separately, like this:

instance of (P31): human (Q5)
sex or gender (P21): female (Q6581072)
country of citizenship (P27): Czech Republic (Q213)

Question 2

Maciej Staręga (Q11765649) has occupation (P106): athlete (Q2066131) and cross-country skier (Q13382608). Should we add only the "sub-occupation" (cross-country skier (Q13382608)) and not the main one (athlete (Q2066131)) or do we have to add both of them?

Thanks! --Stryn (talk) 18:18, 8 December 2014 (UTC)

Question 1: The latter case is correct, since sex or gender (P21) and country of citizenship (P27) are attributes pertaining to the item itself, not the statement that Kristiina Mäki (Q17275699) is a human (Q5). As a general rule, instance of (P31) shouldn't have any qualifiers. Question 2: only the sub-occupation should be included, but do check whether the sub-occupation is a subclass of (P279) the main occupation. —Wylve (talk) 18:25, 8 December 2014 (UTC)
I agree with Wylve on Question 1. Some tools have difficulties accessing qualifications so Option 2 is better for that reason too.
For Question 2 there is another option. Use occupation (P106): athlete (Q2066131) and 'Sport:Cross Country skiering'. This way there is no need to create an additional item with no links like 'cross country skier'.Filceolaire (talk) 20:33, 9 December 2014 (UTC)
Thanks for your answers! So, my question #1 is now clear. But for the 2nd, this is a mess.... there is also a bit similar discussion at User_talk:Stryn#Addition_of_P641_to_players.27_entries. In my opinion all of these should be handled similarly, not depending of the things whether there are sitelinks available or not. There is sitelinks for tennis player (Q10833314) so either we will stop using occupation items at all or we will continue using them. --Stryn (talk) 20:49, 9 December 2014 (UTC)

Importing of Finnish National Gallery's art collections

Hello, we've had a discussion with WMFI about getting some data from external sites to Wikidata. I'm wondering whether it's ok to import Finnish National Gallery's art collections that contains over 36 000 works (painting, graphics, drawing...)? Licences are Creative Commons CC0 1.0 Universal (CC0 1.0) so there shouldn't be any problem. So, the actual question is if they are notable on Wikidata to have an item. Some of those are notable anyway, since there is a sitelink on Wikipedia, for example Q18346876. We will have someone (or many) who can import those data's after we've learned more about bots. Also, I would like to ask that how we can make sure that we don't import data that already exist on Wikidata? Is there any easy way to do this? --Stryn (talk) 20:17, 9 December 2014 (UTC)

What kind of data are available? --ValterVB (talk) 20:50, 9 December 2014 (UTC)
For example artwork's URL at the Finnish National Gallery's site, artwork's description, type of artwork, artwork's identifier, creator of the artwork. There's a lot of data. --Stryn (talk) 20:58, 9 December 2014 (UTC)
@Stryn: Talk to Wikidata:WikiProject sum of all paintings. User:Multichill has developed various tools and scripts that maybe of help. Oil paintings would certainly qualify for notability, as would sculptures; watercolours, sketches, drawings, engravings etc might or might not -- things are still developing. You should also be thinking about creators, and trying to match them to our existing items. Jheald (talk) 22:47, 9 December 2014 (UTC)

Mystery deletions

Two items on my watchist, Varinder Aggarwal (Q18574853) and Red link example (Q18574815) have just been deleted , each with the edit summary "Does not meet the notability policy". I have no record of what these were, nor any prior notification. Can someone tell me, please? And, in future, can we make sure such deletions have their (English of other) label are noted in the deletion log? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:48, 3 December 2014 (UTC)

Hm, the first one slipped trough the bad ones. Restored that. The second one is about a "red link", not clearly what's the point about that. Sjoerd de Bruin (talk) 22:56, 3 December 2014 (UTC)
Please restore it; it's an example red link on Wikipedia. Why was this deleted with no discussion? How many other inappropriate deletions have "slipped through"? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:03, 3 December 2014 (UTC)
What is a "example red link on Wikipedia"? I only see instance of (P31) pointing at a disambiguation page. I'm busy with the notability-backlog at the moment, there are a lot of stuff that doesn't belong here (advertising, not notable people, items with only a label and/or a description). Sjoerd de Bruin (talk) 23:06, 3 December 2014 (UTC)
You have not answered my questions. Your deletion of Red link example (Q18574815), made with no prior notification, has been challenged: please now restore it. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:09, 3 December 2014 (UTC)
I have reviewed the deletion and see it is within the policy regarding notability of items. The item should not be restored until a need in line with the notability policy is proven. John F. Lewis (talk) 23:31, 3 December 2014 (UTC)
An undisclosed, and challenged, deletion should be restored whole the matter is discussed. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:38, 3 December 2014 (UTC)
A challenged deletion is usually sent for another administrator to review, which it has been. Wikidata is special in the fact, all an undeletion will do is add an unnecessary item to the list again when it does not meet our policies. You are more than welcome to bring this matter up on the administrators noticeboard however the decision is likely to be overturned. John F. Lewis (talk) 23:43, 3 December 2014 (UTC)

Is the "notability-backlog" referred to above User:Pasleim/notability? I have just removed two more of my recent creations from there. How is stuff added to it?

Note that not all of my created items are on my watchlist, as they are created in batches using an approved bot account and 'Quick Statement'. I shudder to think that you or anyone else might be deleting them with no warning. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:12, 3 December 2014 (UTC)

Please don't remove your own pages out of that list. It simply lists pages that don't meet the notability policy. The two pages you removed out of the list are clearly not notable, according to the policy. Sjoerd de Bruin (talk) 23:16, 3 December 2014 (UTC)
The page carries a hatnote saying "Feel free to remove false positives from the list". Your claim that Applied Catalysis Award (Q18560122) and Applied Inorganic Chemistry Award (Q18560129) do not meet the notability policy is bogus. What the hell is going on? I again note that you have ignored my question. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:22, 3 December 2014 (UTC)
Sitelinks? No. References? Don't see any on the item. Structural need? Both are not in use at the moment. Sjoerd de Bruin (talk) 23:28, 3 December 2014 (UTC)
You are still ignoring my questions. You have still not restored your challenged deletion. I note you have Journalisted ID (Q18575559), another false positive on the same list. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:32, 3 December 2014 (UTC)
It's listed because there are not incoming links from other items and there are no sitelinks. Sjoerd de Bruin (talk) 23:33, 3 December 2014 (UTC)
I'm on the boundary of deleting it per 'not an item but intended property' John F. Lewis (talk) 23:35, 3 December 2014 (UTC)
You both appear to be ignorant of the second of the three numbered criteria in the notability policy. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:38, 3 December 2014 (UTC)
I don't believe we are. I fail to see the items being described with serious and notable references. I have also gone a head and deleted the item that should be a property pending its creation. John F. Lewis (talk) 23:43, 3 December 2014 (UTC)
And still you ignore my questions. Will you answer them, or not? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:38, 3 December 2014 (UTC)

Shocking. Users create items and they are silently deleted by other users that by some historical coincidence got the label "admin". All the deleters do is write "notability not met", "should be a property" or something alike. No way for non-admins to verify that. Andrea Shan (talk) 00:04, 4 December 2014 (UTC)

This applies to all wikis. If an admin on enwiki deletes a page; there is no way to verify it was correct. That is why avenues exists of having administrators co-review work which happened here. Administrators are elected by the community to carry out the roles. If you feel an administrator is acting inappropriately, avenues exist which are the administrators Noticeboard, requests for comments and requests for removal of permissions. John F. Lewis (talk) 00:12, 4 December 2014 (UTC)
Admins also have a duty to explain their decisions when asked, and to answer good faith questions about items they have deleted when asked. This is not being done here. Thryduulf (talk: local | en.wp | en.wikt) 01:02, 4 December 2014 (UTC)
Well, I'm explaining them here but that's not enough it seems. Sjoerd de Bruin (talk) 08:20, 4 December 2014 (UTC)
Are you? Please see #Questions, below, for a list of questions you have ignored. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:53, 4 December 2014 (UTC)
 Comment User:John F. Lewis That does not apply to all wikis. Not even sure it applies to all Wikimedia wikis. But even if the latter would be true, is that the defense? A certain behavior somewhere else is a defense for doing the same in Wikidata? Have you seen the user decline in the English Wikipedia? The discussion in many internet forums about deletions carried out by admins? Do you want to repeat that here? If I "feel an administrator is acting inappropriately" I shall go where all the admins are sitting together, to the administrators noticeboard? I am happy User:Pigsonthewing came here, I don't watch the "administrators Noticeboard". "Administrators are elected by the community" - I didn't elect any. Andrea Shan (talk) 03:00, 4 December 2014 (UTC)

While there are certainly items that are better off put down, that list seems to be rather broad, and items put down unnecessarily. For example, the list contains this (my first, randomly picked example), where someone clearly put effort into creating it; as a topic, it would certainly stand on its own as a Wikipedia article. Airplanes are notable. So is Varinder Aggarwal (Q18574853), a Fellow of the Royal Society. On Wikipedia, he would have at least one incoming link from some list, but here the arrangement is "reverse". Finally, while Wikidata is growing fast, it is still in a early growth phase; exterminating seeds because they have not yet become flowers will lead to a desert, not a garden. --Magnus Manske (talk) 08:20, 4 December 2014 (UTC)

Questions

@Sjoerddebruin: Questions you have ignored include:

You also ignored my implied question about why I should not remove pages [sic] from that list, given the "Feel free to remove false positives from the list" hatnote. Perhaps you would answer these five questions now? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:53, 4 December 2014 (UTC)

I didn't get a ping, I think because your signature is too complicated for the Echo-extension. Anyway, the answers:
  1. Because it didn't meet our notability policy. There are no clear rules about the deletion of such itemxs, this is the first complaint in months about it. The amount of deletions and the size of the backlog is too high to start a discussion for every single item.
  2. I don't know, "inappropriate" is pov. Other sysops can check my deletion log.
  3. Yes, you've could look at the history of that page to get this answer.
  4. With a bot every day, ask Pasleim for more information.
  5. Removing your own pages is like marking your own edits as patrolled. You can't decide if your own items are notable, with actions like this you're damaging the quality of our data.
Sjoerd de Bruin (talk) 20:22, 4 December 2014 (UTC)

[Outdent]

Pings work well using my identical signature on other wikis.

  1. The number of items where you perceive a problem is not an excuse for acting without discussion; nor for ignoring the matter of gaining community consensus. That there have been no previous complaints is not reassuring, if the deletions are not discussed, and so not seen by the wider community. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:27, 4 December 2014 (UTC)
  2. "Inappropriate" is not PoV; by your own admission, you said the item "slipped trough". Deleting things which "slip through" is not apprropriate.
  3. The history of that page does not answer my question. But now that you have, an unadvertised page in user space is not a satisfactory way for such a system to operate. Greater transparency is needed.
  4. Inadequate. Again, greater transparency is needed.
  5. We have "autopatrolled", so your analogy is flawed. The invitation on top of the page in question is unambiguous, with no such caveat. The damage to the quality of data is by undiscussed and inappropriate deletions such as those under discussion; not the good-faith contributions of established editors in good standing.

Furthermore, form others' comments in this discussion, it is clear that community consensus for your actions is not demonstrated, nor has it been for the flawed interpretation of polices with which you attempt to justify them. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:22, 4 December 2014 (UTC)

Things slipping through is a matter of life. I can't be 100 percent perfect! But at least I expect that "established editors in good standing" know what notability is and how they can make a item more notable. Autopatrolled doesn't make all of your items automatically notable. In the last few days I've deleted items about non-notable YouTube users, self-claimed notable Indian people, empty items and a lot of vandalism. Nobody seems to see that. Sjoerd de Bruin (talk) 22:16, 4 December 2014 (UTC)
I do know what notability is; you may disagree, but your admin bit doesn't make you (nor a pair of you collectively) more right than me, nor give you additional authority to make that call, which is why if one of your undiscussed deletions is challenged, you should undelete. Since I have not claimed that "Autopatrolled makes all of my items automatically notable" (I was merely refuting your analogy in another regard), your comment on that it is a straw man; your list of other deletions is an irrelevant distraction. At least we seem to have established both that false positives "slip through" with your editing; and that you do not know how many do so. That's again troubling. That when issues are highlighted to you in good faith, you do not always reverse your deletions, is doubly so. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:46, 4 December 2014 (UTC)

Notability policy

(Edit conflict with Magnus) Once again I'd like to point out that notability criterion #2 is totally useless as is (cf. Wikidata talk:Notability#Criterion 2). You can either take the first part ("refers to an instance of a clearly identifiable conceptual or material entity") literally, then basically items about everything in the world would be allowed and the deletion of items about someone's best friend or the cheapest car wash of Chicago would be no-gos. Or you could apply some unwritten minimum standards to "serious and publicly available references" and then spend ages for every such item in order to search those references, which can't be given in Wikidata itself in a reasonable way, and afterwards you'll still have discussions like this on whether the references are good enough to make the item notable or not. My proposal: Just remove criterion 2. Does it serve any purpose not sufficiently met by criteria 1 and 3? --YMS (talk) 08:24, 4 December 2014 (UTC)

If you wish to propose changes to the notability policy, I suggest you start a separate discussion (or boldy edit it and see whether or not you are reverted) Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:53, 4 December 2014 (UTC)
It does seem somewhat dubious to delete an item because of its lack of notability, while at the same time ignoring the notability criteria at a whim. Criterion #2 is not split into "parts"; it could esaily be reformulated, without changing its meaning, as:
It refers to an instance of a clearly identifiable conceptual or material entity, which must be notable, in the sense that it can be described using serious and publicly available references.
You can't just delete items because you dislike the sentence structure in the rule that prevents deletion. Please stop further deletions on this basis, and restore the ones you already did. --Magnus Manske (talk) 16:22, 4 December 2014 (UTC)
I rarely deleted anything that could apply to criterion 2. The fact that criterion 2 basically prevents deletions of items at all (as long as "serious and publicly available references" are not properly defined, nor is it known how these references should be depicted in Wikidata) actually is the reason why I never was very active with the deletion requests. For more on these sources, see the discussion I linked. I did not mean to start a new one here, I just wanted to point to this discussion, as I believe this is the root for the problem here and needs to be solved. --YMS (talk) 18:10, 4 December 2014 (UTC)

Break

Well I guess I will chime in here looking at the two currently still deleted items. So, Q18575559 (Journalisted ID). Sure this could be a property, but it is also a clearly identifiable conceptual entity. The claim 'instance of' => 'authority control' also allows us to generate lists of authority controls from the dataset, this is not possible with just a property at this time. (edit: See a section above, this is now possible.. ·addshore· talk to me! 16:06, 4 December 2014 (UTC)) Now that single claim does not represent a structural need according to our current definition, maybe that is something we need to think about..

So now taking a look at the deletion itself, it was deleted with the reason "this should be a property not item", which as far as I can tell is not an agreed reason for deletion? (or am I missing something)? The two points it might link into on WD:DELETION are "Any other common-sense reason for deletion not listed above" and "When discussion on ____ has lead to consensus to delete the item. Any such discussion would run for at least a week, and require ___ support.". Now given this discussion here I am going to go ahead and say that this can not simply be put down to common sense, there are arguments on both sides, thus where is the discussion?

Quickly looking back at WD:N Note that the guidelines below are intentionally left a bit vague. In case of disagreement over the notability of an item, you can launch a request for deletion. The final decision is always up to the community. There is clear disagreement about the deletion of this item hence discussion should happen / have happened prior to its deletion. I am going to go ahead and say that Wikidata:RFD is probably not a suitable venue for these week long discussions and a new venue may be needed? Speedy deletions due to merges, being empty e.t.c and another location for week long deletions with discussion.

Comments? ·addshore· talk to me! 15:54, 4 December 2014 (UTC)

I found its deletion, immediately after I gave it as an example of an item wrongly-listed as "non notable" during a discussion challenging other deletions on that basis, to be a particularly aggressive and pointed action. Regardless of that, it should be restored as a challenged speedy deletion. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:33, 4 December 2014 (UTC)
And has been. Thank you. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:32, 4 December 2014 (UTC)

Taxon deletions

Found via the tool Autolist "claim[70:28319] and noclaim[171] and claim[105:7432]" in between many items named "Brachodes <...>". Probably they also belong to genus Q9662854 (Special:WhatLinksHere/Q9662854): 

Q13443661
This page has been deleted. The deletion and move log for the page are provided below for reference.
09:19, 2 December 2014 Sjoerddebruin (talk | contribs) deleted page Q13443661 (Does not meet the notability policy)

Q13443664
This page has been deleted. The deletion and move log for the page are provided below for reference.
09:21, 2 December 2014 Sjoerddebruin (talk | contribs) deleted page Q13443664 (Does not meet the notability policy)

@Magnus Manske, Succu: What do you think about this? I cannot even look into the items. Deletion with claim "Does not meet the notability policy" should stop all together. Maybe deactivate or something similar. But deletion is evil. User:Lydia Pintscher (WMDE) - see above how upsetting the deletions are. Could there be something like "freeze" or "deactivate"? So one can still look at the content, but item does not show up in normal searches. Andrea Shan (talk) 06:34, 5 December 2014 (UTC)

Such actions are unnecessary. If the item describes a taxon, it is notable, period. The fact that we are even discussing this shows a disturbing lack of mission awareness by the deleting admin. --Magnus Manske (talk) 08:16, 5 December 2014 (UTC)
Both items were merged and thus do clearly no longer meet the notability policy. --Pasleim (talk) 09:14, 5 December 2014 (UTC)
@Magnus Manske: Is a incorrectly named taxon also notable? Sjoerd de Bruin (talk) 11:56, 5 December 2014 (UTC)
The edit summaries cited do not say "duplicate" or "wrongly named", they say "Does not meet the notability policy". This is again about transparency and accountability. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:58, 5 December 2014 (UTC)
Maybe it was the wrong reason... But, if you want more transparency, ask the developers for better automatic summaries. The default ones don't indicate where the item was about... Sjoerd de Bruin (talk) 12:01, 5 December 2014 (UTC)
User:Sjoerddebruin, if the default doesn't match reality, then undelete all items that you deleted with a label that is a false claim. Andrea Shan (talk) 13:58, 5 December 2014 (UTC)
User:Sjoerddebruin: Feel free to improve MediaWiki:Deletereason-dropdown. No need to bug developers for that :-) Multichill (talk) 18:15, 5 December 2014 (UTC)
You, and no one else, are responsible for your edits summaries. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:04, 5 December 2014 (UTC)

Transparency

@Pigsonthewing: You said above "an unadvertised page in user space is not a satisfactory way for such a system to operate. Greater transparency is needed.". Maybe it would help if User:Pasleim could move User:Pasleim/notability to the Wikidata namespace? IMO several further steps for more transparency are needed. Andrea Shan (talk) 06:43, 5 December 2014 (UTC)

Selection criteria?

What are/were the selection criteria being used to make this list? Looking down it, I can see aircraft types, genes (CDK2 for goodness sake), systematic diagnostic classifications, E numbers, significant chemicals -- none of which ought to be appearing on a list like this; and if some have been deleted, like the taxa above, that is simply shocking.

Something we're fundamentally doing on a lot of fronts is trying to establish 1:1 parity with various external datasources, so we can start from the foundation that every object in the external source -- whether it's E-numbers, whether it's genes, whether it's taxa, whatever -- is confirmed as having an item here. Achieving that comprehensiveness is often the basic 'step zero' before you can do anything else, because you have to establish your complete controlled vocabulary as items before you can be confident that you can put into Wikidata any statement referencing it. For people to be removing items that uploaders are depending on to establish completeness in their desired controlled vocabulary is shocking. For those removals to be being made without any notice or notification being made to the uploaders or the projects concerned is even more misjudged. People or projects will think they have established completeness, only to find it has been moth-eaten by deletion out of their view without any reference or warning.

The criteria being used to generate this list appear to be adding very large numbers of false positives that should not be deleted, some of which then have been deleted. At the very least there now needs to be a serious community review of what has been going on to this list, and what has been deleted as a result. Jheald (talk) 09:06, 5 December 2014 (UTC)

As I started to create the list, all items without a sitelink and backlink were added. As the list growth too fast, I selected also all items out with three or more different properties. It is not intendet that all items on the list have to be deleted but they should all be reviewed. --Pasleim (talk) 09:14, 5 December 2014 (UTC)

New list

As I feel discomfort of some users about the notability list, I created a new list which lists potentially non-notable items: User:Pasleim/Review needed. The list should be preciser and more transparent than the old one and every user (not only sysops) can contribute to it. Any feedback is very welcome. Note that at the moment less than 1% of the actual list is shown. --Pasleim (talk) 23:20, 9 December 2014 (UTC)

@Pasleim: I guess we should appreciate what you're trying to do, to keep us clear of creation of vanity items, rubbish items, and other mistakes. But a lot more refinement is needed. After a quick look at some of the things on your list with English-language labels, I saw nothing that screamed to me that it needed to be removed. But I saw a lot that it would make sense to automatically whitelist. If something has a reference in a major database that is linked, it probably has a place here. If something has been created by an established user, it probably has a place here. If something is an example of a geographical feature, or a music album, or a type of vehicle/ship/plane/research institute/railway stations with coordinates, it probably has a place here.
There may be a systematic issue with people creating new items, starting to populate them, but only realising when they try to add a sitelink that in fact we have an item for the thing already -- and them then leaving their new unlinked part-created item cluttering up the database. That may be something we need to think about.
But in most cases, it seems to me that there may be very real costs associated with over-ready deletion -- for example, WDQ searches no longer giving complete returns if items that ought to be on a complete list have been deleted; or creation of a property on another item failing, because the item the property should be referencing has been deleted. The costs of false-positive deletions are significant, we need to be very aware of them.
So how to refine this list? One thing would be to identify what properties (and property-values) the items do include, with a view to whitelisting groups of them, and/or grouping together similar items for consideration together. Another would be to try to do more with who has created the items. What other ideas for refinement can we try to come up with? Jheald (talk) 11:33, 10 December 2014 (UTC)
Another thing would be to send a friendly message people, if a lot of their newly-created items are pinging this filter, and asking whether the creations are part of a strategy or item-creation campaign that the new-item monitors should be aware of (and could then take into account). Jheald (talk) 11:39, 10 December 2014 (UTC)

Talk pages and redirects

Should redirects have talk pages with content? Sjoerd de Bruin (talk) 19:22, 7 December 2014 (UTC)

If it is talk about the redirect, why not? --Diwas (talk) 23:34, 7 December 2014 (UTC)
Talk pages don´t have to exist unless there is a necessity.--Giftzwerg 88 (talk) 12:24, 10 December 2014 (UTC)

Wikidata:WikiProject_Music

I started a SUPER bare bones Wikidata:WikiProject_Music. I based the page on Wikidata:WikiProject_Books and hope to update it more fully soon. My aim is to have a place for those who are helping standardize the way musicians and their creative works are updated, giving examples and instances of the various related properties. FYI if anyone is interested in helping out or otherwise has suggestions. Go team. Sweet kate (talk) 22:23, 8 December 2014 (UTC)

Added a whole bunch of properties and made a logo. Sweet kate (talk) 05:02, 9 December 2014 (UTC)
I _love_ the logo! :) --Lydia Pintscher (WMDE) (talk) 08:58, 9 December 2014 (UTC)
Thanks, Lydia! :) Sweet kate (talk) 18:26, 9 December 2014 (UTC)
Nice, that list of properties is very useful! Thanks :) DSGalaktos (talk) 13:28, 9 December 2014 (UTC)

Holy shit! That logo does not reveal the nature of a book in any way. Perfektinski (talk) 10:14, 9 December 2014 (UTC)

  • Update. Ups, that was supposed the music logo that I looked to. Well, good proof, it does not reveal the nature of music. Holy, shit, put the logo in front of human beings - WITHOUT THE TEXT - and let them guess what it stands for. Perfektinski (talk) 10:18, 9 December 2014 (UTC)
Clearly my sound bar concept doesn't meet your fancy, but at any rate the text is an integral (and carefully designed) part of the mark and will clarify for those who are unclear. Sweet kate (talk) 18:26, 9 December 2014 (UTC)

The logo is great. Good luck with the project! What I'd especially would love to see is to have here data going beyond what MusicBrainz does. I think that's where Wikidata can truly shine. --Denny (talk) 17:01, 9 December 2014 (UTC)

Thanks! I totally agree on making it more than just a clone of MusicBrainz — really utilizing the broader scope of Wikidata to make something else. Sweet kate (talk) 18:26, 9 December 2014 (UTC)
The logo's cool. Love it! Jheald (talk) 18:47, 9 December 2014 (UTC)
Great logo! I'll join, song items need a lot of attention. --AmaryllisGardener talk 17:42, 10 December 2014 (UTC)

Batch editing best practices

I would like to up my game in making batch edits. I understand Autolist as a tool and am understanding queries in general better, but I'm not sure the best way to then make good, up-to-standards edits. I'd like to query items and then add statements to these items. I understand the Autolist way to do this, but it's slow.

Questions:

  • Do I just deal with Autolist's 1-edit-every-10-seconds throttle?
  • Do I do these edits from my own account?
  • Am I supposed to add sources to these statements in the process (like imported from Wikimedia project (P143))? If so, how do I do that in Autolist?

To explain where I'm coming from at the moment, I'd like to batch-edit musician-related items. For instance, all persons in w:Category:American classical cellists can get instrument (P1303) of cello (Q8371).

Some day I will learn how to create a bot, but in the meantime my main goals are to follow proper Wikidata etiquette and not have to go so slowly if possible. Sweet kate (talk) 21:41, 9 December 2014 (UTC)

Quick statements is more quick and more complete. --ValterVB (talk) 22:07, 9 December 2014 (UTC)
Two things:
— Ivan A. Krestinin (talk) 22:10, 9 December 2014 (UTC)
You can request a bot flag for a seperate account to be used with Autolist (this is mostly what I do with AGbot). As long as it's under manual supervision (ie you look over the list first) it should be fine. Andrew Gray (talk) 12:56, 10 December 2014 (UTC)
Good to know. Thanks, everyone! I will request a bot flag for a separate account, explore Autolist and Quick statements, and be sure to source things properly. :) Sweet kate (talk) 16:15, 10 December 2014 (UTC)

A Boy was Born

I wanted to connect the new de:A Boy was Born to en:A Boy Was Born but got an error message: "already an entry". --Gerda Arendt (talk) 10:21, 10 December 2014 (UTC)

Hydro did it allready ([20]). --Succu (talk) 10:24, 10 December 2014 (UTC)
Thank you! --Gerda Arendt (talk) 10:50, 10 December 2014 (UTC)

Broken redirects

Because some sysops still delete instead of redirect: there are some broken redirects. But, how to fix them? Sjoerd de Bruin (talk) 12:20, 2 December 2014 (UTC)

We should have a tool or something that we could edit redirects like Q13927513. Now we can't fix the redirect easily. We can fix it only by reverting, merging again and so on... --Stryn (talk) 17:11, 2 December 2014 (UTC)
(Edit conflict) The only way I see currently for fixing this is to restore them, then merge them again with leaving a redirect this time. But besides this being a rather complicated process (in order to restore them, the merges have to be reverted first) we likely will have double redirects afterwards, which aren't nice, too (which is why there is Special:DoubleRedirects, too). Another option would be to delete the redirects, but this would destroy the whole point of the idea of the redirects. So leave it like it is (with the redirects being quite non-functional, too), until we have (if we ever have) the possibility to edit redirect pages? --YMS (talk) 17:14, 2 December 2014 (UTC)
@‎Lydia Pintscher (WMDE): How do you see this? Matěj Suchánek (talk) 17:58, 2 December 2014 (UTC)
How are these broken redirects created? How would you want to fix them ideally? (Sorry it's late here and my brain is tired...) --Lydia Pintscher (WMDE) (talk) 21:45, 2 December 2014 (UTC)
The main problem is that some users merge items with the merge gadget without creating a redirect or requesting a deletion and therefore leaving an empty item. This is especially a problem when other items or redirects are linking to this empty item. When a sysop is deleting the item the problem just gets visible but it was actually already there before. I have therefore written a small modification of the merge gadget which forces the user either to redirect or request a deletion. If somebody could review it I would be glad. --Pasleim (talk) 23:01, 2 December 2014 (UTC)
@Petr Matas, Pasleim: Interesting, now we have two possible versions od the gadget (comparison between gadget and Petr's version, gadget and Pasleim's version, both versions). Matěj Suchánek (talk) 14:23, 3 December 2014 (UTC)
I wasn't aware of Petr's version. I don't care which approach we take, important is only that no more empty items are created. --Pasleim (talk) 15:08, 3 December 2014 (UTC)
I can just repeat that I think that the redirect should always be created, even if you request its deletion. Otherwise there will be an empty item until it is actually deleted. Petr Matas 18:06, 3 December 2014 (UTC)
 Support I agree plainly... the RFD with "Merged with item X" should systematically be converted into redirections instead of deletions... sometimes, another redirect has already been made from another item, and it breaks the link :/
In fact, the option "suppression" should be simply removed from the Merge gadget.
Only completely blank items (or spam, etc.) should be deleted. --Hsarrazin (talk) 08:28, 4 December 2014 (UTC)
The changes are prepared for review. Petr Matas 17:05, 4 December 2014 (UTC)
Similar discussions are at WD:AN#Double redirects. Maybe duplicate? :P — Revi 04:50, 11 December 2014 (UTC)

One-of-the-types constraint

I am trying to define a constraint for the items which have P1361 (P1361) (they should be instances of one of the given classes). There is no a particular constraint template for such a case, so I have done as advised at {{Constraint:Type}}: created new element Q18613914 as a superclass and used it as a constraint for the property. However, it is pretty clear for me that this class is not notable in any sense and can cause controversy when used as a superclass (e.g. human (Q5) subclass of (P279) Q18613914). --4th-otaku (talk) 03:50, 11 December 2014 (UTC)

Instance of national park

All the items that match claim[31:46169] could be improved by putting a more useful classification, i.e. say what they really are an instance of, US national park, German national park etc. I improved some, but more than 900 are kind of broken. 91.9.120.125 11:56, 10 December 2014 (UTC)

I disagree with what you've been doing so far. The items that you've been renaming are lists, and are more properly labelled and used on WD as lists. You removed the claims that those items were lists: this at the very least is patently wrong.
Besides that, they are not subclasses of parks: the words "parks" etc. in the titles are plural for a reason. Generally there is no meaningful difference between a park in country X and a park in country Y, and in fact the items that you have changed have articles that say "the parks in this country include these ...", and they do not say "the parks in this country are unique and different from parks in other countries for these reasons..." That's what would make them legitimate subclasses, and that is not the case. I also find this type of claim a little fishy, because I doubt Pakistan administers its parks differently from other parts of its territory. --Haplology (talk) 02:18, 11 December 2014 (UTC)
@Haplology: There can be a huge difference with the name «national park» who can go to the exclusive state ownership park in Canada to the protected landscape of UK, and each country have is own legislation for national parks. I also live in a province who name is park «national park», so we have in Quebec national park of Canada (Q1896949) and national park of Quebec (Q3364923). But I agree with you for they shoudn't be administrative division of the country but more a a type of protected areas of this country. --Fralambert (talk) 04:08, 11 December 2014 (UTC)

Textile artists

If I want to indicate that a person is a textile artist (Q10694573) working in tapestry (Q184296), should I qualify textile artist (Q10694573) with genre (P136) = tapestry (Q184296), or something else?

I thought of using material (Q214609), but that is the stuff that the subject is made of, so that doesn't work well here.

Perhaps we should use field of work (P101) = tapestry (Q184296)? - PKM (talk) 22:04, 9 December 2014 (UTC)

In Italian we have the word "arazziere" Google translate it with "tapestry" but I'm not sure if is the correct translation. --ValterVB (talk) 22:19, 9 December 2014 (UTC)
I'd rather use field of work (P101) in a separate statement. -- LaddΩ chat ;) 00:58, 10 December 2014 (UTC)
@ValterVB: I think Google is confused. If I ask English > Italian, I get "tapestry" = "arazzo" (which makes sense; arras is another word for tapestry in English). I would expect "arazziere" to mean "tapestry weaver" or "person who makes tapestry".
@Laddo: The more I think about it, the more I like using field of work (P101). That also works for values like 'Navajo weaving', 'Chilkat weaving'. - PKM (talk) 01:06, 10 December 2014 (UTC)
I like field of work (P101) too, and I agree with Laddo to put it as a separate statement, not a qualifier. Sweet kate (talk) 16:16, 10 December 2014 (UTC)
I made that change on the ones I've been working on. - PKM (talk) 07:31, 11 December 2014 (UTC)

UI ??? language hip-hop

I was at

the label was "Topánky". I clicked [upravit'] to correct the label. Clicking brought me to

Hello Germany! I changed the text in the input field to "topánky", clicked "Bezeichnung festlegen" which means something like SAVE. And the resulting page was

which showed the label "shoe". At

I verified I made the intended change.

So, in this funny process the UI led me from Slovak to German to English.

Some people thought they would like to outperform the user by setting languages based on clever deep data analysis? KISS KISS KISS. Keep it simple STUPID! 91.9.103.243 13:56, 11 December 2014 (UTC)

limitation for adding or removing Wikipedia:Featured articles (Q16465) and Wikipedia:Good articles (Q19765) to items

please set a limitation like (abuse filter) for not confirm users to not add or remove Wikipedia:Featured articles (Q16465) and Wikipedia:Good articles (Q19765) to items. like this. Yamaha5 (talk) 20:17, 9 December 2014 (UTC)

Do you mean that only (auto-)confirmed users could add and remove badges? I would be fine with that, since per my experience, IP's are mostly adding/deleting badges when they shouldn't. Though, I've also seen autoconfirmed users adding false badges, but of course we can't make it as "sysops-only allowed". --Stryn (talk) 20:38, 9 December 2014 (UTC)
Actually, we have Special:AbuseFilter/51 and Special:AbuseFilter/52. What went wrong, Matěj Suchánek?  – The preceding unsigned comment was added by Pasleim (talk • contribs).
I think those filters are just catching the badge additions and removals and not preventing of someone adding of them. --Stryn (talk) 21:03, 9 December 2014 (UTC)
@Matěj Suchánek: In my opinion abuse filter should only allow Auto-patrolled or patrollers or users with more than 10,000 edit for adding or removing badges. now the abuse filter doesn't prevent and it only flags the edits. and some one should check it every time!Yamaha5 (talk) 19:17, 12 December 2014 (UTC)
All sysops are able to manage abuse filters, by the way. Are you all really sure you want to disallow adding/removing badges for IPs, newbies? If yes, then please help me create a warning message. Matěj Suchánek (talk) 19:27, 12 December 2014 (UTC)
@Matěj Suchánek: At local wikis Badges are added or removed by consensus. changing them should be done by trusted users or sysops not any users (like IPs or newbies). the message is here Yamaha5 (talk) 14:01, 13 December 2014 (UTC)
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘
✓ Done Special:AbuseFilter/history/51/diff/prev/667, Special:AbuseFilter/history/52/diff/prev/668. Matěj Suchánek (talk) 14:30, 13 December 2014 (UTC)

Wikidata Hack-a-thon in NYC on December 14th, 2014

Wikimedia New York City is hosting its first WikiData Hack-a-thon/beginners workshop in Dumbo, Brooklyn, New York, on Sunday December 14th from 1-5pm. If you happen to be in the area, it would be great to have you join us! Details can be found at the official event page: Wikipedia:Meetup/NYC/December Wikidata. If you have any suggestions as to Wikidata activities, that would be very helpful as well. Further, if you are a Wikidata expert or would be interested in helping facilitate demos during this event or future NYC Wikidata events, please contact me! (preferably on my Wikipedia talk page). Cheers! OR drohowa (talk) 16:17, 10 December 2014 (UTC)

According to Wikipedia, «The word "hackathon" is a portmanteau of the words "hack" and "marathon"». Thus, I don't see the reason for a dash between 'a' and 'thon'. --Ricordisamoa 08:30, 13 December 2014 (UTC)

Wikidata London Meetup 2

Hi, we are having our second Wikidata London Meetup, it would be great to meet more Wikidatans. Also check the list for organising Wikidata activities in the UK/British Isles. It would be great to see you if you are in London. Fabian Tompsett (WMUK) (talk) 12:17, 12 December 2014 (UTC)

Allow redirect pages to be added?

With the tree species en:Pappea capensis, I cannot add the English page to wikidata "Pappea capensis", as the genus is monotypic, and it clashes with the genus. The consequence is that all the pages about this species do not link up. I thought that in these cases one should allow the redirect page "Pappea capensis" to be added to wikidata. JMK (talk) 14:00, 9 December 2014 (UTC)

en.wiki must decide if they want connect to the genus or if want connect to the species. Wikidata provides both. --ValterVB (talk) 18:35, 9 December 2014 (UTC)
there is a bug for this on Phabricator - T54564. Filceolaire (talk) 20:07, 9 December 2014 (UTC)
If I connect to species, then all the genus pages do not link up. And if I connect to genus, then the species pages do not link up. Sorry ValterVB, but I don't see that as the solution. JMK (talk) 15:31, 11 December 2014 (UTC)
There is no guarantee that this sort of issue can be resolved in any way, using redirects or otherwise. Whether a genus is unispecific is a matter of a taxonomic viewpoint, and it is well established that taxonomic viewpoints are subject to change. Likely the best that could be achieved would be if every Wikipedia would have both a genus and a species page. However, the pragmatic short-term solution would be to put all the iw-links on the species page; this happens a lot. It is not perfect, but we live in an imperfect world. - Brya (talk) 18:43, 12 December 2014 (UTC)
I think that the proposal at Wikidata:Requests for comment/One vs. several sitelink-item correspondence#Use a property will be able to solve it. Petr Matas 20:12, 13 December 2014 (UTC)

Edit request: Gadget-Move.js

I have made a fix for the Move gadget, please apply the changes. At least moving one sitelink at a time works now. Petr Matas 22:21, 12 December 2014 (UTC)

✓ Done, Sjoerd de Bruin (talk) 22:23, 12 December 2014 (UTC)
@Petr Matas: Just remember in the future, the Administrators' noticeboard may be a more appropriate place for edit requests. --AmaryllisGardener talk 00:45, 13 December 2014 (UTC)
And the use of {{Editprotected}}  — billinghurst sDrewth 07:13, 14 December 2014 (UTC)
{{Editprotected}} does not seem to receive much attention, whereas the request made here was fulfilled in 2 minutes. But I think that it is a good idea to use it and, if nothing happens, go to AN. Petr Matas 13:20, 14 December 2014 (UTC)

Unintentional removal of statements in software version identifier (P348)

The problem that users are replacing old versions with new versions, instead of adding a new statement still persists (e.g. https://www.wikidata.org/w/index.php?title=Q1936840&diff=179861989&oldid=179861635). Is there any way to patrol these changes? It would be even nicer if the interface would prompt users that want to replace statements with a warning message. @Okki: Just pinging you, because your in my example. --Tobias1984 (talk) 16:03, 13 December 2014 (UTC)

Someone could write an abuse filter where someone is editing that property type, and return a custom message.  — billinghurst sDrewth 07:12, 14 December 2014 (UTC)
@Tobias1984, billinghurst: I think it is possible to write it in this case. But there is too much to do with AF. Let's start new page Wikidata:Abuse filter and continue there. Matěj Suchánek (talk) 11:28, 14 December 2014 (UTC)

Cycles in parent taxon (P171) tree?

I've noticed the taxonomy tree seems to have cycles in "parent taxon" property, e.g.: Q122422 -> Q728299 -> Q134665 -> Q2254408 -> Q122422. Is it intended or is it some mistake? I thought "parent" relationship should not contain cycles... --Smalyshev (WMF) (talk) 10:34, 14 December 2014 (UTC)

It could be due to preferred or depreciated statements in parent taxon (P171).
Anyways, wikified version of your sample: Squamata (Q122422) -> Lepidosauria (Q728299) -> Lepidosauromorpha (Q134665) -> Sauria (Q2254408) -> Squamata (Q122422). --- Jura 10:40, 14 December 2014 (UTC)
Maybe this revert gives some explanation. And I removed this claim. So the cycle should be broken. --Succu (talk) 11:38, 14 December 2014 (UTC)

Wikidata:Abuse filter

I have just started a new project page about AbuseFilter because the number of requests is rising. Anyone is welcome to improve the project page and to create the request system (maybe similar as on the English Wikipedia). Matěj Suchánek (talk) 11:37, 14 December 2014 (UTC)

The external client site did not provide page information

I'm trying to add n:ru:Беларусь восстановила таможни на границе с Россией to Q18636241, but I get an error message: "The specified article could not be found on the corresponding site. The external client site 'ruwikinews' did not provide page information for page 'Беларусь восстановила таможни на границе с Россией'.". It Is Me Here t / c 13:52, 14 December 2014 (UTC)

@Hoo man:, @aude:: I have seen this twice now for Wikinews. Can you have a look if there is something wrong in the configuration maybe? --Lydia Pintscher (WMDE) (talk) 14:25, 14 December 2014 (UTC)

Optimising the main page for mobile!

The mobile web team at the Wikimedia Foundation are keen to optimise Wikidata for mobile. We've enabled a mobile site which is currently not being redirected to so we can quickly iterate and give a good mobile experience to your projects readers.

The main page is one issue and I've got some fixes I'd like to propose. Copy User:Jdlrobson/minerva.css and then view https://m.wikidata.org/wiki/User:Jdlrobson/MainPage to see my proposed mobile optimised version of your main page.

This fix is simple. Apply https://m.wikidata.org/wiki/Special:MobileDiff/180548244 to WikiData:MainPage/Content/en and friends and add the rule in User:Jdlrobson/minerva.css to MediaWiki:Common.css

I'd also suggest someone edits Wikidata:Main_Page/Contact so that it doesn't use a table as this doesn't render nicely either with the above improvements.

Please let me know if you have any questions. Mobile could prove to be very important to the future of Wikidata and I'm excited to help support it! Jdlrobson (talk) 21:19, 12 December 2014 (UTC)

"Last edited 13 hours ago by Jdlrobson" is nothing for a serious website. Serious would be to give time in UTC, e.g. last edited 2014-12-13 15:30:24". 91.9.116.138 10:09, 13 December 2014 (UTC)
@91.9.116.138: Would you like to open a bug for it on phabricator (http://phabricator.wikimedia.org) so we can discuss this? :) If you haven't an account (and won't create one) let me know it and i will create the task :) Thanks! --Florianschmidtwelzow (talk) 12:40, 15 December 2014 (UTC)
Also this is actually not related to my request... that's just part of MobileFrontend and is what is currently live on English Wikipedia and every other one of our websites. This modification will also help Vector on smaller screens and is really simple so can I get any help adding it? :-) Jdlrobson (talk) 17:26, 15 December 2014 (UTC)

Merge items

Category:Megabats (Q8804807), Category:Megabats (Q7024251). I think these two items has to be merged. What you think and how to do that? --Termininja (talk) 12:01, 15 December 2014 (UTC)

No. The topic is the same, but one is a category by common name, the other by scientific name. Some iw-links need to be moved for consistency. - Brya (talk) 12:07, 15 December 2014 (UTC)

Mnemonic

I have noticed an issue concerning wikidata items Q191062, Q13461032 and Q5169264 (and possibly others?). Obviously I don't understand most of the languages linked to on those pages but going from some of the easier ones to make out, I found that:

Q191062#sitelinks-wikipedia includes:

Q13461032#sitelinks-wikipedia includes:

Q5169264#sitelinks-wikipedia includes:

The english Wikipedia doesn't have an "operation code" mnemonic article but does mention it at en:Mnemonic (disambiguation). Also, some of the articles may combine several meanings (e.g. eo:Memorhelpilo is linked to Q191062 and is categorized under Psychology, Medicine, Equipment, Cognitive Science, Human Sciences, Society, Educational Psychology and Computer Science). benzband (talk) 12:55, 15 December 2014 (UTC)

Don't hesitate to move the site links to items that seem more suitable. --- Jura 17:35, 15 December 2014 (UTC)

Workaround for AuthorityControl gadget

Many users are having troubles with the AuthorityControl gadget and I think that I have found a workaround:

  1. Disable AuthorityControl in your preferences
  2. Add mw.loader.load('ext.gadget.AuthorityControl'); into your common.js

Details: MediaWiki talk:Gadget-AuthorityControl.js#Not working. Petr Matas 04:28, 15 December 2014 (UTC). Update: Petr Matas 15:13, 15 December 2014 (UTC)

Thanks, any idea how to rewrite the gadget using formatter URL (P1630) statements rather than a hardcoded list ? --Zolo (talk) 18:06, 15 December 2014 (UTC)
I'll think about it. Probably it won't be very difficult. Petr Matas 23:08, 15 December 2014 (UTC)
Fomafix has just done it. Petr Matas 00:40, 17 December 2014 (UTC)
Thank you; that seems to work. Any idea why it's necessary? Also, +1 to what Zolo said. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:03, 15 December 2014 (UTC)
It may have something to do with the script loading order and/or dependencies. Does anyone know how to debug the scripts loaded for gadgets enabled in preferences? I can't even find their code in Firefox debugger (see [1] and [2]). Petr Matas 23:08, 15 December 2014 (UTC)
Add ?debug=1 to the URL. --JulesWinnfield-hu (talk) 23:29, 15 December 2014 (UTC)
Yes, that allows me to debug the gadget (thanks), but at the same time it changes the mode of script loading in such a way that the error does not arise. So, I cannot use this to track down the bug. Petr Matas 00:45, 16 December 2014 (UTC)

Make Sandbox more visible

Hey, I did not know we have a Sandbox here too. I think it would be useful to link to it from the top bar, like the enwiki does. Petr Matas 12:11, 16 December 2014 (UTC)

There is one more Wikidata Sandbox (Q4115189), but for different purposes. Paweł Ziemian (talk) 17:05, 16 December 2014 (UTC)
That one and two more sandbox items are referenced from the fixed Sandbox heading. Petr Matas 00:08, 17 December 2014 (UTC)

Native names of organizations

Is there an equivalent of name in native language (P1559) for organizations? - - PKM (talk) 19:42, 14 December 2014 (UTC)

official name (P1448) Sjoerd de Bruin (talk) 19:51, 14 December 2014 (UTC)
Thanks! - PKM (talk) 02:42, 15 December 2014 (UTC)
That is not equivalent. Frank Robertson (talk) 20:45, 14 December 2014 (UTC)
Logically, it's not, I agree. But I can use it to solve my current problem (native-language name of a church or museum when EN wiki translates the name). - -PKM (talk) 21:04, 17 December 2014 (UTC)

Vandalism

Do we have bots scanning for description vandalism like this diff ? Jheald (talk) 14:58, 17 December 2014 (UTC)

No we don't have any. And also, it's impossible to program a bot that knows what is not acceptable description and what is. --Stryn (talk) 15:00, 17 December 2014 (UTC)
I have seen the same sort of thing and I can't help but wonder if these are disgruntled community members who are testing out how vandalism works on Wikidata, because let's face it, it's pretty hard to find these input fields unless you try pretty hard, and the typical vandal is not trying too hard to do something that won't show up somewhere he/she can brag about about to friends. I have been leaving comments like "Please don't vandalize Wikidata" whenever I see these. Jane023 (talk) 15:03, 17 December 2014 (UTC)
Open up any item that doesn't have a label or description in your main language (or English by default), and two big input fields with a "save" link next to them appear right at the top of the page. You don't have to try hard to vandalize those. We have to try hard to detect vandalism in thousands of unpatrolled edits each day, possibly in any of the Idontknowhowmany languages we support. --YMS (talk) 16:16, 17 December 2014 (UTC)
Some vandlism stuff is at User:Pasleim/Vandalism‎. Matěj Suchánek (talk) 16:44, 17 December 2014 (UTC)

Labels and descriptions for Wikinews items

How should we handle items like 2006-07-10: Mundial 2006: Zinedine Zidane nagrodzony Złotą Piłką? Should its English label just be a translation of the article title? Should it be in quotation marks? What should the description be? "WikiNews article" or an actual description of the article? Kaldari (talk) 20:40, 17 December 2014 (UTC)

I think that it don't need to be a direct translation, but something like: "World Cup 2006: Zinedine Zidane won the Golden Ball award". --Stryn (talk) 20:43, 17 December 2014 (UTC)
I guess the label should simply relate to the 'event' that the article is explaining. ·addshore· talk to me! 23:43, 17 December 2014 (UTC)

Freebase - large content donation from Google

Hello! In 2007 Freebase was established as a project like Wikidata. In 2010 that project was acquired by Google. This month Google announced that Freebase was closing, and they would like to make all of its data available for integration into Wikidata. In six months that website is closing completely.

In my opinion, this is a large content donation and much of it would be useful to integrate into Wikidata. A lot of investment went into this, and a lot of benefit could come to Wikidata to capture this donation.

I personally have little ability to understand what Freebase is offering, but just because of the Google brand name, I am interested in doing something to respond to this offer. If anyone has comments or thoughts on this project, please go to Wikidata:WikiProject Freebase and comment, watch the page, or sign on as a participant of the project. I presume that in the coming months there will be discussion about what to take from Freebase and how to keep it in Wikidata. Thanks! Blue Rasberry (talk) 16:34, 18 December 2014 (UTC)

I think Wikidata needs to examine the data carefully before accepting such a "present". Only high quality data with clear sources should be integrated into Wikidata. After all, unreferenced Wikidata content is worthless e.g. for Wikipedia uses. Gestumblindi (talk) 22:02, 18 December 2014 (UTC)
I agree with Gestumblindi. That's why we suggested a curation and reference tool. Your help is very welcome! --Denny (talk) 05:37, 19 December 2014 (UTC)
@Denny: The curation and reference tool is certainly a step into the right direction, but only a first small step. Of course: An URL is still a lot better than nothing, but not exactly a full-fledged reference. At the least, it should IMHO be possible to add some basic info about the source (title of the website, name of the author, maybe even a short quote), and why only support URLs? Many facts are found in printed sources where you can't point to an URL. Often, e.g. Google Books might have digitized a book, but it's not accessible there due to copyright reasons, and for older titles, you don't have it offered as an e-book. Or there are large amounts of newspapers that are still not available in digital form. Also, using a book or journal as the "basic" reference (where you can optionally add an online version) is much more stable and the reference doesn't become as useless as in the case of a dead URL. Gestumblindi (talk) 20:26, 19 December 2014 (UTC)
Gestumblindi, I agree. The reference tool should not be limited to a single URL. Atlasowa pointed me to Citoid, which seems to be a pretty cool tool to get better references. Also, it would indeed be a shame if you couldn't reference books! No, I agree that the reference tool should be flexible enough for this kind of references. Please join on the tool discussion page to make sure these requirements don't get lost. --Denny (talk) 23:51, 19 December 2014 (UTC)

User:Gestumblindi "Only high quality data with clear sources should be integrated into Wikidata." - then 50% or more of all claims should be removed. 80.134.84.12 14:41, 19 December 2014 (UTC)

Or be annotated with references. You can either break something imperfect, or help fixing it. Since a wiki is work in progress, I tend towards the latter. --Denny (talk) 16:01, 19 December 2014 (UTC)

Neglected property proposals

I would like to attract your attention to WD:Property proposal/Sister projects#Similar item. Petr Matas 02:27, 20 December 2014 (UTC)

new code deployed - language fallbacks and property datatype

Hey everyone :)

Wow. What a day... Hard to beat that but let me try anyway. We have just deployed new code. This includes a first version of language fallbacks, a new datatype to link to properties, fixes for a nasty focus issue when adding statements and more performance improvements. Language fallbacks so far only work for the linked items and properties in statements. It does not take into account your babel box. The entity selector does not do fallbacks yet. I'd love to have your feedback feedback on what we have so far to see how we need to further improve language fallbacks.

Cheers --Lydia Pintscher (WMDE) (talk) 01:08, 17 December 2014 (UTC)

That's pretty cool, will it be enabled in Wikipedia too, or do we have to wait for arbitrary access for getting fallback on labels ? --Zolo (talk) 10:05, 17 December 2014 (UTC)
It'll come to Wikipedia too. I wanted to get some feeling first though of how well the system we have works and where we need to tweak it. Also we'll probably need different fallbacks on the Wikipedias. Maybe even let them decide what fallbacks to use? Let's see in a month or two I'd say. --Lydia Pintscher (WMDE) (talk) 10:24, 17 December 2014 (UTC)
Hmm something seems to be broken because I can't edit sitelinks anymore. Jane023 (talk) 13:24, 17 December 2014 (UTC)
We had a problem with that last night which should now be fixed. You might still be hitting a cache. Can you purge the page (add ?action=purge to the URL) and see if that fixes it for you? --Lydia Pintscher (WMDE) (talk) 13:26, 17 December 2014 (UTC)
Sorry not possible - I created 2 new Wikipedia pages and so there was no cache. For the first one I just used the gadget to add another link from another language, and for the second one there was no other language, so I used Special:SetSiteLink. In both cases the wikidata items were pre-existing. Jane023 (talk) 13:32, 17 December 2014 (UTC)
It might still be caching relevant scripts. --Lydia Pintscher (WMDE) (talk) 14:19, 17 December 2014 (UTC)
Just tested with a new Wikidata item and a new Wikipedia article and the link to edit sitelinks still doesn't work for me. But as indicated, using special:setsitelinks works fine. Jane023 (talk) 14:24, 17 December 2014 (UTC)
Try again please after purging. Should be fixed now. --Lydia Pintscher (WMDE) (talk) 10:19, 20 December 2014 (UTC)
Awesome! Would there be an entry for this fallback from the API too? Also, it always feels weird to just copy the label of an entity from one language to an other, especially for names. E.g. afaik, William Shakespeare has the same label in all latin alphabet based languages). Could there be a mention near languages labels field specifying what will be the fallback if let empty? Then, letting empty if the fallback is already the right label could become the official recommendation, sort of a "lazy labelling" policy :D (I'm not up to date on all the discussions, sorry if this was already discussed) Zorglub27 (talk) 16:15, 17 December 2014 (UTC)
The problem is that it might be true for Shakespeare - in Serbian he would be written Šekspir, but they use a cyricllic language so your point is still valid - but how do you recognize the Shakespeare's of the world? I.e. how do you know that a fallback is correct without copy and pasting? I think the language fallback is awesome, but it is a stop gap only. --Denny (talk) 18:45, 17 December 2014 (UTC)
Agreed. There are also use-cases where we really want to have the label even if it is the same as a fall-back. What we'll be playing with though is making it easier to just re-use a fallback and that'd copy it into the proper label for example. --Lydia Pintscher (WMDE) (talk) 22:07, 17 December 2014 (UTC)
@Lydia Pintscher (WMDE): What is the intended use of the property datatype? Can you give an example? Kaldari (talk) 20:44, 17 December 2014 (UTC)
The main use-case is for statements on properties. This way you can for example express that one property is the inverse of another property. --Lydia Pintscher (WMDE) (talk) 22:07, 17 December 2014 (UTC)
@Lydia Pintscher (WMDE): Bug: property value in property datatype is not linked. Please fix it.--GZWDer (talk) 10:31, 19 December 2014 (UTC)
Doh! I created phabricator:T84992 for that. Thanks! --Lydia Pintscher (WMDE) (talk) 13:03, 19 December 2014 (UTC)

Question on inclusion syntax

Does anyone know how to convert a property name to its ID ("father"→P22) ? For use in a WP template, I would like to achieve what parser function {{#property}} does, in order to use existing LUA modules requiring the property id (such as en:Module:PropertyLink). Alternately, simply give me a link to the code of {{#property}}. -- LaddΩ chat ;) 15:00, 21 December 2014 (UTC)

Removing P132

Hey, I removed about 170K P132 (P132) but in lots of cases we have a P132 that is not in instance of (P31) so it's not possible to just remove P132 but since it maybe in subclass of (P279) hierarchy of P31. Is it okay to copy P132 to P31 and remove them and deal with issues later? (for more discussion see User talk:Dexbot#Dexbot Amir (talk) 21:41, 15 December 2014 (UTC)

Why not? It's a clear improvement, namely getting rid of P132 and improve proper resolving via P31-P279. Removal of any redundant claim is an extra step of improvement. But what is to be removed could be a current P31 or a current P132. 91.9.120.70 17:36, 16 December 2014 (UTC)
There are at the moment 27000 P132 claims left. 19000 are about village in Turkey (Q1529096) and 5000 are about civil parish (Q1115575). In this cases, I think we can replace the P31 value by the P132 value. The remaining 3000 claims I prefer to do it by hand as I don't consider it as an improvement if an item has multiple redundant P31 claims. --Pasleim (talk) 06:16, 17 December 2014 (UTC)
Thank you so much Pasleim. I used your hint to remove the P132 and fix P31 now there are in a very low number (and they're lowering as I write this) Amir (talk) 23:40, 17 December 2014 (UTC)
@Ladsgroup: There are also 550 items which have P132 in a qualifier [21]. I think these qualifiers can all be removed. --Pasleim (talk) 12:03, 20 December 2014 (UTC)
@Ladsgroup, Pasleim: better not. Check whether the claim exists on the item page. This was not here 91.9.110.153 00:54, 22 December 2014 (UTC)

You both have bots, why don't you copy the values to P31??? See Q959356 - this is work for bots, not humans. 91.9.110.153 01:07, 22 December 2014 (UTC)

make duplicate

How easy to duplicate an item, change necessary facts before save it as an new item? Thank you, Conny (talk) 17:09, 18 December 2014 (UTC).

I can only think of a hack: Let's say that you want to create a duplicate of item A.
  1. Create a new item B
  2. Merge item A into item B
  3. Delete all sitelinks from item B
  4. Revert the changes of item A
  5. Edit item B
It would also be possible to create a gadget for this purpose. Or does anyone know any better solution? Petr Matas 20:31, 18 December 2014 (UTC)
As a first step, could we have a tool that would export the claims for an item in a format that could be used to populate QuickStatements (Pxxx:Qxxx, etc.). Then we could use the ones we want to populate a new, similar item.- PKM (talk)
Ok, I am developing such script, it's not finished yet, but to get the idea, add importScript('User:Petr Matas/ExportClaims.js'); into your common.js. Hover on the Statements heading of an item and press Ctrl+C. Petr Matas 03:26, 19 December 2014 (UTC)
thanks I'll try that next time I fire up the PC (I do at least half of my editing in my iPad). - PKM (talk) 00:37, 21 December 2014 (UTC)
Doesn't work for me. Copy-paste buffer is not filled. I'll try playing with it later. (Chrome on Windows 7 PC) - PKM (talk) 02:03, 21 December 2014 (UTC)
Let's continue this discussion at User talk:Petr Matas/ExportClaims.js. Petr Matas 22:14, 21 December 2014 (UTC)

Finding ORCID identifiers used in Wikidata & Wikipedia

Apologies to anyone looking for my blog post, Finding ORCID identifiers used in Wikidata & Wikipedia, mentioned in this week's weekly summary, #137, and getting a "403 forbidden" error. My host chose today to kill my website, and their support team have yet to respond. All being well, it should be back online tomorrow. And yes, all the bills have been paid. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:58, 15 December 2014 (UTC)

It's back, now. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:02, 16 December 2014 (UTC)
Such 403 can be avoided by simply posting this little text in Wikidata. So it looks as if you wanted to promote your blog. 91.9.120.70 17:30, 16 December 2014 (UTC)
@91.9.120.70: Please assume good faith and avoid personal attacks. Petr Matas 00:19, 17 December 2014 (UTC)
I avoid assuming good faith if people behave like the one above. That content is strongly related to Wikidata, what reason could there be to not post it here? 91.9.110.153 00:49, 22 December 2014 (UTC)
For example a wish to publish under different license. Quoting from the page editor: "By saving changes, you agree to the Terms of Use, and you irrevocably agree to release your contribution under the CC BY-SA 3.0 License and the GFDL. You agree that a hyperlink or URL is sufficient attribution under the Creative Commons license." Naturally, for the actual reason, you have to ask Andy. Petr Matas 03:09, 22 December 2014 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘I think no one needs to ask Andy Mabbett for posting :

Finding a page based on a property-value pair can be done via

91.9.127.22 12:10, 22 December 2014 (UTC)

And if anyone wants to know more about that, and some caveats and exceptions, or why it's important, they're welcome to read my blog post - which, to reiterate, is Finding ORCID identifiers used in Wikidata & Wikipedia. Thanks for keeping this in the public eye for a whole extra week. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:45, 22 December 2014 (UTC)

Tool for adding references and data to Wikidata

Freebase was launched to be a “Wikipedia for structured data”, because in 2007 there was no such project. But now we do have Wikidata, and Wikidata and its community is developing very fast. Today, the goals of Freebase might be better served by supporting Wikidata.

Freebase has seen a huge amount of effort go into it since it went public in 2007. It makes a lot of sense to make the results of this work available to Wikidata. But knowing Wikidata and its community a bit, it is obvious that we can not and should not simply upload Freebase data to Wikidata: Wikidata would prefer the data to be referenced to external, primary sources.

In order to do so, Google will soon start to work on an Open Source tool which will run on Wikimedia labs and which will allow Wikidata contributors to find references for a statement and then upload the statement and the reference to Wikidata. We will release several sets of Freebase data ready for consumption by this tool under a CC0 license. This tool should also work for statements already in Wikidata without sufficient references, or for other datasets, like DBpedia and other machine extraction efforts, etc. To make sure we get it right, we invite you to participate in the design and development of this tool.

I hope you are as excited as I am about this project, and I hope that you will join me in making this a reality. I am looking forward to your contributions! --Denny (talk) 19:24, 16 December 2014 (UTC)

Will Wikidata be used for the Google Knowledge Graph? Sjoerd de Bruin (talk) 19:57, 16 December 2014 (UTC)
The Google Knowledge Graph relies on structured data from all over the Web, and Wikidata is already part of that. --Denny (talk)
@Denny: Will the freebase identifiers still serve a purpose if the site is shut down? Also: Will the former editors of Freebase receive an invitation to join Wikidata? --Tobias1984 (talk) 23:01, 19 December 2014 (UTC)
What do you mean with 'invitation'? Our message post was pretty explicitly pointing to Wikidata - do you mean something else?
The Freebase identifiers will be used to map the entities for the data releases we are planning, so they are quiet crucial for that step. --Denny (talk) 23:47, 19 December 2014 (UTC)
@Denny: I actually meant after the data releases. Will the identifiers still lead to a page in e.g. 5 years? - About the invitations: I thought that maybe anyone with more than 1000 edits on Freebase would get an Email inviting them to Wikidata:WikiProject Freebase. --Tobias1984 (talk) 11:28, 20 December 2014 (UTC)
I'll check back on both - but it may take a while, a lot of people are already on their holiday breaks. --Denny (talk) 16:19, 22 December 2014 (UTC)

Can we associate Wiktionary links with article topics?

A topic like "Architecture" or "Zebra" will have both a Wikipedia article and a Wiktionary entry (or Wiktionary entries in many languages). Is there any reason we don't provide a place to associate the Wiktionary link to the Wikipedia topic? BD2412 (talk) 04:06, 17 December 2014 (UTC)

Wikipedia articles describe concepts or groups thereof, while Wiktionary entries describe words or phrases. Each word or phrase can have multiple completely different meanings (in dependence on language and context), which are all collected in one entry. Identically-named entries in different language editions correspond to each other 1:1, so there is no need to provide explicit links between them. Petr Matas 06:22, 17 December 2014 (UTC)
Even so, the Wiktionary entry for a common noun like "zebra" will primarily correspond to the Wikipedia article for "Zebra"; a Wikipedia entry like "Architecture" has a describable relationship with the Wiktionary entries for "architecture", "architectural", and "architect" among others. I think there would be some utility there. BD2412 (talk) 18:14, 19 December 2014 (UTC)
@BD2412:, FYI: from my point of view zebra (Q32789) should be connected with , and architecture (Q12271) with . I suppose this (and your proposal too) will be done with some kind of properties. --Infovarius (talk) 21:30, 22 December 2014 (UTC)
We will need to do considerable custom development for Wiktionary before we can add it. The planning is at Wikidata:Wiktionary. --Lydia Pintscher (WMDE) (talk) 18:24, 22 December 2014 (UTC)

Tree Of Life

Hey,

Since the Tree of Life by Denny is outdated, we thought it was a nice idea, to have a new one, to have an overview over the biological taxonomy on wikidata. Not only to have a nice looking tree but also to see, where errors are and to correct and update it. Right now, there are 660775 Items in the tree. Even though the change of names can be seen instantly, because this is based on the API, changes in the order need an update of the whole tree, because it's based on wikidata dumps. (This one on the most recent one from 15.12.2014)

Here you go, this is the new tree of life, made with a lot of love: https://tools.wmflabs.org/tree-of-life/

If you have any corrections, additions or features you want to add, feel free to ping me, send me a mail, submit a patch or file an issue on github. The repo for the tree is on https://github.com/frimelle/tree-of-life

Cheers, --Frimelle (talk) 16:35, 18 December 2014 (UTC)

This is awesome! Thank you.
What I always try is, to find humans. Here I gave up, though. Would you know if humans are connected already in your ToL? --Denny (talk) 18:18, 18 December 2014 (UTC)
Afaik human is not in the tree because of some missing link. --Lydia Pintscher (WMDE) (talk) 19:54, 18 December 2014 (UTC)
@Lydia Pintscher (WMDE): Taken out of context your comment is really funny. I guess the theory of evolution had a good run ;) --Tobias1984 (talk) 20:16, 18 December 2014 (UTC)
All intentional ;-) --Lydia Pintscher (WMDE) (talk) 20:46, 18 December 2014 (UTC)
The missing link is deuterostome (Q150866) parent taxon (P171) Bilateria (Q5173), but it is contained in Wikidata. Nice to see you having fun. :) Petr Matas 21:12, 18 December 2014 (UTC)
it was displaying weirdly on my screen, the elements of the page were overlapsing or requiring horizontal scrolling so I made a pull request to fix that, hope it helps :) Bests, Zorglub27 (talk) 11:06, 20 December 2014 (UTC)
Thanks, awesome you contributed! --Frimelle (talk) 16:35, 20 December 2014 (UTC)
Wikidata holds something like three times as many items as that that should fit. A quick look at angiosperms shows that this tree is missing more than 95% of taxa. Also Wikidata is intended to hold multiple trees of life, so any one 'tree' is likely to be weird. - Brya (talk) 17:49, 20 December 2014 (UTC)
They are not connected to an item in the tree then through the parent taxon property. --Lydia Pintscher (WMDE) (talk) 18:26, 22 December 2014 (UTC)
Less than 5 percent of items with taxon name (P225) (ca.1,900,000) have no parent taxon (P171). The query TREE[2382443][][171]| finds more than 1,700,000 connections to Biota (Q2382443). --Succu (talk) 18:41, 22 December 2014 (UTC)
Do you have examples of something that should be in the tree because it is connected on Wikidata but is not in the tree? Would it help if Frimelle dumps all the unconnected ones in a special tree and those can be worked on from there? --Lydia Pintscher (WMDE) (talk) 08:56, 23 December 2014 (UTC)

Why is this tree of life so disconnected from actual life? I just made Flipper (Q15831235) an instance of dolphins, but that class doesn't resolve to the class of all classes. 91.9.127.22 00:50, 23 December 2014 (UTC)

Are you talking about issues in the tree Frimelle created? Or the data here in Wikidata? It is unclear from your description what the actual issue is. --Lydia Pintscher (WMDE) (talk) 08:54, 23 December 2014 (UTC)

UI - order of events in significant events

For Q23380 I got : funeral, baptism, marriage, marriage.

Would it be possible that the Wikibase programmers make this more intuitive, e.g. by sorting by "point in time" which existed for each of the aforementioned events? 91.9.110.153 03:18, 22 December 2014 (UTC)

I think that a statement "this property is an ordering key" should be added to point in time (P585) and we will probably need a new property for that. Note that it is also possible to order the statements manually. Petr Matas 06:14, 22 December 2014 (UTC). Update: Petr Matas 11:34, 22 December 2014 (UTC)
Thank you very much, interesting idea. Regarding the diff see below. 91.9.127.22 11:59, 22 December 2014 (UTC)
User:Petr Matas - if sorting is done by such statements, it should go to significant event (P793), because in other context 'point in time' might not be the default sort key. 91.9.127.22 00:43, 23 December 2014 (UTC)
I think that the preferred ordering key could be specified by another statement: significant event (P793) ordering key point in time (P585). Petr Matas 08:05, 23 December 2014 (UTC)

UI - diff with changes but saying -no difference-

Petr Matas posted this one-edit-diff. It says in edit summary "(‎Changed claim: Property:P793: Q35856)" but in the diff-area it says "(No difference)". Could the Wikibase programmers make this more intuitive? Was there a change or not? If there was, what was changed? 91.9.127.22 11:59, 22 December 2014 (UTC)

Is changed the order of the claims. --ValterVB (talk) 14:14, 22 December 2014 (UTC)
Could the Wikibase programmers show that? Also, how can the change of order be done? 91.9.127.22 17:22, 22 December 2014 (UTC)
That's tracked at phabricator:T59918. --Lydia Pintscher (WMDE) (talk) 18:28, 22 December 2014 (UTC)
Open since 2013-12-03, Happy birthday dear bug! ... and blocked by "No, we have postponed this to be part of the UI redesign." ... See you in 2020. 91.9.127.22 00:38, 23 December 2014 (UTC)
An interim solution has been proposed there. You are welcome to implement it yourself or bring in a programmer to do it. Petr Matas 07:54, 23 December 2014 (UTC)

wikidata properties listed in a JSON file as key-values

I needed a simple per-language list of wikidata properties ids and labels for an application prototype, so I made a little script to request them from the API: https://github.com/maxlath/wikidata-properties-dumper

the output looks like this: { 

   "properties": {
       "P6": "head of government",
       "P7": "brother",
       "P9": "sister",
       "P10": "video",
       "P14": "highway marker",
       "P15": "road map",
       "P16": "highway system",
       "P17": "country",
       "P18": "image",
       ...

and that's very handy :)

Is that a duplicate of an other existing tool? I couldn't figure out how to extract those list from the dumps, so the current script is a bit brutal with the API, making currently about 34 requests at a time (= 1700 properties / 50 max ids per request), is that a problem? Any feedback welcome! Zorglub27 (talk) 10:40, 23 December 2014 (UTC)

Try http://quarry.wmflabs.org/query/1248 --- Jura 12:46, 23 December 2014 (UTC)
wow, wasn't aware of Quarry, awesome! Thanks!  – The preceding unsigned comment was added by Zorglub27 (talk • contribs).
I also didn't see that page yet. Really nice. Just the documentation is a little short: https://meta.wikimedia.org/wiki/Research:Quarry --Tobias1984 (talk) 16:16, 23 December 2014 (UTC)

Tree Of Life

Hey,

Since the Tree of Life by Denny is outdated, we thought it was a nice idea, to have a new one, to have an overview over the biological taxonomy on wikidata. Not only to have a nice looking tree but also to see, where errors are and to correct and update it. Right now, there are 660775 Items in the tree. Even though the change of names can be seen instantly, because this is based on the API, changes in the order need an update of the whole tree, because it's based on wikidata dumps. (This one on the most recent one from 15.12.2014)

Here you go, this is the new tree of life, made with a lot of love: https://tools.wmflabs.org/tree-of-life/

If you have any corrections, additions or features you want to add, feel free to ping me, send me a mail, submit a patch or file an issue on github. The repo for the tree is on https://github.com/frimelle/tree-of-life

Cheers, --Frimelle (talk) 16:35, 18 December 2014 (UTC)

This is awesome! Thank you.
What I always try is, to find humans. Here I gave up, though. Would you know if humans are connected already in your ToL? --Denny (talk) 18:18, 18 December 2014 (UTC)
Afaik human is not in the tree because of some missing link. --Lydia Pintscher (WMDE) (talk) 19:54, 18 December 2014 (UTC)
@Lydia Pintscher (WMDE): Taken out of context your comment is really funny. I guess the theory of evolution had a good run ;) --Tobias1984 (talk) 20:16, 18 December 2014 (UTC)
All intentional ;-) --Lydia Pintscher (WMDE) (talk) 20:46, 18 December 2014 (UTC)
The missing link is deuterostome (Q150866) parent taxon (P171) Bilateria (Q5173), but it is contained in Wikidata. Nice to see you having fun. :) Petr Matas 21:12, 18 December 2014 (UTC)
it was displaying weirdly on my screen, the elements of the page were overlapsing or requiring horizontal scrolling so I made a pull request to fix that, hope it helps :) Bests, Zorglub27 (talk) 11:06, 20 December 2014 (UTC)
Thanks, awesome you contributed! --Frimelle (talk) 16:35, 20 December 2014 (UTC)
Wikidata holds something like three times as many items as that that should fit. A quick look at angiosperms shows that this tree is missing more than 95% of taxa. Also Wikidata is intended to hold multiple trees of life, so any one 'tree' is likely to be weird. - Brya (talk) 17:49, 20 December 2014 (UTC)
They are not connected to an item in the tree then through the parent taxon property. --Lydia Pintscher (WMDE) (talk) 18:26, 22 December 2014 (UTC)
Less than 5 percent of items with taxon name (P225) (ca.1,900,000) have no parent taxon (P171). The query TREE[2382443][][171]| finds more than 1,700,000 connections to Biota (Q2382443). --Succu (talk) 18:41, 22 December 2014 (UTC)
Do you have examples of something that should be in the tree because it is connected on Wikidata but is not in the tree? Would it help if Frimelle dumps all the unconnected ones in a special tree and those can be worked on from there? --Lydia Pintscher (WMDE) (talk) 08:56, 23 December 2014 (UTC)

Why is this tree of life so disconnected from actual life? I just made Flipper (Q15831235) an instance of dolphins, but that class doesn't resolve to the class of all classes. 91.9.127.22 00:50, 23 December 2014 (UTC)

Are you talking about issues in the tree Frimelle created? Or the data here in Wikidata? It is unclear from your description what the actual issue is. --Lydia Pintscher (WMDE) (talk) 08:54, 23 December 2014 (UTC)

UI - order of events in significant events

For Q23380 I got : funeral, baptism, marriage, marriage.

Would it be possible that the Wikibase programmers make this more intuitive, e.g. by sorting by "point in time" which existed for each of the aforementioned events? 91.9.110.153 03:18, 22 December 2014 (UTC)

I think that a statement "this property is an ordering key" should be added to point in time (P585) and we will probably need a new property for that. Note that it is also possible to order the statements manually. Petr Matas 06:14, 22 December 2014 (UTC). Update: Petr Matas 11:34, 22 December 2014 (UTC)
Thank you very much, interesting idea. Regarding the diff see below. 91.9.127.22 11:59, 22 December 2014 (UTC)
User:Petr Matas - if sorting is done by such statements, it should go to significant event (P793), because in other context 'point in time' might not be the default sort key. 91.9.127.22 00:43, 23 December 2014 (UTC)
I think that the preferred ordering key could be specified by another statement: significant event (P793) ordering key point in time (P585). Petr Matas 08:05, 23 December 2014 (UTC)

UI - diff with changes but saying -no difference-

Petr Matas posted this one-edit-diff. It says in edit summary "(‎Changed claim: Property:P793: Q35856)" but in the diff-area it says "(No difference)". Could the Wikibase programmers make this more intuitive? Was there a change or not? If there was, what was changed? 91.9.127.22 11:59, 22 December 2014 (UTC)

Is changed the order of the claims. --ValterVB (talk) 14:14, 22 December 2014 (UTC)
Could the Wikibase programmers show that? Also, how can the change of order be done? 91.9.127.22 17:22, 22 December 2014 (UTC)
That's tracked at phabricator:T59918. --Lydia Pintscher (WMDE) (talk) 18:28, 22 December 2014 (UTC)
Open since 2013-12-03, Happy birthday dear bug! ... and blocked by "No, we have postponed this to be part of the UI redesign." ... See you in 2020. 91.9.127.22 00:38, 23 December 2014 (UTC)
An interim solution has been proposed there. You are welcome to implement it yourself or bring in a programmer to do it. Petr Matas 07:54, 23 December 2014 (UTC)

wikidata properties listed in a JSON file as key-values

I needed a simple per-language list of wikidata properties ids and labels for an application prototype, so I made a little script to request them from the API: https://github.com/maxlath/wikidata-properties-dumper

the output looks like this: { 

   "properties": {
       "P6": "head of government",
       "P7": "brother",
       "P9": "sister",
       "P10": "video",
       "P14": "highway marker",
       "P15": "road map",
       "P16": "highway system",
       "P17": "country",
       "P18": "image",
       ...

and that's very handy :)

Is that a duplicate of an other existing tool? I couldn't figure out how to extract those list from the dumps, so the current script is a bit brutal with the API, making currently about 34 requests at a time (= 1700 properties / 50 max ids per request), is that a problem? Any feedback welcome! Zorglub27 (talk) 10:40, 23 December 2014 (UTC)

Try http://quarry.wmflabs.org/query/1248 --- Jura 12:46, 23 December 2014 (UTC)
wow, wasn't aware of Quarry, awesome! Thanks!  – The preceding unsigned comment was added by Zorglub27 (talk • contribs).
I also didn't see that page yet. Really nice. Just the documentation is a little short: https://meta.wikimedia.org/wiki/Research:Quarry --Tobias1984 (talk) 16:16, 23 December 2014 (UTC)

Floor leader property (properties?)

I was thinking of proposing the property "floor leader" for use on infoboxes like en:Template:Infobox legislative session. However, those infoboxes have four different kinds of floor leaders: lower-house government, lower-house opposition, upper-house government, and upper-house opposition.

As far as I know, Wikipedia's syntax for Wikidata items does not allow you distinguish between multiple values for a property. For example, if there were four values under "Property:floor leader", I couldn't tell a Wikipedia Infobox to look for the one with qualifiers "of->senate" and "of->majority".

Therefore, should I propose that we create four new properties for floor leaders? --Arctic.gnome (talk) 23:59, 23 December 2014 (UTC)

Library systems and museum networks

Do we have a preferred instance of (P31) value for a centrally-managed group of libraries, museums, etc.? I'm seeing a variety of values (and even subclasses, which is clearly wrong). Some are foundations, some are corporations, but most are ... something else. 'Organization' is too vague. - PKM (talk) 02:19, 24 December 2014 (UTC)

changing issue

I cannot edit wikidata links since yesterday. Is there any problem. Could you check it? This is the only solution for me :/--Sabri76 (talk) 10:01, 23 December 2014 (UTC)

I didn't have any problem to add the link. Do you get any error messages or why you can't add links? --Stryn (talk) 10:05, 23 December 2014 (UTC)
I've faced with this issue for the first time. When I tried to click "edit" icon, nothing happened and there is no such error messages or action.--Sabri76 (talk) 10:27, 23 December 2014 (UTC)
Actually, I don't speak French, so I didn't understand. However, now, I don't have a problem and I can change them. That's interesting...--Sabri76 (talk) 22:02, 24 December 2014 (UTC)

What to do when the only article linked from a wikidata item is removed?

I've found cases where an article in a wiki is removed, for instance, because the subject was considered not relevant enough. Next, the link to the article in the wikidata item is removed. Thus, an orphan item remains with no possibility of being linked again. What should be done in that case? BT --Discasto (talk) 01:33, 25 December 2014 (UTC)

I'd say -- If it meets Wikidata's notability guidelines, leave it be. If not, request deletion. (And err toward keeping the item.) - PKM (talk) 01:47, 25 December 2014 (UTC)
If there are statements e.g. with external identifiers or dates related to the person, or there are sources for statements, my decision is to keep it. Of course, Wikidata's notability is more important. Matěj Suchánek (talk) 09:40, 25 December 2014 (UTC)

Sourcing Question

I am slightly unclear on exactly how to source certain statements, for example in Meneldil, should I list the source as stated in: The Lord of the Rings or stated in: Appendices of the Lord of the Rings? Furthermore, if more specific is preferable, is there a way that I could specify Apendix A, section ii?

You have to create a new item for a specific edition of the book and use chapter (P792), section, verse, paragraph, or clause (P958) and page(s) (P304). Snipre (talk) 18:50, 25 December 2014 (UTC)
Sorry, but I don't think you've answered my question, am I supposed to reference the entire Lord of the Rings (book) or just the Appendices (smaller section of said book)?

post weekly summary here?

Hey folks :)

As our community is growing we need to make sure we distribute important information to everyone who needs it. One way to do that is the weekly summary that exists since the beginning of the project. It is posted to a number of user talk pages, the mailing list, twitter, facebook, google+ and identi.ca. I was thinking it makes sense to also post it here. Can I get a yay/nay on this please? --LydiaPintscher (talk) 18:43, 23 December 2014 (UTC)

 Support --Hsarrazin (talk) 19:03, 23 December 2014 (UTC)
 Support Why not all languages of project chat with a notice to translate the message? --Tobias1984 (talk) 19:15, 23 December 2014 (UTC)
I had not thought of that. Good point. --LydiaPintscher (talk) 19:24, 23 December 2014 (UTC)
 Support - PKM (talk) 20:05, 23 December 2014 (UTC)
If this is unanimous on Saturday, I'll add it to the spam list for the summary. If not, I'll let the week take its course. (FYI) John F. Lewis (talk) 20:13, 23 December 2014 (UTC)
 Support Like it. Sweet kate (talk) 20:42, 23 December 2014 (UTC)
 Support Great! --AmaryllisGardener talk 20:50, 23 December 2014 (UTC)
 Support, sure, as John F. Lewis said, add it to the spam list. ;) --Stryn (talk) 21:37, 23 December 2014 (UTC)
 Support. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:43, 23 December 2014 (UTC)
 Support Do not see any issues.--Ymblanter (talk) 10:26, 24 December 2014 (UTC)

Simon Cowell

I just reverted back some vandalism on the Simon Cowell Q162629 entry, but I am not experienced enough to know how to handle the other issue I found. Q5976940 is the book "I Don't Mean to be Rude, but", the author appears as Simon 'Bootylicious' Cowell. It is the link to Q162629, but is showing the vandalised label. When I edit, the corrected name "Simon Cowell" appears but the save link is disabled. Presumably the text is cached but is there anything I can to to updated the entry? Periglio (talk) 11:02, 26 December 2014 (UTC)

@Periglio: The quick fix was just to take out the info and immediately reinsert it. —Justin (koavf)❤T☮C☺M☯ 11:25, 26 December 2014 (UTC)
Usually what helps is purging the page. Add ?action=purge to the end of the URL to do that. --Lydia Pintscher (WMDE) (talk) 11:26, 26 December 2014 (UTC)
@Lydia Pintscher (WMDE): I tried that with Ctrl+Shift+R, which usually works the same as Purge but it didn't work for me. —Justin (koavf)❤T☮C☺M☯ 11:28, 26 December 2014 (UTC)
?action=purge does a bit more than that. --Lydia Pintscher (WMDE) (talk) 11:29, 26 December 2014 (UTC)

Are we still running interwiki bots?

I just discovered that this category was created on October 23 (two months ago) and until five minutes ago had no Wikidata item. This is not the first time I see that. For example, I am active on AfD of the English Wikipedia; most of the articles (which have to be there at least a week before they get deleted) have no Wikidata item at the moment of deletion. The situation with other projects is even worse. For example, on Russian Wikivoyage we have a large number of pages on cultural heritage like this one. Some of them are around for a year, and none of them has a Wikidata item except for those we have manually added to Wikidata. I am afraid this sitiation is not really acceptable, and I would strongly encourage owners of interwiki bots run their bots on a regular basis over all Wiikimedia projects eligible for having Wikidata items.--Ymblanter (talk) 10:39, 26 December 2014 (UTC)

Hi Ymblanter, I wouldn't call it interwiki bots. That makes me (and probably other bot operators too) think of the bots we had running around a couple of years ago to spread the links. You want a bots to create new items for new articles on a regular basis.
Regular maintenance with a bot is always a bit of a problem. As a bot operator it's more fun the solve incidental problems (initial import of items) than keeping up with maintenance (import of new items). I started User:JanitorBot with this problem in mind. The idea is to have one bot operated by multiple people doing the boring maintenance tasks.
Pywikibot already contains a bot to do this task, but the problem is that Wikibase doesn't have an api version of en:Special:UnconnectedPages (phab:T51137), so pywikibot doesn't have a generator for it either (phab:T56738). We have to do a database query for each wiki ourselves. That's a bit of a pain. For example query and result for the Dutch Wikipedia.
I'll see if I can spend some time on the JanitorBot in the next weeks. Multichill (talk) 11:35, 27 December 2014 (UTC)
The maintenance task gets hindered as we agreed not to create items for all pages, e.g. /doc subpages, archiving pages, soft redirects, meetup pages and misspelling articles are all excluded. However, not only bots can create items but there is also Wikidata item creator with which actually everyone can do this task. --Pasleim (talk) 12:01, 27 December 2014 (UTC)

VisualEditor News #10—2014

VisualEditor
Screenshot showing how to add or remove columns from a table

Did you know?

Basic table editing is now available in VisualEditor. You can add and remove rows and columns from tables at the click of a button.

The user guide has more information about how to use VisualEditor.

Since the last newsletter, the Editing Team has fixed many bugs and worked on table editing and performance. Their weekly status reports are posted on mediawiki.org. Upcoming plans are posted at the VisualEditor roadmap.

VisualEditor was deployed to several hundred remaining wikis as an opt-in beta feature at the end of November, except for most Wiktionaries (which depend heavily upon templates) and all Wikisources (which await integration with ProofreadPage).

Recent improvements

Basic support for editing tables is now available. You can add and delete tables, add and remove rows and columns, set or remove a caption for a table, and merge cells together. To change the contents of a cell, double-click inside it. More features will be added in the coming months. In addition, VisualEditor now ignores broken, invalid rowspan and colspan elements, instead of trying to repair them.

You can now use find and replace in VisualEditor, reachable through the tool menu or by pressing ^ Ctrl+F or ⌘ Cmd+F.

You can now create and edit simple <blockquote> paragraphs for quoting and indenting content. This changes a "Paragraph" into a "Block quote".

Some new keyboard sequences can be used to format content. At the start of the line, typing "* " will make the line a bullet list; "1." or "#" will make it a numbered list; "==" will make it a section heading; ": " will make it a blockquote. If you didn't mean to use these tools, you can press undo to undo the formatting change.

There are also two other keyboard sequences: "[[" for opening the link tool, and "{{" for opening the template tool, to help experienced editors. The existing standard keyboard shortcuts, like ^ Ctrl+K to open the link editor, still work.

If you add a category that has been redirected, then VisualEditor now adds its target. Categories without description pages show up as red.

You can again create and edit galleries as wikitext code.

Looking ahead

The current VisualEditor design will be replaced with a new theme designed by the User Experience group. The new theme will be visible for desktop systems at mediawiki.org in late December and on other sites in early January. (You can see a developer preview of the old "Apex" theme and the new "MediaWiki" one which will replace it.)

The Editing team plans to add auto-fill features for citations in January.

Planned changes to the media search dialog will make choosing between possible images easier.

Let's work together

— Elitre (WMF)

19:00, 26 December 2014 (UTC)

Elitre (WMF), bit pointless spamming 10 pages on this wiki with an editor that is not used here. I removed the pages. Multichill (talk) 20:39, 26 December 2014 (UTC)
But VE is actually used here, though ofc not in the main namespace and so. --Stryn (talk) 21:02, 26 December 2014 (UTC)
Oh? Can you give an example Stryn? It doesn't seem to be enabled in this namespace, user namespace or any of the talk namespaces I checked. Multichill (talk) 11:01, 27 December 2014 (UTC)
Yeah, it's only a beta feature (opt-in) at the moment. --Stryn (talk) 11:03, 27 December 2014 (UTC)
Exactly. Despite people barely knowing this, almost 100 people on this project are already using VE, so I can actually see the point. Is there a better page I should deliver to? Thank you. --Elitre (WMF) (talk) 12:33, 27 December 2014 (UTC)

Optimising the main page for mobile! (again)

I'm still yet to get a reply about this (see earlier request and would really appreciate identifying someone who can help me make the home page better suited for mobile devices. Maybe User:Legoktm or User:SPQRobin or User:Stryn based on recent Common.css edits? Jdlrobson (talk) 05:40, 24 December 2014 (UTC)

@Jdlrobson: all those changes are done. But for some reason it doesn't show up yet on https://m.wikidata.org/wiki/Wikidata:Main_Page. Maybe it just takes some time? --Stryn (talk) 09:13, 24 December 2014 (UTC)
Heh, ok, of course I had to mark it for translation, done now, and now it looks like it should. Then someone should change those ugly tables on Wikidata:Main Page/Contact to something else. --Stryn (talk) 09:21, 24 December 2014 (UTC)
I took the liberty of getting rid of the table. —Wylve (talk) 12:11, 24 December 2014 (UTC)
There seems to be something broken in the code, see https://m.wikidata.org/wiki/Wikidata:Main_Page?uselang=en. I couldn't figure out what should be fixed and have no time to look at it more right now. --Stryn (talk) 13:15, 24 December 2014 (UTC)
I think I have fixed the code. Please mark it for translation, thanks. —Wylve (talk) 14:22, 24 December 2014 (UTC)
Thanks! Yes it looks good now. And thanks for Matěj Suchánek for marking it for translation. --Stryn (talk) 15:02, 24 December 2014 (UTC)

User talk:Stryn it looks great! Thank you so much for getting this done! next step mobile editing!! :-) Jdlrobson (talk) 01:30, 28 December 2014 (UTC)

property "introduced feature" (P:P751)

For me it's not clear if the "introduced" just refers to the series the item belongs, or it's meant it's a feature which is general introduced for the first time.

Examples:

Q18681531: is backup encryption just used the first time in the MS SQL Server series or the first product ever which supports backup encryption (in this case I know it just refers to the MS SQL series)

Q1469766: is this the first Handley Page aircraft that contains an aircraft lavatory or was this the first aircraft at all which has an lavatory?

This uncertainty is for every statement which use this property (http://www.wikidata.org/w/index.php?title=Special%3AWhatLinksHere&target=Property%3AP751&namespace=0&uselang=de)

Either we define the meaning (scope) of this property in the description ("item related to the product line / company" or "overall, NOT related to product line / company of this item") or we've to use qualifier each time using this property which qualifies the scope what exactly "introduced" is refering to.

My proposal is to use two properties: one for item related meaning and one for overall meaning. If others agree it would be nice someone could make the property proposal (or just create it when all agrees).

Kind regards --77.239.38.225 20:36, 27 December 2014 (UTC)

Property proposal votes needed

I need to fast-track (if possible) 5 properties with medical identifiers. Please consider voting on the proposals starting with the linked one: Wikidata:Property_proposal/Natural_science#ICD-10-PCS --Tobias1984 (talk) 23:23, 27 December 2014 (UTC)

Uploading a statistics table (csv) to WikiData

How would I go about uploading a CSV or table of statistics to WikiData that would be used on a particular Wikipedia article? Here's my specific use-case:

I want to upload the table "US Crime Rate by Year" for use on the wikipedia article, "Crime in the United States". The CSV table has the following column headers: Year, Population, Total, Violent Crime, Property Crime, Murder, Forcible Rape, Robbery, Aggravated Assault, Burglary, Larceny, Auto-theft. Then there's 55 rows under those column headers containing how many of each type of crime occurred during each year.

How would I import that table into WikiData so that I can query that table, and dynamically dump the result into a sortable table and SVG chart graphic that I've added to the Wikipedia article?

My goal is to generate a table and graphic that could be dynamically updated each year by just adding a single row to the wikidata source table, instead of having to manually add values to the SVG graphic code and/or wikipedia tables (both on that article or on any other Wikipedia articles that reference the same data). --Chrisbat92024 (talk) 22:30, 26 December 2014 (UTC)

@Chrisbat92024: Hi. Before adding any value, you have to create the corresponding property. Wikidata uses the system "one property to one value". So you have to define the properties you want to describe the table and then to find some approval from the community to create these properties. For that propose the properties under Wikidata:Property proposal. Then once you have the properties created you can ask a bot to fill them on that page Wikidata: Bot requests. Snipre (talk) 12:57, 27 December 2014 (UTC)
@Snipre: Hi Snipre, I'm a little unclear though on how I would structure this data in order to be able to query it to form something that resembles the original data structure (in order to make a graphic or table from it). Normally if I stored these stats in MySQL, they would be in the form of a "list of lists containing property=>values", like the following:
[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]
Where as the wikidata structure seems to be a 1-dimensional list of property-values:
[List Name (Ex. "List of Crime Rates in 1973")]=>
[Property (Ex. "Murder")]=>[Value]
[Property (Ex. "Theft")]=>[Value]
[Property (Ex. "Assault")]=>[Value]
Since multi-column tables are "lists of lists of property-values" and not just a "list of property-values", how would I go about importing that into Wikidata so that I could query it in a logical fashion when creating tables or graphics? Thanks for your help! --Chrisbat92024 (talk) 22:31, 28 December 2014 (UTC)

Roadmap for Wikidata?

I hope this isn't OT, but I was wondering where I could find a roadmap for this project.

As a patroller, I'm espspecially interested in whether there are any plans for an automatic cross-wiki update mechanism for infoboxes. If you have a look at e.g. soccer player info (e.g. current team, number of goals) or country/city info (such as number of inhabitants), these are currently updated on an individual basis on each and every Wikipedia edition. For one thing, that means there are a lot of patrollers checking the same updates on lots of wikis, and I'm frankly not sure that every minor change is being fact-checked. A centralized system under strict auspices would relieve patrollers all over the worls of some tedious routine work and also make sure that the factual integrity is taken care of. Obvioulsy, this means that any changes in Wikidata would have to be monitored vigorously, as they'd propagate to Wikipedia editions all over.

Is this use of Wikidata being considered or already in the works? Asav (talk) 10:55, 29 December 2014 (UTC)

The roadmap is at Wikidata:Development plan.
Current usage of data per project is at Category:Templates using data from Wikidata (Q11985372) or you can check {{ExternalUse}} on a property talk page. Matěj Suchánek (talk) 11:10, 29 December 2014 (UTC)
Thanks! Much appreciated! Asav (talk) 11:37, 29 December 2014 (UTC)
In the tennis project I have edited a lot of infoboxes in many languagesto keep some of the statistics that are on Wikidata up to date across many languages. Unfortunately, not every language community likes it the same that a Dutch guy comes along to edit their infobox-templates, so some of the edits were reverted, but on the Spanish and Russian wiki's the statistics come from Wikidata for the last 18 months. And I see a huge group of users keeping those statistics up to date, and the good news is, that that will keep the article of a player up to date in many languages in one go. Edoderoo (talk) 15:05, 29 December 2014 (UTC)

Help! Trying to create an article

To whom it may concern, We are a well established financially secure independent record label. We would like to create an article with our information available here as well as our country music artists that we sign to the label. Looking forward to hearing from you. Have a great day!

All the best, 23 Muzik Group (www.23muzikgroup.com)  – The preceding unsigned comment was added by 23 Muzik Group (talk • contribs).

@23 Muzik Group: English Wikipedia is what you're looking for, but please have a look at this first. --AmaryllisGardener talk 15:41, 29 December 2014 (UTC)

input for article placeholder

Hey folks :)

People have been bugging me for a while now about this so I started a page to gather input.

If you are searching for a topic on a Wikipedia for example but it doesn't have an article about it then we can look at Wikidata and see if there is a matching item for that topic. If we have data about it on Wikidata then we can show some kind of placeholder. But how should this look like and work? I'd love your input at Wikidata:Article placeholder input. Please keep in mind the important points at the beginning of that page.

Also if someone comes up with a better name for this that'd be awesome.

<3

Cheers --Lydia Pintscher (WMDE) (talk) 18:11, 29 December 2014 (UTC)

WDQ not updating ?

Is it just me, or is WDQ no longer getting synchronised with Wikidata updates?

eg: 

  http://tools.wmflabs.org/autolist/autolist1.html?q=items%5B2910848%5D%20AND%20claim%5B373%5D

returns nothing, even though Q2910848 was updated with a P373 on 12 December at 20:09 (diff).

Similarly 

 http://tools.wmflabs.org/autolist/autolist1.html?q=STRING%5B1367%3A%22master-of-the-fogg-pieta-active-c-13101340%22%5D

returns nothing, even though Q1751201 was updated with this property on 12 December at 20:48 (diff).

It's frustrating trying to undertake projects (eg in this case: finding Q-numbers for painters in public holdings in the UK) if one can't reliably update the project's tracking pages.

I know Magnus has been poorly, so perhaps it's unfair to raise this at this time. But surely such a mission-critical tool should not still be depending only on Magnus? -- Jheald (talk) 19:08, 17 December 2014 (UTC)

Hi James, http://wdq.wmflabs.org/stats contains some information about sync status (look at the timestamp). That looks ok, but some updates might have gotten lost.
WDQ started as a fun test toy and like a lot of fun test toys in ICT, it's importance grew over time, but it's infrastructure (technical and people) didn't. So yet again we end up with a community best-effort supported tool. This problem keeps popping up and we as movement don't seem to be able to crack it. One of the solutions we talked about in the past is that the WMF and/or chapters adopt important tools. The big hurdle to take here is the quality and language of the code. Multichill (talk) 19:59, 17 December 2014 (UTC)
@Multichill: Interesting. The first query is now working. The second still isn't. (Nor in this form. Though this works; but that doesn't help me search). Jheald (talk) 20:11, 17 December 2014 (UTC)
Yuvi from the Labs team has been actively working with Magnus on improving reliability of the WDQ service in its current state, and WMF will develop a scalable production-grade query service as a top priority in partnership with the Wikidata team; see this page for current notes on the technical assessment and of course please do chip in.--Erik Moeller (WMF) (talk) 21:25, 17 December 2014 (UTC)
I've just switched wdq.wmflabs.org to run on two instances load balanced instead of one, which should definitely help reduce random outages. Let me know if there's anything wrong / missing. Yuvipanda (talk) 08:02, 19 December 2014 (UTC)
Thanks @Yuvipanda: that seems a lot better. I've updated the pages, and most of what I've added recently does now seem to be there; but I did still notice a few anomalies:
  • Domenico di Zanobi - has a P1367 Autolist; but not if you try to search on it Autolist. Added 13 Dec, 00:30 diff
  • Stanley Spencer - P650 found by Autolist [22]; but not by WDQ: [23]. Added 18 Dec, 00:05 diff.
  • Antonio Vázquez - P1367 on Autolist; but not if searched for (Autolist). Added 13 Dec, 00:01 [24]
  • Francisco de Osona - P1367 on Autolist; but not if searched for (Autolist). Added 13 Dec, 19:25 diff
  • Gerino da Pistoia - P1367 on Autolist; but not if searched for (Autolist). Added 13 Dec, 19:40 diff
  • Master of the Baroncelli Portraits - P1367 on Autolist; but not if searched for (Autolist). Added 15 Dec 23:31 diff.
These are just the first ones I have looked at. There could be 20 more... Jheald (talk) 18:29, 19 December 2014 (UTC)
I also noticed something strange with Autolist :
http://tools.wmflabs.org/autolist/autolist1.html?lang=fr&start=250&q=claim[107%3A215627]%20and%20noclaim[31]#
returns 2965 items], among them many are items deleted weeks ago… :/
how can a deleted item still appear in such a list ? if due to old dumps, would it please be possible to have a check that indicates that the item no longer exists ?
… it is very frustrating, while trying to clean up a list (here P107) and still having a high number of items, when, in fact, they no longer exist on wikidata, and very hard to check what are the still remaining wrong items :/ --Hsarrazin (talk) 14:54, 23 December 2014 (UTC)
The tool is updated based on Special:RecentChanges so you can do an edit to an item to trigger an update. Doesn't scale but gives you at least a way to get rid of some items in your lists. Multichill (talk) 20:38, 29 December 2014 (UTC)

WikidataQuery API

I need help. How do I create a query which would include only Wikipedia articles and not Categories and Templates? I need it for this tool.--Kohelet (talk) 21:51, 25 December 2014 (UTC)

You can add &q=claim[31:5] to the url. Sjoerd de Bruin (talk) 22:01, 25 December 2014 (UTC)
Thanks, but I can't try it, because tools on wmflabs seem to be down since yesterday.--Kohelet (talk) 08:49, 27 December 2014 (UTC)
It's wrong. It's just people, but I need all articles.--Kohelet (talk) 01:35, 30 December 2014 (UTC)

Painful slow progess - properties for deletion that still are used

  • Property:P766 (DEPRECATED) event location - 1,182 pages - closed 7 December 2014
  • Property:P45 (OBSOLETE) grandparent - 863 pages - closed 28 December 2013 - !!! Almost one year, ouch.
  • Property:P107 (OBSOLETE) main type (GND) - 77,626 pages - closed 1 October 2013 - !!! More than a year, double ouch.
  • Property:P70 (OBSOLETE) order - 568 pages - closed since 12 February 2014 - will that be around in February 2015?
  • Property:P741 (OBSOLETE) playing hand - 1,546 pages - closed 28 December 2013 - Almost one year, again an ouch.
  • Property:P132 (Will be deleted) type of administrative territorial entity - 27,027 pages - ONE YEAR OF DISCUSSION. closed 7 December 2014 - now again, one user is starting discussing the agreed upon replacement. Not blaming him, since he is one of the few removing many of these deprecated properties.

91.9.120.70 17:56, 16 December 2014 (UTC)

On the other side at least P107, P766, P70 and P132 are being worked on right now. There are also P513 and P643 which are deprecated, too. Matěj Suchánek (talk) 18:14, 16 December 2014 (UTC)
Based on the issue with qualifiers, P741 should probably be kept. --- Jura 07:55, 17 December 2014 (UTC)
FYI, on Property talk:P741 I started a discussion how to proceed with P741.--Pasleim (talk) 13:14, 30 December 2014 (UTC)
Property:P107 was, and still is, a very widly used property. I believe it was the single most used property when it was proposed for deletion. /ℇsquilo 10:09, 17 December 2014 (UTC)
This would be a good occasion to create {{Sofixit}} ;-)
But seriously, these properties have been deprecated so these will be deleted eventually. If you think it's important to do this sooner than later, help replace them. I rather have deprecated properties stay around a bit longer so we can properly replace them than having a quick and dirty removing/replacement leaving us with a mess to clean up. Multichill (talk) 11:42, 17 December 2014 (UTC)
It's a very long work to change P107 (P107) to instance of (P31) as the values accepted are not the same… 
I've been working on (claim[107:215627] and noclaim[31]) for some time now, but with Autolist that has problems updating (and also wmftool down) it is very slow. Also, many, many items are NOT Q5, so it has to be done manually, one by one… many are also empty items, resulting of uncomplete merges, and quite a lot of them have only chinese, japanese, georgian, or other languages that need a speaker of these language to be understood... to complete this work, a taskforce specialized in translation to English (at least) would be very helpful - since those are generally names :) --Hsarrazin (talk) 21:04, 17 December 2014 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ User:Hsarrazin's comment is a good reply showing that User:Multichill's SOFIXIT does not work, since the tools are a mess. Bots could do it all in a day. Regarding "It's a very long work to change instance of (P31) to P107 (P107) as the values accepted are not the same" - That is FUD. Current situation is P70 was deleted less than two hours after thread start. The others are still there. 80.134.84.12 14:49, 19 December 2014 (UTC) @Multichill "I rather have deprecated properties stay around a bit longer so we can properly replace them than having a quick and dirty removing/replacement leaving us with a mess to clean up." - Dirty is to have outdated properties not replaced with what was the result of Properties for Deletion. And the results for all of the above are VERY easy to implement. 80.134.84.12 14:58, 19 December 2014 (UTC)

Dirty is to replace in the following items P107 with P31: Q5025639, clothing in Mauritius (Q5135571), climate of the British Isles (Q5133613), Climarice (Q5133387) --Pasleim (talk) 15:15, 19 December 2014 (UTC)
Would User:Pasleim substantiate her claim? Why should that be dirty? 80.134.84.12 15:19, 19 December 2014 (UTC)
It's dirty because it includes propagating of wrong claims and the information gain is zero. To be honest, I think the data quality of P107 is poor. So I would really appreciate it if we have the time to create correct and useful claims instead of blind replacements. BTW, why are you doing such edits [25] [26]?
Would User:Pasleim substantiate her claim that the information gain is zero? I contest it, since it would immediately place items in the P31/P279-tree - there for anyone to perform automated checks on integrity. Regarding the edits - P107 is deprecated and replacing it with P31 or P279 was agreed to be the way. Cleaning up and improving is a completely different task. Not only the data quality of P107 is low, also that of P31 is. 91.9.121.131 22:05, 20 December 2014 (UTC)
sorry, been away for some days : and corrected my obvious inversion above
what I meant by "the values accepted are not the same" is that a lot of items (if not most items) stated as P107:Q215627 (persons) are NOT Q5 - they are groups, teams, duos, anything but Q5… which means that the job has to be done by hand and no bot can do it. Autolist can help, but just converting P107:Q215627 to P31:Q5 would not only propagate error, it would prevent to find it easily… at least, we know that all P107 have to be checked… once the items are all mixed with others, it will be much more difficult to find the errors…
I completely agree with User:Multichill : better keep the deprecated properties for some more time, and take time to do it properly, rather than a mess :) - on the other hand, it would be useful to block and prevent adding of deprecated properties to items ;) --Hsarrazin (talk) 14:05, 23 December 2014 (UTC)

P45

A bot can do it in a day. Wikidata:Requests_for_deletions/Archive/2013/Properties/4#Property:P45 says

80.134.84.12 15:14, 19 December 2014 (UTC)

Agree, this replacement could do a bot. --Pasleim (talk) 22:38, 19 December 2014 (UTC)

P132

Look at Q90 - to copy all the values in P132 by hand to P31 takes a lot of time if done via the Wikidata UI. A bot could do it, and editors could spend time on something more useful. 80.134.84.12 15:13, 19 December 2014 (UTC)

So what now? Are we collectiong knowledge or statements?

I already thought my statements about a hoax death of Axl Rose would be reverted and so it's that was happened (http://www.wikidata.org/w/index.php?title=Q11885&diff=180869978&oldid=179265474&uselang=en).

Yeah, it was just a hoax news. However, I was part of the Wikidata community in early beginnings and back then the goal of Wikidata was to collect not only knowledge, but statements, whatever true or not. See Denny's Blog: http://blog.wikimedia.de/2013/06/04/on-truths-and-lies/. Even he wrote it's just his point of view, the goal of such a project like Wikidata should be clear.

Sure, if I would make a statement like "person x died on xx.xx.xxxx" it's not worthy to include this statement here in Wikidata. But about the Axl death hoax there was several news sources which reported so and even Axl Rose himself denied his death the next day. I gave my statement the "deprecated" rank, which is especially designed for that purpose. However, others in this community think such statements doesn't belong to Wikidata at all and removed the statement.

For myself, I would support to collect also wrong statements, if there is a specified amount of sources which claims the statement. It gives e.g. an overview how much someone already was declared dead, even if it's just kind of a fun fact. It gives also the possibility to link to a source with disproof the hoax (in this case: http://loudwire.com/axl-rose-death-hoax-still-pay-taxes/) with a specific subproperty like "disproved on (URL)"

However, even some others here would agree me, there would be ever and ever others who remove such statements.

Kind regards --46.255.171.231 11:37, 24 December 2014 (UTC)

 Support to include notable disproved statements. By the way, knowing that a claim was made and later disproved IS knowledge as well. Petr Matas 15:11, 24 December 2014 (UTC)
 Oppose Wikidata gives people access to true knowledge, we don't put theories that turned out to be false in our items, even if a deprecated rank is used. --AmaryllisGardener talk 15:22, 24 December 2014 (UTC)
There is no correct knowlege base, explanation: https://www.youtube.com/watch?v=gAtn-4fhuWA --FischX (talk) 13:02, 25 December 2014 (UTC)
 Oppose Wikidata is not ready for this. For instance, there's no way to form queries that take the rank into account. I also believe that adding statements that we know are false will hurt the credibility of Wikidata when the time comes to use it on Wikipedia. Pichpich (talk) 00:34, 25 December 2014 (UTC)
 Support All my experience with Wikipedia suggests this is absolutely necessary. If we don't include such statements and mark them false, others will only go on finding them on the internet and adding them in a less controlled way. It can also be highly necessary for matching to external sources -- it is valuable to know what ULAN says for an artist's date-of-birth/work period/date-of-death because this is widely used by other websites; even if eg RKD artists may have much more detailed precise information. Jheald (talk) 08:46, 25 December 2014 (UTC)
As Jheald says, some deprecated statements are useful, if only because it tells people that we know about this claim, but that it is false, and should not be added with a non-deprecated rank. @Pichpich, the time has already come to use data in Wikipedia, and deprecated statements are not really an issue, as they are filtered out by default. That said deprecatedness should be much more prominent in the Wikidata UI, as most readers will not notice it.--Zolo (talk) 09:12, 25 December 2014 (UTC)
I'd prefer some way to filter out deprecated statements in searches, etc, first. I do think that they should be included eventually though. Ajraddatz (talk) 17:45, 25 December 2014 (UTC)
 Oppose We are collecting knowledge, not hoaxes and errors. It is possible to mention things like that in article on wiki, but i do not think that it should be on wikidata. --Jklamo (talk) 18:06, 25 December 2014 (UTC)
 Comment We probably need some sort of notability policy for statements that are more likely to be incorrect. I think that obvious errata should not be included. Wikidata can't include every conspiracy theory that is trending on Twitter for a few days. --Tobias1984 (talk) 09:32, 26 December 2014 (UTC)
Indeed. Give me a subscription to a dozen tabloids and I can spend all day adding rumoured facts about celebrities and marking them deprecated along the way. We clearly want to avoid this. Pichpich (talk) 22:22, 26 December 2014 (UTC)
 Oppose We are storing information for the use of Wikipedia, which requires accurate information. Hoaxes could lead to troubles with en:WP:BLP, just for example.--Jasper Deng (talk) 07:09, 27 December 2014 (UTC)
 Comment how do you know what is true and what is not? Eg. statements on Jesus (Q302)... Do we even care about truth? I think we should only care about references and reliability. Cdlt, VIGNERON (talk) 11:38, 29 December 2014 (UTC)
 Comment Brilliant question. I am all for verifiability and sources, and less for "The Truth". There might be sensitive cases, where we might be a bit more stringent, though, especially in the data about living people (or, reasonably alive people). Furthermore, in this case, any respectable source that has fallen prey to the Axl Rose Death Hoax has, I assume, issued a correction, and I don't think we can find any good source which is not retracted, or do we? Other than that, I think this was pretty much the right approach: statement was marked deprecated, sources were given. It is just that the sources were retracted and thus in this case it does not make much sense to keep it, but yes, it is clearly a borderline case, and a good test. --Denny (talk) 18:08, 30 December 2014 (UTC)

Asking for permission from external sites to import data

The French statistics bureau has recently released new census data. If we could import them to Wikidata, that would be a nice occasion to swich from ad hoc local templates to a nice Wikidata format. The trouble is that the data do not appear to be incompatible with CC0, though they seem to be compatible with Wikipedia's license. I understand that questionning the choice of CC0 is complete taboo, but is there any guidance on how we can ask the institute for permission to use their data ? --Zolo (talk) 07:53, 30 December 2014 (UTC)

We need a system like in commons: [27]. The way to ask the permission is not the role of wikidata, but the implementation of a system to check the rights yes. For the permission, better try to use the french wikimedia chapter and to write an official letter to the administration or the Ministre. But definitively CC0 is the most important obstacle in importing data in Wikidata. Snipre (talk) 21:12, 30 December 2014 (UTC)
From what I know of CC0, they would need to release all rights to the data and thus be opening it to use by anyone without attribution - I doubt they would sign on for that. It would be nice if we could included some data under different licenses as well, though that might not be ideal either. Ajraddatz (talk) 02:43, 31 December 2014 (UTC)
I do not think they need to formally release their data under CC0, which they seem unlikely to do on a large scale at the moment. But I think we can get a permission for using some of their data even if they do formally release the full database under a suitable license.
We ought to be clearer on what exactly we need before uploading data. For instance here, the guy in charge of open data at the French National Library said that we could use his data, and he was backed by user:Remi Mathis who works at the library and is the Presisdent of Wikimedia France. That was originally considered ok, and then it was said here, it was said that we really need a more formal authorization. Public open data websites are supposed to work for the public good, and advertise a lot on how their data are made to be reusable. I suspect many non-profit websites would be rather pleased if we used some of their data. But at the same time, they may not want to shout: "we have a CC-attribution license, but there are always exceptions".
Otherwise, yes, doing something with Wikimedia chapters might be a good idea.--Zolo (talk) 16:29, 31 December 2014 (UTC)

de - en

Please connect de:Kloster Allerheiligen (Schweiz) and en:Kloster Allerheiligen, Schaffhausen --Gerda Arendt (talk) 20:59, 31 December 2014 (UTC)

✓ Done--Ymblanter (talk) 22:07, 31 December 2014 (UTC)
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat/Archive/2014/12&oldid=1102551278"