Wikidata

Wikidata:Project chat/Archive/2013/04

< Wikidata:Project chat‎ | Archive‎ | 2013
This page is an archive. Please do not modify it. Use the current page, even to continue an old discussion.

Use of P131

I was just reading the various statements for Germany (Q183), and found the country is in the administrative unit (Property:P131) Central Europe (Q27509). To me that sounds utter nonsens as Central Europe isn't an administrative unit.

However it is true that Germany is in the geographic region Central Europe. So I wonder if P131 is wrong worded and also includes geographic regions, or if the statement should be deleted, or if the property should be replaced by another one. Byrial (talk) 20:32, 28 March 2013 (UTC)

It should be deleted or replaced by another statement. IMHO the alias "region" in English is misleading. --  Docu  at 20:35, 28 March 2013 (UTC)
Aggree, P131 should include administrative items only. RobotMichiel1972 (talk) 22:29, 28 March 2013 (UTC)
At least this is what it was designed for.--Ymblanter (talk) 11:07, 1 April 2013 (UTC)

Username block

Do we have any policy regarding users being blocked for their usernames? I'm referring specifically to User:Fapper69420, which seems a bit offensive. FrigidNinja 15:08, 30 March 2013 (UTC)

If you start to block names which may be offensive to someone, you may have to block many serious contributors as different things are offensive in different languages. I have no idea what "fapper" means to you. The word is not in en.wiktionary. I think it is better to block based on actions. Byrial (talk) 18:51, 30 March 2013 (UTC)
"Fap" is slang for "masturbate", and a "fapper" is one who masturbates. FrigidNinja 20:07, 30 March 2013 (UTC)
Considering his sole edit is him liking pp in his bum bum, it's hard to assume good faith based on the username. Blocking.. Regards, — Moe Epsilon 23:44, 30 March 2013 (UTC)
We can always start with http://commons.wikimedia.org/wiki/Commons:Username_policy since it's multilingual and then edit it as needed. Regards, — Moe Epsilon 00:48, 31 March 2013 (UTC)
This will be a policy that would be very difficult to use, no matter how it is interpreted. — Jeblad 01:28, 31 March 2013 (UTC)
I don't care about a user's username unless his edits aren't disruptive. I'm strongly against 0 edit blocks, just because of the username (except it is a cross-wiki vandal; then it should be reported to stewards and be locked then). Regards, Vogone talk 01:46, 31 March 2013 (UTC)
I think we need policies, when there are conflicts - not earlier :) . Greetings, Conny (talk) 13:19, 1 April 2013 (UTC).

Blocking bots: Houston we have a problem…

Due to excessive lag we will have to turn off (block) all bots. After a grace period bots will be slowly turned back on. Please add the name of your bot below together with a reason why it should be exempt from the blocking. This message can be removed in 24 hours. Thank you for your understanding. — Jeblad 00:07, 1 April 2013 (UTC)

Skynet - vital to the security of the world :) FrigidNinja 01:10, 1 April 2013 (UTC)
I guess it was to obvious, it is 1th of April, but we actually has a problem with getting the changes fast enough out to the clients. It will probably be solved in short time, without turning the bots off! :D — Jeblad 11:11, 1 April 2013 (UTC)
Yeah, those developers really need to solve that problem with getting the changes to the clients. ;)--Snaevar (talk) 11:32, 1 April 2013 (UTC)

User don't stay logged in

I've seen few notices on Finnish Wikipedia, that some users don't stay logged in when they come from Wikipedia to Wikidata. They always need to log in manually here, But when they go from Wikipedia to Commons or other projects, they stay logged in, so what could be the problem? --Stryn (talk) 07:38, 1 April 2013 (UTC)

The most likely reason is that the browser is rejecting an cookie from wikidata.org. Check whether the browser accepts cookies from third parties. Also either the option of accepting cookies from websites needs to be ticked, or there needs to be an exception that allows wikidata to set cookies.--Snaevar (talk) 11:22, 1 April 2013 (UTC)
The pages where you leave fiwiki, does they use a link ala d:Q1? If so does they link to //fi.wikidata.org/wiki/Q1? The url should be //www.wikidata.org/wiki/Q1, if not try to purge the page and check whats happen when you then try to traverse the link. — Jeblad 11:29, 1 April 2013 (UTC)

CommonsDe/Relinker

Tracked in Bugzilla
Bug 46358

Do we have any solutions for moved and deleted files yet? -- Lavallen (block) 09:18, 1 April 2013 (UTC)

This known problem. --EugeneZelenko (talk) 13:36, 1 April 2013 (UTC)

Lousy interwik link(s)

In English Wikipedia, the en:Rifle (disambiguation) page links to the French disambig page fr:rifle. But the French for rifle is fusil. The French disambig page for fr:fusil links to the English disambig page en:Fusil, but fusil is not the English for the French fusil (rifle)! Someone has clearly created interwiki links thinking that words spelt similarly in each language have the same meaning - they don't. (And then there are the Catalan, Polish and Spanish links to consider as well). I don't normally moan about these things - I put them right - but this is ineditable from WIkipedia. Emeraude (talk) 17:36, 1 April 2013 (UTC)

I think this is exactly what we want. A disambiguation page lists pages whose title basically matches a certain sequence of letters. This is even clearer on fr.wiki where disambiguation pages are called page d’homonymie. For instance, en:Rifle (disambiguation) has an entry for en:Rifle, Colorado which has nothing to do with firearms. If fr.wiki creates an article for Rifle, Colorado, it will be listed on fr:rifle and not on fr:fusil. So these interwiki links are correct. Pichpich (talk) 18:10, 1 April 2013 (UTC)

Make MediaWiki:Gadget-AuthorityControl.js a default gadget ?

Hi all. Many statements about external IDs, like "VIAF ID: XXX" have been created. However, those properties currently just look like basic strings. That looks rather cryptic, and might even scare away newbies. MediaWiki:Gadget-AuthorityControl.js adds an external link to them, making their usefulness much more obvious. Can I make it a default gadget ? --Zolo (talk) 07:56, 26 March 2013 (UTC)

  Support (obviously) --Ricordisamoa 10:18, 26 March 2013 (UTC)
  Support — Ayack (talk) 10:59, 26 March 2013 (UTC)
  Support Looks good to me. Ajraddatz (Talk) 14:08, 26 March 2013 (UTC)
Definitely. --Izno (talk) 14:24, 26 March 2013 (UTC)
  Support — ΛΧΣ21 04:08, 27 March 2013 (UTC)
  Support ···日本穣? · 投稿 · Talk to Nihonjoe · Join WP Japan on enwiki! 16:05, 28 March 2013 (UTC)
  Done --Zolo (talk) 15:10, 2 April 2013 (UTC)


Additional question: How should it be named ?

  • AuthorityControl ?
  Support to keep it consistent with usage elsewhere. ···日本穣? · 投稿 · Talk to Nihonjoe · Join WP Japan on enwiki! 16:05, 28 March 2013 (UTC)
  • ExternalLinker ?

  Support, as a more clear and general name (and actually, the script adds an external link, it does not add an authority control) --Zolo (talk) 09:30, 27 March 2013 (UTC)

I think handling of this property is wrong, it should be an identifier passed on as a value in an iw-link and turned into a link accordingly. This should be done by the core software, not by a gadget – even if the gadget is a stop-gap measure. — Jeblad 11:26, 28 March 2013 (UTC)
For many properties, it would probably be better to have something resembling sitelinks (having a source for them do not necessarilty make sense either). Isn't that on the plans for Wikidata's future ? --Zolo (talk) 15:47, 28 March 2013 (UTC)

Spaghetti of gadgets

I wonder if we should have a stricter policy on gadgets, it is to easy to slap together some js-code and turn it on for everybody. If some functionality is considered so important that it should be turned on for everybody, then it should probably go into the extension and have proper tests and go through a review that is not merely a popularity measure. That means more work for those that write gadgets, but I think the quality of the code will increase.

Perhaps a policy like: "Gadgets should only be used for non-essential functionality the user can turn on herself, and for functionality that is unrelated to the Wikibase extension. Any essential core functionality that is turned on for everybody should be available as part of the extension and adhere to all necessary processes."

It would mean follow the defined process, write a bug in bugzilla about what you would like to do, create a [http:gerrit.wikimedia.org patchset in gerrit] with the necessary code, and make the necessary changes so your code can be merged. — Jeblad 12:01, 29 March 2013 (UTC)

Well... while RequestDeletion and DeletionHelper are properly for usage on Wikidata, I think that move.js would be more appropriate in the Wikibase extension. --Ricordisamoa 12:53, 29 March 2013 (UTC)
Nope. Gadgets allow the community to get what it wants (or needs) now. Making a bug means the community waits for that functionality to be implemented.
Ideally, both are done, where the community thinks there should be an improvement in the core functionality. Or perhaps the developers could check on a regular basis to see if there is something turned on for everyone. If and when the functionality is implemented in an extension or in core, that same functionality can be turned off in the interface.
Unless there's a particular gadget causing you or the other devs grief? In which case, you're welcome to bring that up for discussion, or perhaps even have the devs unilaterally remove the gadget. --Izno (talk) 13:20, 29 March 2013 (UTC)
This isn't about writing a bug and then sitting down waiting for the extension devs to do the coding for you, they won't, its about setting standards and asking the gadget developers to do the coding the right way. If it was up to me I would say that no gadgets should be turned on that hasn't been through a review, including a security review, they should have proper documentation, and they should all have proper tests. I would even say that the special page should include tests that should pass before the gadgets could be turned on at all. — Jeblad 18:00, 29 March 2013 (UTC)
Oh, see, you didn't say that. And no, I completely disagree still. Especially, how do you produce the adoption necessary to include these as either a) core functionality or b) as implemented gadgets? Your way... that doesn't happen, because we're then back to "oh, here's a random javascript page that may or may not be updated in a vacuum, etc." symbolic on en.wikipedia. Additionally, users may not want to submit to the non-SUL environment of the repository and in fact, the submission of patches in Git or any other source control.
If you see a problem with a gadget, as with all things wiki, you are free to provide your feedback (or in the case of code, even edit it yourself if it's that big an issue, though I'm pretty sure you don't have admin privileges?). Otherwise, this hampers the deployment of functionality to the users necessary to experiment and use the code. --Izno (talk) 18:48, 29 March 2013 (UTC)
One of the potential benefits of gadgets may be that we can dispose of them once they are no longer useful. For instance, I would say that we can get rid of "sluprinterwikis" now that we have almost every Wikipedia article here (but I know removing things people have got used to is not always easy)--Zolo (talk) 14:10, 2 April 2013 (UTC)
We are very far from having every article with interwikilinks here. Wikidata:Wiki import task force--Saehrimnir (talk) 15:18, 2 April 2013 (UTC)
I do not think the remaining items would need slurpinterwikis, as they either do not have many sitelinks or have incorrect one. There may be non-main space pages where that could still be useful, but I was making a general point rather than a specific proposal.

Widget to add language links on the Wikipedias directly

Apperently from Status Report #51:

  • Widget to add language links on the Wikipedias directly: added setting to enable/disable it per wiki and made it available for logged-in users only
  • Widget to add language links on the Wikipedias directly: improved layout / size

This is live now but the Link under the Languaglings still directs me to wikidata and no widged opens. When will it be live for all users also the not logged in. This is the most important feature of Phase one which because only this really makes of Interlanguagelinks not more complicated than before the transition (if it can also make suggestions for weather Articles in other Languages exist based on translation or same title that would be awesome).--Saehrimnir (talk) 17:58, 31 March 2013 (UTC)

If something is in a status report it means someone has done some work on it. It does not mean it is in production, but you would usually see something at the test repo or client the following week. — Jeblad 21:27, 31 March 2013 (UTC)
Yes thanks but what puzzled me was the made available to logged in users part and that it did not appear in this weeks report.--Saehrimnir (talk) 04:22, 1 April 2013 (UTC)
It shouldn't yet be available to logged in users either. Is it? --Lydia Pintscher (WMDE) (talk) 12:11, 2 April 2013 (UTC)
No it is not but I thought it should.--Saehrimnir (talk) 13:56, 2 April 2013 (UTC)

Strategy for Phase 2

Described below is a strategy how Wikidata can provide easier correctness, completeness and up-to-date data on Wikipedia's.

Wikidata is easier maintained because it is a central database. This might getting used to, as the interface is different from Wikipedia and stored on on sister project, but remember this is stored 1 time for over 200 Wikipedias.

Correctness is better guaranteed as on the Wikipedia-way where every Wikipedia can have different values. Outdated values and mistakes are prevented by using Wikidata. Wikidata can still cater for different Point-of-Views so deliberate different values on different Wikipedias are still possible.

Wikidata combines all data for all Wikipedias so Completeness is higher than on any single Wikipedia.

Up-to-date data: An update on Wikidata could have immediate effect on each Wikipedia, provided the Infoboxes are set to do so. For example the Date and Place of death of a person or a change of the Mayor of a city.


Strategy for Phase 2

1. Bots are currently updating Wikidata from Wikipedias.

2. Update templates on Wikipedias. Wikidata should only fill missing information, this will ensure a smooth transition. This can be achieved by using Template:Wikidata.

3. Compare Wikipedia data with Wikidata data. If data is not equal: Check if Wikidata can be improved. Template:Wikidata could be expanded to compare Wikipedia data with Wikidata data.

4. Volatile data: Volatile data is data that is changing over time, like population numbers, the current person holding a position (like Mayor). These would be the first to be maintained centrally via Wikidata. If Wikipedia data equals Wikidata data: Remove the Wikipedia parameter, but only if you are confident, and in a few months time.

5. Static data: Static data is data that does not change over time, like the capitol of a country, the birth date of a person, the taxonomic tree. If Wikipedia data equals Wikidata data: Remove the Wikipedia parameter, but only if you are confident, and in a few months time.

6. Update templates on Wikipedias in a way that it can be used completely without parameters. The parameters should remain, but should give a clean Infobox, even if Wikidata does not contain information. In this way Infoboxes without parameters can be added to stubs or any other article.

I hope this makes sense. HenkvD (talk) 18:28, 1 April 2013 (UTC)

Comment. If we store volatile data here and set infoboxes on all Wikipedias to change based on that data, then one vandal could affect dozens or hundreds of wikis by vandalizing the data stored here. Jakob Megaphone, Telescope 18:58, 1 April 2013 (UTC)
Re: "the Mayor of a city". We normally do not have articles on any wikipedia for the majority of the mayors in our cities/regions world wide. How are we going to solve such issues? The property for mayor today only accept items. What happens when (s)he is replaced with somebody who has no item. Will properties in the future be flexible to accept both strings and items? Do we have to create items without sitelinks for persons who, most likely, never will have any article on any Wikipedia? -- Lavallen (block) 19:16, 1 April 2013 (UTC)
In my opinion moving data from Wikipedia to Wikidata is a suboptimal solution, Wikidata should capture and store data from the real sources. The templates at the individual Wikipedia projects should then be changed to use values from the items in Wikidata if they exist and is not overridden by locally set values. Slowly the local infoboxes will shrink until the call will become nothing more than {{infobox}}, or possibly some template type depending on the context. — Jeblad 20:09, 1 April 2013 (UTC)
Central repository doesn't ensure correctness, Wikidata concept isn't a solution against error: only a strong discipline based on references can ensure that characteristic and right now with the present wikipedia templates harvesting I am not sure that we are going in the right direction. Wikipedia are not the best data sources, we have to find primary sources for data, sources which are complete or offer a better overview of their relevant field than addition of data from different sources. Honestly we need task forces defining good data sources and working with bots to extract those data when possible and to fill wikidata. If data extraction can't be made by bots an manual extraction is necessary organized by the task force.
Don't focus on wikipedias and how they will use data, it is not the main task of wikidata: we inform wikipedia of how they can use data from wikidata and they will organize themselves for including data. They will do only if wikidata can provide good data so we have to focus only on that.
Then we have to solve the problem of data for property based on item: do we want to create an item for those data even when no article in wikipedia exists ? Snipre (talk) 22:40, 1 April 2013 (UTC)

Hazard-Bot

Hello. For quite some time, since seeing the number of deletion requests my bot has at times (recent example), and especially even when Legobot reports some of them from before, I think it might be useful to have Hazard-Bot as an adminbot to delete pages that it has blanked, rather than requesting deletion of them. I've had this opinion since during the bot's trial, but I had decided to wait for two main reasons:

  1. I doubt the community would want a temporary administrator to have an "indefinite" adminbot account
  2. I wanted to improve the accuracy of the bot to minimize false positives as greatly as possible.

I believe these have been solved:

  1. The community recently had my adminship reconfirmed (thank you)
  2. For some time now, I've rewritten the code for increased accuracy and efficiency, including the addition of other features, and it has been running fairly well as far as I know. The only issue I've been aware of recently was one with reporting the items on RfD, though this has nothing to do directly with the deletions or otherwise removals of the bot. However, as shown in the example above, this has been fixed. As we currently have no agreed process of requesting adminbots, I've decided to ask for some community input here, the possibly, based on that, put the bot through an RfA (or use the decision from the discussion here). Thank you. — Hazard-SJ  ✈  04:13, 26 March 2013 (UTC)
  •   Support Wikidata is a technical Wiki so needs bot a lot and needs adminbots as well, besides he and his bot has done good works Amir (talk) 04:52, 26 March 2013 (UTC)
  • AFAIK, before an admin deletes an item, he/she checks the history for vandalism. If two items were merged, admins also move labels/descriptions/aliases, at least in major languages. If this is correct, will your bot do the same? --Michgrig (talk) 07:55, 26 March 2013 (UTC)
    • Actually, no. My bot does not merge items (that's not a part of any of it's tasks, and it's more suited to be done by a human or semi-automatically, not fully automatically). The purpose of it having admin rights would be to delete pages it blanked, because it has no sitelinks left (the one it removed would have been one for a page deleted from a Wikipedia, so it should fail WD:N). — Hazard-SJ  ✈  02:57, 27 March 2013 (UTC)
  • I can support for items, which are non-notable entities (pages deleted on Wikipedia). But other items needs checking. --Stryn (talk) 09:30, 26 March 2013 (UTC)
    • Could you please be more specific about what needs checking? The bot updates sitelinks of moved pages before starting to process the deletions (which it uses the API to get, in a way such as to exclude undeletions etc.), and before attempting to remove the sitelink, it checks to ensure the page still does not exist on the Wikipedia first. Only if only one sitelink was on the page, it would be deleted. — Hazard-SJ  ✈  02:57, 27 March 2013 (UTC)
      • Oh, looks like I was wrong. I checked deletion requests made by your bot, and I didn't saw any mistakes. It was Legobot, whose some requests have not been deleted. So I Support. --Stryn (talk) 07:51, 27 March 2013 (UTC)
  •   Support I did check the archive of deletion requests from now to 24th of March and as far as I can tell all of the pages HazardBot nominated for deletion where deleted in that timeframe. With that in mind, I do support this request.--Snaevar (talk) 10:38, 26 March 2013 (UTC)
  •   Support only deleting the items which were deleted from the only Wikipedia, and only after checking that the item was indeed deleted, not moved with the redirect suppression.--Ymblanter (talk) 12:34, 26 March 2013 (UTC)
    • That code has two main parts: processing page moves, and processing page deletions. It first goes through the move log for the wiki before it goes through the deletion log, so if it is a move, it should have already been corrected. Also, yes, only if the only sitelink is removed it is requested for deletion, so only in such a case it would delete. — Hazard-SJ  ✈  02:57, 27 March 2013 (UTC)
  •   Support Per Ymblanter. Ajraddatz (Talk) 14:27, 26 March 2013 (UTC)
  •   Support — ΛΧΣ21 04:07, 27 March 2013 (UTC)
  •   Support but it must check if some other sitelinks have been removed without a reason. --Ricordisamoa 08:16, 27 March 2013 (UTC)
    "Without a reason"? If I'm thinking what you're thinking, "reason" would be "edit summary", which, for items, are software-generated. — Hazard-SJ  ✈  05:23, 28 March 2013 (UTC)
    No, not edit summary; in many cases, a sitelink has been removed by anonymous users because the page has been deleted; else, if the page still exists, and doesn't have an associated Wikidata entry, Hazard-Bot should not delete the item, but maybe revert the removal and/or log a warning. --Ricordisamoa 06:09, 28 March 2013 (UTC)
    Unfortunately, your argument is still unclear to me. The Wikidata entry is the item. Did you mean Wikipedia? If so, my bot would only delete the pages it blanks, not what other users blank. Does my clarification help? If not, please rephrase. — Hazard-SJ  ✈  19:49, 28 March 2013 (UTC)
    If no other sitelinks have been removed, then it could delete the item. --Ricordisamoa 19:51, 28 March 2013 (UTC)
  •   Support in my opinion it can also deletes when item is blanked and it has only one user contributor or the first user and the last user are the same (creator blanked the item or made item as blank). it can check this case one day later to let the user to complete his/her item Yamaha5 (talk) 15:13, 27 March 2013 (UTC)
    Generally speaking, blank items don't usually hang around for so long, and things such as the default notability criteria would complicate such a task either way. It might be better to let admins continue doing this manually. — Hazard-SJ  ✈  05:23, 28 March 2013 (UTC)
  •   Oppose I do not think automatic deletion is a good idea. This was part of one of the early API modules, but it was removed because there was no simple way to figure out why an item was emptied. The argument was that "some user by intent or accidently empties an item for content, and then the item is automatically removed – this effectively gives the user admin rights". I think the argument is still valid. [That said, I do think we need a good way to find items that are "empty" in some way, including but not limited to those without sitelinks.] — Jeblad 11:37, 28 March 2013 (UTC)
    • Actually, this request is for deleting pages for which my bot has removed the last link from, after checking that that page has been deleted from the language's Wikipedia. That is to say, items which fail WD:N by not having any sitelinks, considering that my bot removed the only remaining sitelink. — Hazard-SJ  ✈  19:41, 28 March 2013 (UTC)
    What if I remove all (legitimate) sitelinks from an item, and add an alibi sitelink to a user page (or an article to be deleted, or ...) afterwards? Will the bot remove this, and delete the item? --YMS (talk) 19:52, 28 March 2013 (UTC)
  •   Oppose See Jeblad: This would effectively give every user, including vandals and newbies, the right to delete any item. --YMS (talk) 11:55, 28 March 2013 (UTC)
  •   Support As far as I can see, Hazardbot would only be deleting items where the bot has removed a deleted article from the sitelinks list and has in the process left the item empty. I have yet to see any deletion request that Hazardbot has proposed that has not needed to be deleted by an admin. It would really help reduce the workload at RfD. Del♉sion23 (talk) 18:29, 28 March 2013 (UTC)
  •   Support I see no reason not to. Go ahead and open the RFA. :-) Regards, Vogone talk 19:46, 28 March 2013 (UTC)
  •   Comment Even though pages deleted in Wikipedias and there is no sitelink any more, it doesn't necessarily mean the item fails WD:N because it can still meets the default notability criteria (this situation is rare but still possible). So I think the bot should also check whether the item is an orphan or not, and only delete when it does. --Stevenliuyi (talk) 20:03, 28 March 2013 (UTC)
    Yes, I'll add some code to check for this on the Wikipedias soon. Even if the bot doesn't pass it's RfA (which I'm about to file), it will check before reporting to WD:RFD. — Hazard-SJ  ✈  23:00, 28 March 2013 (UTC)
  •   Support Javad|Talk (10 Farvardin 1392) 15:25, 30 March 2013 (UTC)
  • I would oppose this, per the edit that Merllwbot made that spurred me to investigate a particular bot's behavior. There are a number of instances where the bot made the wrong edit. Can HazardBot show that such a problem would not be an issue? --Izno (talk) 15:35, 30 March 2013 (UTC)
    I believe that for the most part it has been. Also, I have a few fixes planned which would possibly exclude other possibilities. I'm not sure what that bot was doing, but I'm more able to speak about my bot :D — Hazard-SJ  ✈  05:13, 3 April 2013 (UTC)

Featured Items

Should we have "Featured Items" (like Featured Articles on WP and Featured Picture on Commons)?

Items with complete sitelinks, descriptions, statements and references, could become   Featured Items.

Opinions? --Ricordisamoa 21:21, 28 March 2013 (UTC)

For this purpose we could also just make one or two example items, which we can link somewhere at the start or help page. It isn't necessary to have hundred of examples. --Nightwish62 (talk) 08:16, 29 March 2013 (UTC)
  •   Oppose this name in favor of good items. In my opinion, we can't make items brilliant in quality - we can only make them good; there isn't much room to grow.--Jasper Deng (talk) 22:38, 28 March 2013 (UTC)
  •   Support in principle – it may be too early. I doubt any item would pass the mark at this stage in time, especially as we haven't yet got a fully implemented version of Wikidata in order to add references and various other data that would be needed in order to pass strict criteria. The best we can hope for right now is an "ok item". Del♉sion23 (talk) 22:47, 28 March 2013 (UTC)
  • Eventually this makes sense, once phase III is online, and we can hold up certain items as "how things should be done" but until then, there isn't much of a way to judge yet. "Model items" might be another idea of what to call the idea. Courcelles (talk) 23:01, 28 March 2013 (UTC)
    • I like the idea of model items better, at least as long as not all planned features have been implemented. Having such model items could help standardization in terms of expected properties for certain types of items, and serve as references during discussions of proposed new properties. -- Daniel Mietchen - WiR/OS (talk) 01:42, 30 March 2013 (UTC)
  • This may be a good idea in the future, but with most of the features of Wikidata not completed yet, this would result in an ever-changing standard. --Rschen7754 23:11, 28 March 2013 (UTC)
  •   Oppose --an item can't really be brilliant or engaging. And as for good items, don't most items have sitelinks from the beginning? (At least I add all the Wikipedia links when I create a new item). Jakob Megaphone, Telescope 23:26, 28 March 2013 (UTC)
  •   I Support the idea of having some assesment process to identify our best work. be it featured, or good, or whatever, that will encourage users to work beyond the label and (sometimes) descriptions. If we want to be a database, we need to encourage improvement of items. Maybe not now, mostly because phase 2 and 3 are still not deployed, but this will be useful for later. — ΛΧΣ21 00:47, 29 March 2013 (UTC)
  •   I Support I think it would be a great way to show new editors how to help effectively. --Macadamia1472 (talk) 01:54, 29 March 2013 (UTC)
  •   Oppose as a list of links (which basically all everything here is) can't really be made great. It's just a list of links with some brief identifying information at the top to help people who stumble across the unuseful item title know they are at the right place. What exactly is there to source? All that's happening is connecting one Wikimedia site to another. Basic. Let's not complicate things unnecessarily. ···日本穣? · 投稿 · Talk to Nihonjoe · Join WP Japan on enwiki! 04:56, 29 March 2013 (UTC)
    Sourcing will be functional at some point, and they are for the properties. Sourcing is needed for any claim. We can't simply say X is Y without something to prove it. Regards, — Moe Epsilon 06:14, 29 March 2013 (UTC)
    That's what the original articles are for. It's extra and completely unnecessary work to have to source every interwiki link. People complain about having to watch "another million pages for vandalism" (above), yet this would be far more work than allowing people to add userpage interwikis here. ···日本穣? · 投稿 · Talk to Nihonjoe · Join WP Japan on enwiki! 16:07, 29 March 2013 (UTC)
    There's more to Wikidata than an interwiki holding pen. The claims per phase 2 are and will be sourced at some point, and those will need sources. --Izno (talk) 18:50, 29 March 2013 (UTC)
  •   Oppose. I think many people doesn't understand what a database really means and did never work with a real database like MySQL or MSSQL. Database means a collection of raw data, full stop. It's a serious matter. What we need in Wikidata are data collectors, which isn't really an exciting task to do. There is also non reason why non-contributors should visit Wikidata at all and jump from one featured items to another. It isn't something like an article where you can pass time and enjoy by reading featured articles one after another. Moreover, starting with 'featured items' would encourage for other non-data-relevant-meta-stuff and at the end we didn't only have 'featured items', but also stuff like 'items stubs', 'NPOV items', 'item of the day' and so on. I really like the clean and tidy interface Wikidata offers at moment. There is no unnecessary stuff which distract me while collect data. Introduce things like listed above would only mess up this nice look and give non advantage in my opinion. Also, the discussion about honor an item as featured or not would only consume times which is better used by collecting data instead of just talking about. --Nightwish62 (talk) 08:11, 29 March 2013 (UTC)
    • In my opinion, it's important to encourage people to contribute to the reliability and authenticity of our information. Certainly if an item is missing lots of sources, but could have a lot of information, there's room to improve.--Jasper Deng (talk) 08:43, 29 March 2013 (UTC)

I didn't expect to get support, and I wasn't too much sure of it, either. :) --Ricordisamoa 09:01, 29 March 2013 (UTC)

I like the idea of measuring quality and encouraging improvement. A system based on the system used on the English Wikipedia could probably be used. Multilingual is important so the number of descriptions, sitelinks etc should be part. For a certain topic (for example a person or a building) if all relevant claims are present (each task force could maintain that). Because it is all about linking the quality of the linked items should be taken in account too. We should be thinking about such a system, but I think it's still too early to actually implement it. Multichill (talk) 12:39, 30 March 2013 (UTC)

I think the infoboxes should be part of the featured criteria. --Goldzahn (talk) 13:21, 30 March 2013 (UTC)

How to harvest data in the future

Harvesting data with bots, especially "a bot that is mine", is fun but it doesn't scale very well. If something happens, you are the only one that knows how to do that specific task. If that specific task is collecting population figures about a country, well then it can take years before anyone else picks up the task because nobody knows how to do it. To make the data collection sustainable and scalable we need some better way to do it than sitting in the dark with our own bots. It works now when a large amount of data needs to be moved from one very homogeneous source, but in the future when we need to maintain a large number of import solutions for heterogeneous sources it will not work. Anyone with ideas how to do this in a scalable way? — Jeblad 10:52, 2 April 2013 (UTC)

+1 BIG one: it need a well think toolkit/toolset. Using common languages (python, JS, php?), with proper modules:
  • listing : get list of all your targets (ex: enwiki [Italy + cities + above 100.000hab] → articles names list of 40 Italian cities above 100.000 ),
  • reading> scraper : get data from WD,
  • writing> seeder : push data into WD
The API (http://wikidata.org/w/api.php) is not enough. I will do my part by sharing my python scraper. :] --Yug (talk) 11:47, 2 April 2013 (UTC)
My understanding is that every task like this should start with the list of jobs: What and how should be done. (And what should not be done). Do we have such a list?--Ymblanter (talk) 11:55, 2 April 2013 (UTC)
We have to wait first on numeric and time datatype before doing any big work in data importation in order to avoid to rework data extraction. Then we have to have a clear idea of the properties we will need for data importation. As first approximation I would say we have to be able to provide the complete set of parameters for infoboxes. Then for the data import itself the best way is to allocate some bots to the different task forces and the work has to be defined and organized in a task force. Even if the bot operator is an expert in the field he wants to work it is better to first discuss once with other persons in order to have a complete overview of the import task. The correct procedure for me is the nexte one:
  1. creation of task forces according to topics/items
  2. definition of the minimal set of parameter for the different topics/items (this set can be extended later but we need to start with the list of most crucial parameters to see possible synergy between properties) (task force)
  3. a set of good references has to be selected (task force)
  4. a complete example including data, qualifiers and references has to be presented before any scale-up (task force + bot)
  5. importation by bots has to be organized by task force (task force+bot)
You can already prepare a system for bot request but if different persons asked different taks without any discussion before you will have to undo what you did because no preparation was made. Bots are not the most important part but properties definition and references selection yes.
Snipre (talk) 12:05, 2 April 2013 (UTC)
(edit conflict) Hi, I've been working on such a system and should be done by the end of the week. The basic concept is that you give it a list of pages (generator), which can be just about anything. Then you choose what you want to do, update description, set labels, add claims, etc. Then a script calculates the changes needed to be made, and pushes it into a public API for any bot to update onto the wiki. There are some safe guards like requiring an admin/trusted user to approve it. It also has the capability to remember requests for the future, so you could say "update latest_version parameter from infobox software" every month. Legoktm (talk) 12:08, 2 April 2013 (UTC)
Possible project page Wikidata:Scalable harversting
@Lego: Nice. Please as much as possible make autonomous "modules" as I noticed: listing, reading, writing. Indeed, one may already have his own list, or just willing to get data, not write as bots usually do, etc. --Yug (talk) 12:17, 2 April 2013 (UTC)
The project is online. I will provide my script soon, a short but working python scrapper. You are welcome to provide your modules as well.--Yug (talk) 13:12, 2 April 2013 (UTC)
I have a kind of very vague idea about a special page where you can upload some kind of list or table, and then this list or table is split into entries, and those entries are feed one by one into a Lua script that processes them. After processing the result is merged into an item and the diff is shown to the user. The user can then chose to save before he progress to the next entry. Because the Lua scripts will be saved for later use, and they are saved on this site, anyone can step in and do updates as long as there is some documentation that says where the script applies. The basic idea seems doable if the sources are well-structured. Examples are datasets like the one from Statistics Norway for population figures. — Jeblad 15:12, 2 April 2013 (UTC)
In PHP, there is also this toolkit that provide an easy to use interface to edit Wikidata. Tpt (talk) 19:07, 2 April 2013 (UTC)
My point is that this doesn't solve the scalability issue at all, it only keeps the current situation. We must get maintenance of values out of the individual computers of bot operators and into some form that can be reused by more users, otherwise we will create a system that scale with number of bot operators. That is not what we want because those are several orders more sparse than normal editors. We must somehow transform that knowledge into a form that other editors can use and reuse. — Jeblad 21:53, 2 April 2013 (UTC)
@Jeb: are you planing a web apps ? I created a script-based project (Wikidata:Scalable harversting toolkit) to answer our need, but I increasingly wonder if I misunderstood your project (I guess because I'am deeply diging into script creation recently). Yug (talk) 23:18, 2 April 2013 (UTC)

Project to watchs

Hello all, I added the following template to the top of the Project chat's page. The aims is to list all meta-page or projects we currently push forward. The meta-pages, which will ease the life of later users. There is a good dozen of project annonced on this Chat page, it may be good to keep them visible so people can jump in at will :) So go ahead, let's list your projects. --Yug (talk) 14:56, 2 April 2013 (UTC)

Yug, thanks for helping, but please don't try to reinvent the wheel. Multichill (talk) 18:15, 2 April 2013 (UTC)
Can you specify. I'am aware and unsatisfied by the main page which doesn't list or map the emerging projects. Yug (talk) 23:20, 2 April 2013 (UTC)
WikiData: Emerging projects and meta-pages [edit]

Items that have no English equivalents

Hello. I'm a newbie in Wikidata. I have a question: Is it possible to create a new item without any English label and description? It's because I really want to link some articles written in my first language to other Wikipedia projects' equivalents but none of them are English, so I don't know if I should "translate" these titles into English; in many cases, there's no way to translate them into English. For example there is a Chinese article (about a cake) that needs to be connected to another Korean article but there's no English vocabulary for that cake, so how can I deal with this? Thank you. Fondofgeography (talk) 16:36, 26 March 2013 (UTC)

Yes, you can create a new item without any English label and description and that is exactly what you should do.--Snaevar (talk) 17:06, 26 March 2013 (UTC)
Many thanks! Fondofgeography (talk) 04:08, 27 March 2013 (UTC)
And as a courtesy, you may add the English description that this is a cake, so that everyone can understand what the item is about.--Ymblanter (talk) 08:24, 27 March 2013 (UTC)
Why not add the Chinese, preferably with a transcription, on English Wiktionary with a "request for definition" or request that a Chinese entry be created? There are Chinese speakers there, some native speakers, who might be able and willing to research the item. Explanatory links and a talk page comment would help. I specifically mention transcription because many words without obvious exact corresponding English lexical terms are simply borrowed in the form of a simplified Latin spelling. English Wiktionary has standards for inclusion that require relatively little attestation. If there were an attestable English term, it would certainly simplify reference to the Chinese for many users. 72.225.236.92 09:25, 3 April 2013 (UTC)

Removing label together with interwikilink

Hi I am trying to resolve some Interwiki conflicts. My Problem is that if I split items (disambiguation pages) so I remove an interwikilink in one Item but the label stays behind and then I cant create the label in the new Item because the description is the same. I think if an Interwiki link is removed from an item that is done because its wrong that means in >>50% of the cases the label is also wrong so it should be removed automaticly. What do you think?--Saehrimnir (talk) 14:04, 2 April 2013 (UTC)

The label and description for language may be right, even if the article is wrong. But even if the label and/or description is wrong, it is better to correct them than remove them. Byrial (talk) 14:48, 3 April 2013 (UTC)

Specificity beyond items

Being strongly interested in aviation I have found myself leaning towards editing the items for many airlines, airports and actual aircraft. The problem I am having is that for most aircraft there is only one article on a series of aircraft models. For example, 'Airbus A320' covers the A318, A319, A320 and A321. Whilst this is technically correct it under represents the airline's fleet as the A320 family covers aircraft from 107 to 220 seats. Ultimately the reader of a article would expect data such as 'Boeing 737-900ER' for which it is extremely unlikely there will ever be an article for.

The way that seems intuitive to me is to create one item for the Boeing 737-900ER, one for the 737-700 etc. The problem with this, however, is that I can only link one of them to the Wikipedia article. I am therefore wondering if it is good 'etiquette' to have a range of items for individual models, each with the property 'sub-class of.' The only alternative I can see is to leave it as one item per family and use qualifiers to specify scope but I worry that this would lead to an absolute mess.

This problem also has the potential to apply to airports. When data-value statements finally arrive I imagine that properties such as 'runaway length' and 'runaway surface' will begin to appear. The only way I can imagine it being properly implemented is to create one item for each runaway (which won't have individual Wikipedia articles), apply the statements to those; then to the airport the property, 'has runaway' could be added.

As I am sure you understand this problem is going to get a lot worse when data-value statements come out and aircraft specifications start being added. Each model would have different specifications, which could be sorted by qualifiers, but as I said above I fear that would be messy.

Ultimately what I am asking is therefore whether it is stylistically correct to add items without articles in these contexts or whether we need some redirection system. --Macadamia1472 (talk) 04:39, 3 April 2013 (UTC)

  • How do I technically, from a WP-article, reach the properties of an item, that is not directly connected to the article? Q1022452 is an article about 2 urban areas. The most logical place to put the properties (population etc) of these urban areas is in the items related to each urban area Q3277661 and Q3277804. But how do I summon these properties in a wp-article related to Q1022452? -- Lavallen (block) 14:20, 3 April 2013 (UTC)

An "official site" property?

Before I even attempt to propose this, I'd like to ask here: could an official site property work? A number of infoboxes have a field for an "official site," related to a person (in a variety of fields), organization (from companies to governments), audiovisual or performing arts works, etc. And of course there's en:Template:Official website. Is it technically feasible to create such a property here? Would the URL of such a link be a "string," according to our terminology? thanks, O Wise Ones. Shawn in Montreal (talk) 14:04, 3 April 2013 (UTC)

Yup! See Wikidata:Property_proposal/Pending#Official_website_.2F_Offizielle_Homepage_.2F_Site_officiel_.2F_Sito_ufficiale. Legoktm (talk) 15:03, 3 April 2013 (UTC)
Great, well, it's in the works, I see. Shawn in Montreal (talk) 15:22, 3 April 2013 (UTC)

Fuzzybot

Moved from the Administrators Noticeboard--Ymblanter (talk) 09:05, 25 March 2013 (UTC)

On Help:Editing/de i found that Fuzzybot has replaced the German translation in some parts by an updated version in English [here] and some more parts in the next following actions. Can somebody stop Fuzzybot and undo the changes? It makes just no sense to translate the same section again and again because somebody changed a link or added a word.--Giftzwerg 88 (talk) 07:47, 25 March 2013 (UTC)

That's the translationadmin's fault (see page translation log). The best would be if you would go to their user talk pages and ask them always to check the box "Do not invalidate translations" in the translationadmin interface, if the changes are minor. Regards, Vogone talk 09:52, 25 March 2013 (UTC)
No, these changes was not a minor. But earlier this bot did not touch the translated text, but only tagged outdated strings so that they appear pink background (see first version of File:Translate manual - Translate example - 10. Outdated clicking.png). Now (since around February) bot behaves differently and this is very inconvenient.
Also (perhaps for the same or similar reason), version comparison for source textx is actually broken (see this example).
Maybe this related with migration into new translation interface ("TUX")... I'll ask Siebrand, and I expect him to comment here.--Kaganer (talk) 10:32, 25 March 2013 (UTC)
Mhmhm… I see. But why did they change the interface to TUX? I don't really like it. :-S Vogone talk 10:55, 25 March 2013 (UTC)
This crazy idea only ;) I don't know what may be really. Maybe Siebrand can open our eyes to? --Kaganer (talk) 13:34, 25 March 2013 (UTC)
But if somebody changed a link, you must update your translation in most cases. And if a word is added, you cannot be sure in advance that the translation may remain the same. Of course, if only a typo is corrected, this definitely does not influence other languages. But I'm sure that in other cases the translation definitely needs revision.
Besides, the translation memory suggests the old variant to you, doesn't it? So, you do not have to translate from scratch each time. --Michgrig (talk) 10:21, 25 March 2013 (UTC)
This is not related to TUX, it's Bug 44328 – Replace outdated translations with source text on translation pages that was requested by some Wikidata users... I had announced it on m:Meta_talk:Babylon#Bug_44328_.E2.80.93_Replace_outdated_translations_with_source_text_on_translation_pages. Personally, I didn't like this change at all; it was thrown in everyone's throat for Wikidata, though, so if not even Wikidata likes it I wonder what sense it makes. --Nemo 12:42, 25 March 2013 (UTC)
It must come to an end that translation admins do any things without agreeing with translators.
Sorry, breaking here. To be fair, the problem reported in this section is rather that some wiki translators got a behaviour in the extension to be changed, implying it was the will of the community, and translation administrators weren't warned of the change. Now it turns out that maybe the community doesn't even agree on that software change. --Nemo 11:19, 26 March 2013 (UTC)
And I have even the impession that every translation admin does what he thinks right. AFAIK there is even no talk page as central meeting place especially for translation admins and translators. If an issue happens you must find out which translation admin did the change or triggered FuzzyBot. I'd like to have a link to the translation in the sidebar of the Main Page as it is on Meta. I have a feeling that translators are the stepchildren here on Wikidata. The translation admins even prescribe what is allowed to translate and what not. Until now images were included in the translation tool. therefore I could make own images in my languages. Now they are removed from the translation tool so you can't replace the English original any more. If I include the image for my language version in the translation tool, the image will appear twice in the translated document, in English and in my language.
NB: About illustrations in the translated pages - this was my initiative with using {{LM}} to automated construct localised names for mediafiles (see documentation). Localized versions of the files will be connected automatically, as soon be uploads to the repository. This are worked at the Help:Alias (please compare /en and /fr). --Kaganer (talk) 14:12, 25 March 2013 (UTC)
And, when are informed all translators? One user decides to do that, something is changed, nobody knows why and nobody knows what will replace it. We need badly a talk page for the communication between translation admins and translators and there should be a real dialog between both user groups. It is prepostereous that translators have to gather the relevant information from everywhere. --Michawiki (talk)
Translators don't read noticeboards. The place for such information is the message documentation, i.e. the /qqq subpages of the translation units whose content is displayed while translating. Probably Kaganer thought it was obvious that one has to read the documentation of a template on its man page in order to use it, but it seems you think he was wrong. The solution is then to ask him to add such documentation (probably with a short template), then he'll tell you if he can do it or he needs help, etc. --Nemo 11:19, 26 March 2013 (UTC)
Excuse me for my voluntarism. But this is a wiki;) I expect that if someone has any questions, he asks. I follow this way, and others too. Here it is enough just to find out exactly who made ​​a strange change. Until now, there were no translators noticeboard - probably need to create it. Unfortunately, I cannot keep track of this English-language chat, so I will not write here. If will created a chat for translators only - I connect to it.--Kaganer (talk) 17:04, 26 March 2013 (UTC)
@Nemo: What do you think who reads and writes here and discusses with you? And there is no translation noticeboard until now, so how can you claim that translators don't read noticeboards? We need a point of discussion and not a documentation to solve current issues. Why you don't want that such a talk page doesn't exist? Are there any privileges of translation admins that translators are not allowed to know? What's your problem? I can just shake my head about such opinions like yours. --Michawiki (talk) 20:33, 28 March 2013 (UTC)
I think Kaganer already replied. --Nemo 20:12, 30 March 2013 (UTC)
As far as concerns the translation tool: It is even worse: The diff function doesn't work correctly. Left it shows the whole text instead the previous text for the corresponding messages. You haven't a proper control any more what the difference is. You must check the present translation to find the difference. May be there isn't any difference in the translation language then because there was any spelling error in the English text only. Sometimes I see even the translated text left instead of the previous English text. AFAIK the translation tool is reworked. I hope it will work better in future than now. --Michawiki (talk) 13:55, 25 March 2013 (UTC)
The broken diff is a recent bug, introduced by I don't know what. It never has been like that, it has worked like a charm for at least two years I think. Nobody reported it on bugzilla yet, and I've never seen it outside the WMF wikis. Please file a bug against mw:Extension:Translate and they'll try to get it to the relevant people... --Nemo 11:19, 26 March 2013 (UTC)
There is no filed bug about this, that's true. But I reported it in bug 44328, see comment 9. In comment 10 Niklas Laxström answered that this is a transitional issue because of internal changes. For me the wishful thinking of those who decided to file this bug is the cause for this issues with the translation tool. --Michawiki (talk) 11:38, 26 March 2013 (UTC)
Thanks for pointing it out; I remembered reading it somewhere but forgot where. Anyway, I spent last 40 min investigating and I was told that there is no way to fix it without updating the translation (which is a catch 22): removing !!FUZZY!! and re-adding it at least hides the diff (which is useless anyway). The problem is less big with the new interface because the huge diff is collapsed by default, so it doesn't bother you, unless you click it.
IMHO, you should ask a bot to fetch all the outdated translations looping through all the message groups and languages like [1], add !!FUZZY!! when missing, removing it everywhere, readding it where it was removed. --Nemo 12:07, 26 March 2013 (UTC)
This affair becomes more and more idiotic! Why the translations are set to untranslated? From my point of view there is no reason for that! Repair simply the translation tool and don't experiment with it! Now translators have to make even a backup copy to avoid doing the same work twice. The translation tool was used for long years but such issues there weren't. And only because somebody had a new idea. I should ask a bot! A bot is a program! I don't know who runs a bot. That's a task for a translation admin but it seems that just those translation admins are the cause for these issues. Instead of doing their work properly the translation admins discuss if they should be elected or if anybody can be admin if he want to be it. Where that leads to we see now. --Michawiki (talk) 01:30, 27 March 2013 (UTC)
I'm not a translation admin here so I don't know what to answer. Again: if you have a problem with something a translation admin did, I suggest to tell him. I see there are 23 here, probably not all of them were trained yet. --Nemo 20:12, 30 March 2013 (UTC)
That makes this affair even worse. That namely means that no translation admin discussed with us. 23 translation admins and we don't read even a word from them. I should contact a translation admin? I must find out which translation admin is responsible for this disaster, which of those 23. Shall I send a message to all 23 translation admins? You have strange ideas. Instead of helping to finish this disaster you stand up for this terrible situation. Therefore I suggested a talk page as central point for translation admins and translators. If they are not trained they should contact the translation admins of Meta or give up. I know at least three of them from Meta and translatewiki.net: Bjankuloski6, Jon Harald Søby and Amire80. --Michawiki (talk) 21:20, 30 March 2013 (UTC)
I correct myself: Kaganer is a translation admin and discussed with us. And it seems that he does good work. --Michawiki (talk) 21:40, 30 March 2013 (UTC)
I don't see a difference with the rest of the actions on a wiki: you check the history/logs and ask the responsible. As for the noticeboard, if you care so much about it, please stop talking and create it (Kaganer already agreed). I'm not "standing up for this terrible situation", I'm just pointing out that endlessly repeating your complaints on this page won't bring you anywhere. --Nemo 09:31, 31 March 2013 (UTC)

[2] No comment. --Giftzwerg 88 (talk) 00:01, 28 March 2013 (UTC)

No. THAT's "no comment". Whole Polish text gets erased and new English one is inserted. Disgraceful. Kaligula (talk) 00:31, 30 March 2013 (UTC)
Kaligula, that's an entirely different matter. On Meta, translation admins don't mark existing pages for translation without importing all old translations. You should tell the one who marked that page to copy all the previous translations (but you can do it yourself, too). --Nemo 20:12, 30 March 2013 (UTC)
That's the issue here in Wikidata. You can't see the difference between old version and new version any more. The translated document is in English as well now. That's the greatest rubbish I've ever seen. The way in Meta is the the correct way and in Wikidata it worked also well before somebody changed anything. Since then nothing works correctly. There are 23 translation admins and nevertheless the translation tool doesn't work properly. When will you realize that the work of translators is important and should be properly supported? --Michawiki (talk) 21:00, 30 March 2013 (UTC)
Michawiki, I share your feelings but frankly your ranting is making this discussion fairly useless. --Nemo 09:31, 31 March 2013 (UTC)
I think the best way would be (instead of wiping translated text) to put info on top of translated page that it's content may be inaccurate and out-of-date; also this should include link to the diff of en-page making it easy for translators to check what's changed (a diff which compares old version to current version, what makes the diff up-to-date every time the source gets changed). Boom! and that's my opinion. Kaligula (talk) 15:41, 1 April 2013 (UTC)
Okay, now I see there's a transate tool. Weird that nobody said. Nvm, I translated it anyway. Kaligula (talk) 17:22, 1 April 2013 (UTC)

I try to make a bot according to the instructions of Nemo. I hope to run it next week. --β16 - (talk) 21:23, 3 April 2013 (UTC)

Proposal to make Help:Description an official guideline

Seems stable enough at this point.

  Support Kaldari (talk) 08:09, 2 April 2013 (UTC)
The label, alias and description are language specific help pages and should not be official guidelines unless we make them translatable, move them to the project namespace and make them project-wide. I would rather keep them as help pages until we know enough to make them project-wide, and not make a bunch of pages with disparate and language-specific rules. — Jeblad 10:42, 2 April 2013 (UTC)
The label, alias and description are language specific and therefore we should not make them translatable. --β16 - (talk) 21:36, 2 April 2013 (UTC)
The actual field set are language specific, but the help pages should not be language specific. Actually the fields use, constraints and limitations, are not language specific at all. — Jeblad 21:44, 2 April 2013 (UTC)
But how do you even begin to translate guidelines on whether or not to use capital letters into a language that doesn't even have a script that uses them? Del♉sion23 (talk) 23:18, 2 April 2013 (UTC)
Hey guys, description is just a single field that is the same across all languages. And it's already translatable. And what do you mean by making it "project-wide"? It's already project-wide. Every item has a label and a description. I don't understand how any of your discussion relates to my proposal. Kaldari (talk) 20:11, 3 April 2013 (UTC)
The discussion is not about descriptions, but about the guidelines for them. They are language wide not project wide at time. --Sixsi6ma (talk) 22:12, 3 April 2013 (UTC)

Thank you!

Heya folks :)

I wanted to say a big thank you in the name of the dev team to all of you. The first year is over and it's been an amazing ride for us. We're really looking forward to the next year. Thanks to all of you for being amazing and helping make this project rock. It's way beyond our expectations and we can only hope that it'll continue like that. Thanks for believing in us and sticking with us also when it was sometimes a bit painful.

Here's to year 2 of Wikidata! We're only just getting started ;-)

--Lydia Pintscher (WMDE) (talk) 12:58, 3 April 2013 (UTC)

Thank you and all the dev team that make this all possible. Filceolaire (talk) 18:47, 3 April 2013 (UTC)

Thank you to all! --β16 - (talk) 21:27, 3 April 2013 (UTC)

Q8827939 and similar

Do such categories (see w:en:Category:Suspected Wikipedia sockpuppets and w:en:Category:Wikipedia sockpuppets) really fit into project scope? I think they are user-space related. --EugeneZelenko (talk) 13:16, 3 April 2013 (UTC)

IMO We should not add those pages into Wikidata. There will never (or nearly never) be any interwiki links. --Stryn (talk) 14:00, 3 April 2013 (UTC)
I agree. Unless we have a clear vision as to how Wikidata items would be helpful for this, we should avoid them. We can always make an exception for the precious few categories where we have undeniable evidence that a user abused multiple accounts on more than one wiki. Pichpich (talk) 14:22, 3 April 2013 (UTC)
I know some socked at commons and enwp, other were socking at enwp and other language Wikipedias (mostly their mother language) and why shouldn't these categories added here? Mabdul (talk) 14:31, 3 April 2013 (UTC)
These pages and categories are not related to the content-ns anywhere, no-matter how X-wiki they are. Interwiki can be solved without Wikidata in those cases. - Lavallen (block) 14:45, 3 April 2013 (UTC)
True, we can handle it without Wikidata but I don't really mind if we do. The point is that sockpuppets active on more than one wiki represent a ridiculously tiny portion of the sockpuppet categories and at the very least, the default policy should be that we don't create items for sock categories but that, as always, common sense exceptions can be made. Pichpich (talk) 14:55, 3 April 2013 (UTC)
I don't think these belong here, they are not part of the "project" or helping to correlate user-facing content. Besides, having these might encourage interwiki socking as a type of "shrine building" as the userspace tags have been known to encourage prolific sockers on enwp. Courcelles (talk) 23:54, 3 April 2013 (UTC)

Rename AuthorityControl Gadget and add commonsmedia script

Hi guys, I've created User:Bene*/commonsmedia.js which adds the possibility to show an image's preview in Wikidata itself and thought it would fit into the default Gadget Authority control. But if we added it there, the name wouldn't fit so I propose to rename it to something better like ExternalLinker as mentioned above and add the commonsmedia feature to this gadget. Regards, --Bene* talk 16:24, 3 April 2013 (UTC)

That's convenient, though I supposed it is a fairly long "spaghetti". If there is no performance issue, I would support adding it by default (I faitly recall that image preview what included in the demo at some point, but it then disappeared). --Zolo (talk) 17:47, 3 April 2013 (UTC)
Commonsmedia isn't embedded due to some legal issues, but I think it will be fixed. — Jeblad 19:18, 3 April 2013 (UTC)

Event vs. Term

Why does World War II have a GND type of 'term'? It seems like an event to me. -- Ypnypn (talk) 18:18, 3 April 2013 (UTC)

GND is a file maintatained by the German national library. We have to use their data. They usually only see events like festivals, tradefairs or conferences as "events", historical events are mostly just "terms". It may have to do with the German language, where "Veranstaltung" is only a "planned event", while "event" in a broader sense would be "Ereignis". --FA2010 (talk) 18:27, 3 April 2013 (UTC)
The recurring problem with the GND type is that we're calling it "main type" as if it's absolutely needed and that we're following suboptimal translations like "event" rather than something clunky like "planned event" that would minimize errors (and frustration) by non-German speakers. And we should simply replace "term" by "everything else" since that's clearly the best description. Pichpich (talk) 20:03, 3 April 2013 (UTC)
Yup, "event" only in the meaning of sport evant, not countries fighting each other. Please also take a look at Wikidata:Infoboxes task force/events. --Kolja21 (talk) 19:58, 3 April 2013 (UTC)

Question

Hi! How could I know that ru:Папкович, Пётр Фёдорович has the entry Q4344703? I created Q9496873 for pt:Peter Feodorovich Papkovich, that was deleted.--Kaktus Kid (talk) 19:01, 3 April 2013 (UTC)

Disambiguation policy

Do we have a policy on items for disambiguation pages? I ask as I have found two items that would fit the label "Russia", description "Wikipedia disambiguation page". One at Q182648 appears to be for any article with the title "Russia", whereas Q232858 is a disambiguation page the translation of "Russia" into each language. I would have thought we would only group sitelinks together if they are the same word, not different words with the same meaning. Thus the former item would be filled with all disambiguation pages for the word "Russia". However, does this mean that the latter item should be broken down into many different items, as most have different spellings? And what do we do for non phonetic languages such as Chinese where the word is never likely to be the same spelling? Del♉sion23 (talk) 17:38, 31 March 2013 (UTC)

It's informal. The last time I brought it up I got "match the words together for disambigs", rather than "match the content or topic". --Izno (talk) 18:45, 31 March 2013 (UTC)
Actually there is Wikidata:Disambiguation_pages_task_force which tries to create a policy.--Saehrimnir (talk) 19:20, 31 March 2013 (UTC)
There was a vote a few weeks back; there are no official guidelines as yet, but we are working on it. --Sixsi6ma (talk) 19:27, 31 March 2013 (UTC)

I've separated them into Q182648 (Russia), Q9059689 (Russland), Q399309 Rossiya/Rossija, and Q232858 (Éguó) the Chinese pronunciation. Hope this fixes it. Del♉sion23 (talk) 19:56, 31 March 2013 (UTC)

I think Q232858 was an interesting example of "topical disambig"! It's obvious that there exists a set of pages in each language united by own translation of "Russia". Why not to join those lists in one item? Infovarius (talk) 20:07, 1 April 2013 (UTC)
Because, as Delusion already said, we link disambiguation pages of the same word, not of a translation of one of its meanings. --Sixsi6ma (talk) 21:51, 3 April 2013 (UTC)
I believe that this is too strict constraint. And there are no monochromatic separation "article/disambig" - there are disambigs that have a specific topic. Infovarius (talk) 13:01, 4 April 2013 (UTC)

q959163

Please add pt:Política da União Europeia and ro:Sistemul politic al Uniunii Europene. Thanks to IE8 I cannot do this myself. --193.18.240.18 06:48, 4 April 2013 (UTC)

I've added the pt article. The ro article is already linked to Q748720. --Spischot (talk) 07:19, 4 April 2013 (UTC)
ro now moved as well. --Spischot (talk) 07:22, 4 April 2013 (UTC)
  Resolved. Thank you, Spischot! --193.18.240.18 07:28, 4 April 2013 (UTC)

q4167946

Please add de:Eighty Mile Beach. Thanks --193.18.240.18 07:38, 4 April 2013 (UTC)

  Done --Michgrig (talk) 07:53, 4 April 2013 (UTC)

Two VIAF identifiers

The author Carlos Castaneda (Q158878) has two VIAF identifiers associated with him. I added the second one, since the VIAF database contains two numbers for him. Probably the duplicate numbers are due to a mix-up regarding his year of birth, see for example [3] or [4].

Anyhow, I seems to me there should be some place where this should be reported, but I don't know where. Gabbe (talk) 07:39, 4 April 2013 (UTC)

The only place I know of is w:en:Wikipedia:VIAF/errors. You should ask User:Maximilianklein‎. --Nemo 09:37, 4 April 2013 (UTC)

Category:Sapindaceae

Cannot add frr:Kategorie:Sapindaceae to Q8704363. Any ideas? --Murma174 (talk) 08:40, 4 April 2013 (UTC)

If you click Details in the error message, you'll see the reason:

Site link frrwiki:Kategorie:Sapindaceae already used by item Q9755722.

So, you just need to merge the two items. --Michgrig (talk) 08:54, 4 April 2013 (UTC)
Thanks for your quick reply. Merged and Rfd done. --Murma174 (talk) 09:07, 4 April 2013 (UTC)

Redirects

a few months ago I was discussing merges on project talk and it was pointed out that any tool to do this would have to leave behind a redirect from the deleted item to the merged-to item to maintain data integrity. Here we are months later and this 'Redirect' feature has not yet been coded but now we have a live system and so we have County Carlow with type of administrative unit given as q5307224 (Deleted item).

When will deletions stop and be replaced by redirects? Filceolaire (talk) 09:12, 4 April 2013 (UTC)

We have not gotten to it yet. You can follow the progress on bugzilla:38664. --Lydia Pintscher (WMDE) (talk) 10:43, 4 April 2013 (UTC)
This shouldn't be a technical fix. Administrators deleting properties without checking for backlinks should be trouted and asked to fix the links before deletion, or they should ask the editors who RfDd an item to fix the backlinks. --Izno (talk) 13:08, 4 April 2013 (UTC)

Item number 10,000,000

We're soon to hit item number 10,000,000! Just saying :) --Kristian Vangen 06:45, 4 April 2013 (UTC)

You mean 10,000,000?--Jasper Deng (talk) 06:46, 4 April 2013 (UTC)
Of course I do. Just testing you all ;) --Kristian Vangen 06:50, 4 April 2013 (UTC)
On the Finnish language correct number is 10 000 000 (without commas). 10,000,000 is just 10 :) en:Decimal_separator#Examples_of_use. --Stryn (talk) 08:33, 4 April 2013 (UTC)
Reached.. :) --Nizil Shah (talk) 12:08, 4 April 2013 (UTC)
Third straight milestone to be a category, interesting. Courcelles (talk) 13:24, 4 April 2013 (UTC)
Seeing as we hit this one in four days when we'd been fairly stable at 9-10 days for each the last 3 milestones (congrats to all of our wonderful bot ops, by the way!), it would be great if we could get some more discussion at WT:MP#Millions before we hit Q11000000.</canvassing> — PinkAmpers&(Je vous invite à me parler) 20:25, 4 April 2013 (UTC)
I think this one might just be an anomaly, it seems my bot created 608,683 items in the past 24 hours. It's probably not going to be editing that fast from now on, so it should be back to our normal rate... Legoktm (talk) 00:55, 5 April 2013 (UTC)

Should the big ones get all the info ?

Using an example to make things more concrete:. What info go into "California" and what should go into "Demogaphics of California" ? I would say everything in California, because:

  • demographic info is needed for the main article about California and many more articles have an article about California than about the demograpgics of California. Fragmented info would be harder to find
  • there is quite much potential overlaps between demographics of California and other items like "economy of California".

I think we should leave articles like "Demographics of California" essentially blank. In any case, we need to decide on that. --Zolo (talk) 17:15, 4 April 2013 (UTC)

This is not a matter of “size” but of the item type and the aplicability of a certain property. A state (such as California) has a population. A "Demographics of California" does not. Instead of deciding that some items should not have any property we'd better look how properties match – which could result in not having applicable properties for some items. --Spischot (talk) 17:33, 4 April 2013 (UTC)
Yes that makes sense, though I imagine that in practice the item would be considered to mean "item gathering data about the demographics of California". --Zolo (talk) 17:49, 4 April 2013 (UTC)
By the way, would it technically be possible to access data from another item like {{#property:p123|q456}}? --Bene* talk 19:34, 4 April 2013 (UTC)
Yes, see meta:Wikidata/Notes/Inclusion syntax#Accessing Item Data --Spischot (talk) 21:12, 4 April 2013 (UTC)

New user warning templates

Since a lot of IPs seem to go around adding languages for descriptions, I propose a series of user-warning templates to deal with that:

Uw-addlng1:

  Hello and welcome to Wikidata. I noticed that you added a language to the description section of an item. I have undone this change as that is not what a description section is for, please see Help:Description for more information. If you have any questions, you can leave a note at my talk page. Thanks!

Uw-addlng2:

  Please do not add languages to description sections of items as it is unconstructive. You may wish to see Help:Description for what is expected of a description. Thank you.

Uw-addlng3:

  Please stop adding languages to description sections of items. It is considered disruptive and may result in a block.

Uw-addlng4:

Same as Uw-vandalism4


What does everyone think? Jakob Megaphone, Telescope 20:23, 4 April 2013 (UTC)

These are not very explicit on what "adding languages" would mean - I understand them, but that's because I'm an experienced user.--Jasper Deng (talk) 20:26, 4 April 2013 (UTC)
Yeah. To be honest, I think this is more of our fault than anything else... the "add" button is rather hard to find. A far better solution would be either a more intuitive page layout, or some clearer instructions on every page (the idea of embedding some within a namespace parser on MediaWiki:Sitenotice has crossed my mind). — PinkAmpers&(Je vous invite à me parler) 20:29, 4 April 2013 (UTC)
What is adding a language to a description in this context? I would understand it as adding a description in some language, but as that is fine it must mean something else. Byrial (talk) 22:36, 4 April 2013 (UTC)
I'm talking about stuff like this. Jakob Megaphone, Telescope 22:39, 4 April 2013 (UTC)
Hmm, that is odd. I'm surprised that's common enough to warrant a template, but I take your word for it. Do you think it would be useful to link Help:Description from the higher-level warnings as well? Superm401 - Talk 22:43, 4 April 2013 (UTC)
Added the description help page to uw-addlng2. Probably isn't needed on 3 because the higher level warnings tend to assume the user is intentionally disruptive (judging by this). Jakob Megaphone, Telescope 23:21, 4 April 2013 (UTC)
I still don't get it. Why is this a common mistake? In any case, the warning template points to a page that does not explain the problem. Obviously it's an unacceptable to have a template that in essence says "stop doing this thing that I'm too busy to explain to you or you'll be blocked". Pichpich (talk) 23:38, 4 April 2013 (UTC)
I think it's a variety of reasons... people may be wondering how to see the label in their preferred language, or how to add a link, or (if they're really clueless) they might think that adding a language automatically translates the article. And the description box is simply the most obvious empty text field. As I said, though, I think warnings are the wrong direction here. The majority of users only do it a few times and then never edit again... it's basically just a form of (good-faith) test edit. (This is part of the reason we allow use of rollback on test edits.) Personally, if I were ever going to block someone for doing this, I would first stop and thoroughly explain things to them. My advice to recent-changes patrollers: Rollback and move on. — PinkAmpers&(Je vous invite à me parler) 00:34, 5 April 2013 (UTC)

Really odd interlanguage linkage

Originally posted at w:Wikipedia talk:Wikidata#Interwiki conflicts.

The English interlanguage link on so:Soomaaliya leads to en:Template:Cities of Somalia instead of en:Somalia. What's weird is that the item here (Q1045) looks just fine. Does anybody have any idea what's going on with this? The Anonymouse (talk | contribs) 04:30, 5 April 2013 (UTC)

Yup. Someone put the langlinks on so:Template:Magaalooyinka Soomaaliya after the </noinclude>. I've moved the end tag, and the article should be fine now. — PinkAmpers&(Je vous invite à me parler) 05:06, 5 April 2013 (UTC)
Hmm, good catch. The Anonymouse (talk | contribs) 06:01, 5 April 2013 (UTC)

Adding a property to several items in the same Wikipedia category

Is there an easy way (i.e. not creating a bot) to add the same property to several items based on a Wikipedia category (for exemple)? Thanks. Ayack (talk) 16:30, 4 April 2013 (UTC)

Wikidata:Tools/Array_properties_gadget. HTH, --Polarlys (talk) 17:40, 4 April 2013 (UTC)
Thanks but unfortunately it doesn't seem to work... Ayack (talk) 08:21, 5 April 2013 (UTC)
Uhoh, what's not working? Legoktm (talk) 09:12, 5 April 2013 (UTC)
[5] Seems fine to me? Legoktm (talk) 09:13, 5 April 2013 (UTC)
Yes, sorry, the request is fine, but I thought that your bot wasn't completing it. Maybe I have just to learn patience... Thanks. Ayack (talk) 09:22, 5 April 2013 (UTC)
Ah, sorry it doesn't do it immediately. An admin has to move it to User:Legobot/properties.js and then the bot will do it. I'll take a look and move some over now. Legoktm (talk) 09:26, 5 April 2013 (UTC)
Great! Thanks a lot! Ayack (talk) 09:29, 5 April 2013 (UTC)

Filter recent changes by language

Tracked in Bugzilla
Bug 44138

What do you guys think about adding a feature that could filtering Special:RecentChanges by language for the labels, descriptions, and sitelinks? Since properties are multilingual, I think that they probably shouldn't be filtered if we were to implement this. FallingGravity (talk) 07:45, 5 April 2013 (UTC)

I thought I smelled a bug here :) FallingGravity (talk) 08:52, 5 April 2013 (UTC)

Q10361461

Once more IE8 stops me from doing this myself: Please add de:Schattenfinanzindex and en:Financial Secrecy Index. Thank you --тнояsтеn ⇔ 09:08, 5 April 2013 (UTC)

  Done in Q5449645 instead. Legoktm (talk) 09:10, 5 April 2013 (UTC)

Parsing Wikidata

I want a list of all items that have the same String value of Property:P345, sorted by IMDb identifier. Quite possibly there are none, in which case I would just like that fact stated affirmatively.

Trawling through Special:WhatLinksHere/Property:P345 manually is too much work, so I was thinking asking a bot to do said task once a week. Since this merely involves parsing Wikidata (rather than editing any items), I wonder if there are simpler ways of doing it other than by bothering Wikidata:Bot requests? Gabbe (talk) 09:33, 5 April 2013 (UTC)

Better work offline using a dump of wikidata DB. See Wikidata:Database download 141.6.11.15 10:56, 5 April 2013 (UTC)

Opening Wikidata entries from Wikipedia categories and adding data

Hi! I want to add the same description to all Wikidata entries of articles in a certain Wikipedia category. Any ideas how to automate this? --Polarlys (talk) 17:47, 4 April 2013 (UTC)

Ask a bot for this! --Ricordisamoa 18:25, 4 April 2013 (UTC)
Thank you for your answer, but this was not the question. I want to work on entries. --Polarlys (talk) 19:08, 4 April 2013 (UTC)
You can create your own bot (see Wikidata:Creating a bot if you don't know how to do it) and request a botflag for it at WD:RFBOT. --Stevenliuyi (talk) 19:15, 4 April 2013 (UTC)
I don’t want to use a bot. I asked for supporting scripts for editing entries. --Polarlys (talk) 19:39, 4 April 2013 (UTC)
For now there is no automatic way to do what you want without a bot, but you can use the gadget "autoEdit" (you can find it in your preferences) and Items by cat to speed up your work.--Stevenliuyi (talk) 19:53, 4 April 2013 (UTC)
Thank you. --Polarlys (talk) 22:37, 5 April 2013 (UTC)

Logbook of phase II bot jobs and gathered data

I am wondering what robot jobs for collecting phase II data that are ongoing or completed? Is there any infobox in some Wikipedia version, from which all relevant properties are imported? What external sources are completely imported?

I would like to see some kind of logbook mechanism of ongoing and completed import jobs. There are several conceivable options for this:

  • In the field "Robot and gadget jobs" on the property documentation form on the top of each property talk page. See for example Property talk:P107.
  • On the WD:phase II (infobox taskforce) pages, and some subject specific taskforces, under each infobox mapping.
  • On the WD:Bots pages, or each bot user page.
  • For each imported statement, a reference should state which Wikipedia version (or external source) that it is imported from, and maybe also which infobox.

Which options do you think are best? Mange01 (talk) 15:57, 5 April 2013 (UTC)

Please move...

Q10384003 from Rebeldes de Canadá to correct Rebeldes en Canadá. Sorry, my fault. --Si! SWamP (talk) 20:18, 5 April 2013 (UTC)

done by myself. If I only knew it is *that* easy.........:-)) --Si! SWamP (talk) 20:20, 5 April 2013 (UTC)

Wikidata:Requests for comment/Defining bureaucrats

Go comment! --Rschen7754 21:38, 5 April 2013 (UTC)

Merging properties

Qualifiers will be implemented in the near future (qualifiers are in test on the test system, see here or here for examples) so we can think to merge some properties and to use qualifier to distinguish the different requirements associated with the items.

Languages

language: we have now three properties dealing with language: Property:P103 (native language for person), Property:P37 (official language for country) and Property:P364 (original language for books ant other art works). I need for reference purpose to create a new property about language in order to indicate the language used in the reference allowing wikipedia to choose the most appropriate one (if there are several references in different languages).

My proposition is to merge all properties and to use qualifiers (native, spoken, official, original,...) in order to give additional information to the statement or the references. -- Snipre (talk)

  Support, but it may be nice to leave a message on the property's talk page as well (actually, I had already made a related case on Property talk:P364 before I saw this proposal)--Zolo (talk) 16:57, 6 April 2013 (UTC)

citation of reference origin

We have two properties used to define origin of citation, Property:P92 (legal basis) and Property:P248 (stated in). The first one deals only with legal document, the second for all the rest (book, newspaper, speech,...).

As there are no reasons about having a specific property for legal document and not for others I propose to merge these two properties in a property (stated in/by) and if there is a need of specifying that the document is a legal one, qualifier can be used. Snipre (talk) 09:51, 30 March 2013 (UTC)

  maybe, but we first should know how qualifiers work. --Ricordisamoa 09:54, 30 March 2013 (UTC)
Just go on the test system. Snipre (talk) 09:56, 30 March 2013 (UTC)
Take for example Q1545193. This is a Rijksmonument with id 19264. Would you do this in a different way with qualifiers? Multichill (talk) 16:38, 30 March 2013 (UTC)
Support: maintaining a clear cut difference appears very difficult. However, looking at use cases, P:P92 does not just provide a source, it often also says what event made the statement true. I think we should have "created by" property or something (it would be used as a qualifier or as main property, not in sources) --Zolo (talk) 16:55, 6 April 2013 (UTC)

Property:P348 - Stable version (software only)

Not about merging, but about renaming this property to just 'version'. For the same reason as above: A qualifier can be used to mark the version as stable or not. Ok, need an additional property which used values like 'stable', 'unstable', 'preview' and something like that. Renaming it to 'version' we could also use this for non-software related subjects, see here (so 'version' would be the qualifier). As there are only 20 links to the 'stable version' at moment, they are quickly added with the 'stable' qualifier. --Nightwish62 (talk) 16:31, 30 March 2013 (UTC)
I agree with you Nightwish62. --Viscontino talk 11:36, 6 April 2013 (UTC)

Can't see outdated texts

On Translation of the wiki page Wikidata:About (fi) is "6% outdated", but I can't see any outdated texts there. I have also tried to save all messages there many times, but nothing changes. I don't see what needs translating. Currently it shows that 94% is translated, but it's not possible to get 100%. --Stryn (talk) 19:29, 5 April 2013 (UTC)

Hmm... who did and what? I would like to know how it's now translated 100% and not outdated anymore. --Stryn (talk) 20:22, 5 April 2013 (UTC)
It was done on request of a Wikidata user: bugzilla:44328. Ah, but I see that you updated it now; the only option left is looking very carefully at the numbers and using Special:Translate directly. I don't understand though: the page you linked should have contained what you needed. --Nemo 07:21, 6 April 2013 (UTC)
Now same with this page: "83% translated, 0% proofread, 17% outdated". But I can't see there any outdated texts. --Stryn (talk) 07:44, 6 April 2013 (UTC)
Ok, that was because latest edits were not marked for translation, I guess. But now it's also 100% translated. --Stryn (talk) 11:16, 6 April 2013 (UTC)

Continuation of the weekly newsletter

Since the first year of Wikidata development is over I think it is time to move the weekly newsletter more into the hands of the community. From now on it'll be drafted at Wikidata:Status updates/Next. Please help make it rock! I'll continue to send it out on Fridays for now. It has a considerable audience by now. Let's make good use of it :) --Lydia Pintscher (WMDE) (talk) 23:04, 5 April 2013 (UTC)

Thank you for doing this work. I like to read this weekly updates to see the progress Wikidata is making. Could you please explain me, what exactly "Implemented string formatter" from this week status update means? --Nightwish62 (talk) 09:18, 6 April 2013 (UTC)
I'll ask for clarification. --Lydia Pintscher (WMDE) (talk) 10:50, 6 April 2013 (UTC)
Thank you for the work lydia, but can you split the structure of the newsletter into 2 distinct parts: one for the development, press and events which is relevant for you or at least from the wikidata team and the rest where the community can collaborate ? Snipre (talk) 10:34, 6 April 2013 (UTC)
Does it need to be two distinct parts? I think collaborating on one big thing is much better. --Lydia Pintscher (WMDE) (talk) 10:50, 6 April 2013 (UTC)

Doubt

Hi everybody. Sorry aboout my English. Is this the same of that? Merge? How to do that? Best wishes from Brazil. Sturm (talk) 23:54, 5 April 2013 (UTC)

I've merged the links and requested deletion of 'that'. FallingGravity (talk) 05:34, 6 April 2013 (UTC)
Help:Merge gives instructions on how to merge items. However, I don't think it has been translated into Portuguese yet. Del♉sion23 (talk) 12:15, 6 April 2013 (UTC)

Auto Welcome Bot

Please see the request linked below for an auto welcome bot for wikidata. Wikidata:Requests_for_permissions/Bot/SamoaBot_4 ·Add§hore· Talk To Me! 17:41, 13 April 2013 (UTC)

I withdrew my RFBOT. Thanks. --Ricordisamoa 19:13, 13 April 2013 (UTC)

Statistics of users by edit count

Hi, my bot (User:InkoBot) could create a list of all Wikidata users sorted by editcount as the list for bots. I have two questions:

  • does InkoBot need a flag for creating such lists?
  • Is an opt-in list or something like that required for analyzing an user's edit count?

Regards, IW (wikidata addict) 12:11, 6 April 2013 (UTC)

The English Wikipedia practice is that users who do not want to see their name on the list can replace themselves with a <placeholder>. I suggest to do the same here.--Ymblanter (talk) 12:25, 6 April 2013 (UTC)
  Oppose See this RFC. BTW, there is already this; and any user can easily generate such lists via a JavaScript query. --Ricordisamoa 12:48, 6 April 2013 (UTC)
I made some additions to the bot's contribution list that are not part of the statistics you linked. In addition to that, I do not think that everybody can easily create and analyze an API query. Regards, IW (wikidata addict) 14:23, 6 April 2013 (UTC)
Looking through some stats it doesn't seem meaningfull to make stats over edit count. How to handle users running bots on their own account? How to handle bots that is actually under supervision? This seems like a mess to me. Users are active or they are not. Period. — Jeblad 15:58, 6 April 2013 (UTC)
  Agreed --Ricordisamoa 16:07, 6 April 2013 (UTC)

Ok, you are right. I would propose another solution, like de:User:Beitragszahlen. This bot enters the edit count to a user's subpage if he wants, e.g. for using in templates or on the user page. IW (wikidata addict) 18:25, 6 April 2013 (UTC)

It still sounds as a futile task to me... and this script can add a real-time edit counter to userpages without editing them. --Ricordisamoa 18:39, 6 April 2013 (UTC)
Personally I don't care what other people think, I find a list of bots by edit count be useful when comparing the amount of progress (I try and keep a close eye on how many edits my bot has queued) my bot has made compared to others that are working on similar tasks. I guess its a wikt:YMMV kinda thing. Legoktm (talk) 18:55, 6 April 2013 (UTC)

Properties preceded by and followed by

Have have some proposals for these two properties (preceded by and followed by). As you know they could be used for many items: months, awards, asteroids, presidents... Some articles may need to use this property for different items. Like Albert Einstein, as you can see in Portuguese Wikipedia or François Hollande in English Wikipedia. In order to solve this, I will make some proposals.

Proposal 1

The property could be used in, and just in, sources section. Imagine with the item Albert Einstein. It could be used in awards, for example Nobel Prize in Physics, in sources section, in would have the year (future feature, I hope) and other two properties preceded by and followed by in the category.

Proposal 1.1

I tried to explore the initial proposal, so, if possible, I would like to see this implemented. The two properties should be merge in just one, with two sub-properties. In sources section, for Nobel Prize in Physics (in the item about Albert Einstein), it would be like this:

preceded by year (or date) followed by
Charles Édouard Guillaume 1921 Niels Bohr

If there is no value to the property, it would be used no value to designate that.

This proposal would be good to be the basis of templates like this w:en:Template:Succession box (Sarilho1 (talk) 15:08, 6 April 2013 (UTC))

Proposal 2

If the first proposal (even without the sub-proposal) is to difficult to implement, it could be added to the sources section (being the preceded by and followed by the principal properties) what that properties refer to.

Please give your opinions and/or proposals for this problem. - Sarilho1 (talk) 14:31, 6 April 2013 (UTC)

Comments

I assume you mean qualifier rather than "source". If so, I also think that proposal 1 would ne clearer (as has already been mentionned in User_talk:Denny#preceded_by). We could also tell a bot what sort of statements needs a "preceded by" qualifier to help with maintenance. However, I do not see the point of merging "preceded by" and followed by".--Zolo (talk) 14:52, 6 April 2013 (UTC)

I forget that! This proposal would be the basis of succession boxes (w:en:Template:Succession box) - Sarilho1 (talk) 15:08, 6 April 2013 (UTC)
Ok, but I do not see how why it requires merging both properties. I do not see anything wrong with using two properties in to build the box. --Zolo (talk) 15:32, 6 April 2013 (UTC)
Maybe this could be done automatically without using my proposal. And maybe it is a bit complicated. - Sarilho1 (talk) 17:13, 6 April 2013 (UTC)
I see why you want to clarify what series applies (presidents, Nobel Prize winners, etc.) (I've also been using these for successor states). A qualifier may help with this. As far as merging them, I don't think proposal 1 will work (it's not a triplet). But a bi-directional property might when those are available. Superm401 - Talk 23:43, 6 April 2013 (UTC)

Infobox rockunit

I have a general question about articles using infobox rockunit. Has wikidata gathered those properties yet? I tried to fill it some properties for Q1554536 but many properties (e.g. underlies, overlies) are still missing from the properties. Should I request them or wait for phase 2 to finish? --Tobias1984 (talk) 18:07, 6 April 2013 (UTC)

You should request them @ Wikidata:Property proposal/Term (it's the dub of Wikidata:Infoboxes task force/terms). --Kolja21 (talk) 22:21, 6 April 2013 (UTC)

API questions: wbeditentity and wbsetreference

The examples given in the API don't seem very thorough yet. What does setting exclude=ns on wbeditentity do? Can wbeditentity be used to set properties or only labels, descriptions, etc? What should the submitted JSON look like for a setting an actual reference? I was surprised that the UI asks me to choose a property before I can give a reference value. Which property or properties are we supposed to use? Wakebrdkid (talk) 19:49, 6 April 2013 (UTC)

As far as I know currently only labels, descriptions, aliases and sitelinks (on items) can be set via wbeditentity. (Don't trout me if I am wrong) All the other properties, values, sources etc. have to be set one by one. -- Bene* talk 06:41, 7 April 2013 (UTC)

Shapefiles and interface to Wikiatlas

There has been a lot of discussion about links to other WMF projects. One project that hasn't been mentioned is WMF's stealth project - WikiAtlas. The atlas which appears when you click the globe next to coordinates. This not only works for locations on Earth. There are WikiAtlases for our Moon, Mars and various other celestial objects. It is an amazing project that deserves more publicity.

At present it gets location info from Wikipedia. At some stage in the future it will, I presume, get this info from Wikidata, along with sitelink info so it can offer articles in multiple languages.

This is all very well but we need something else. Shapefiles. Files which define areas (country boundaries for instance) or polylines (the branching lines of a river or a railroad network).

I don't know how this can be achieved technically but I really want each administrative unit to have a link to an map which shows its boundaries overlaid on an interactive map. I believe the wikidata page for that unit should have a link to the shapefile which defines those borders.

I know it is not going to happen soon but I would like to make sure it is on the wish list for future stages. Filceolaire (talk) 20:55, 30 March 2013 (UTC)

Filceolaire, there is an initiative on the side of Sharemap.org to create online Shapefiles objects which will be used by wikipedia. The project still need official support from the wikimedia foundation. Yug (talk) 21:29, 30 March 2013 (UTC)
Wikidata will support the datatypes geographic location (see also the most recent draft) and geographic shape. While the geographic location is announced to become available soon [6] and the implementation is already pretty advanced [7] the information on the implementation of geographic shapes is very rare. --Spischot (talk) 06:45, 31 March 2013 (UTC)
Just for your info: geographic shape as a data type is planned but it is unclear at the moment how exactly it'd work and when we'll get to it. That's also why there is so little info about it. We just have not figured it out yet. --Lydia Pintscher (WMDE) (talk) 09:30, 31 March 2013 (UTC)
Jut saw Wikidata:Property_proposal/all#Openstreetmap_IDs. Open Street Map Relations seem to do this job. Filceolaire (talk) 10:05, 7 April 2013 (UTC)

Sources that use waivers

Some sources use waivers to convey that they want backlinks to their datasets. [8] (For the whole website) [9] (Part about SSR) In Europe there is something called "databasevern" or "database protection", that is the individual facts are not protected but the overall database is protected. In those cases some of the owners wants a backlink to the original dataset. I think this is very close to a waiver, but don't know enough about it. [10] Perhaps someone with more knowledge on the field can elaborate. This is also very close on our existing references, so it is basically an expectation on a social contract like "use of this data imply that you should use a reference to the original source". This is even close to provenance information in a linked data setup. A more elaborate schema would be to have an item for the database (would this be called the provenance?) with links to the waiver (my first two example links) and to the norm (for example Open Data Commons Attribution-Share Alike norm), and then link to this item from all the places that uses this database as source and also have reference links to the actual place in the source database were the fact is stored. That would solve our "systematic use" of those databases, even if our own database is CC0, and even if the individual facts are free to use as they would be in public domain. Does this make sense? — Jeblad 17:53, 3 April 2013 (UTC)

Swedish copyright-law protects databases in 15 years from publication. § 49 Upphovsrättslagen. (i.e. sv:Katalogskydd)
Statistics Sweden allow full use of many of their databases if they are attributed in some way. -- Lavallen (block) 18:13, 3 April 2013 (UTC)
Also Norwegian copyright-law protects databases 15 years from publication. § 43 Åndsverksloven. (i.e. no:Databasevern) — Jeblad 19:24, 3 April 2013 (UTC)
All EU countries have the 15 years protection for databases. It might be possibly to force to show the reference link inside the Wikimedia family but what happens when someone takes the Wikidata databases and claims CC0 doesn't require such a link? I think before collecting data from sources which are not 100 % okay this should be checked by a lawyer. This is one of the core questions for Wikidata. NNW (talk) 15:24, 6 April 2013 (UTC)

We would also need to state clearly whether need to fully comply with local IP laws or just with US database rights that seem to be rather permissive. Commons usually respect copyright laws of both the US and of the file's country of origin, but it makes some exceptions, and I really do not know whether we can have similar rules here. It would be really useful to have a summary of legal constraints before we start serious imports from external sources. That would be really annoying if we and up deleting everything we have imported --Zolo (talk) 10:10, 7 April 2013 (UTC)

Bureaucrats?

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

The request for comment has started, and this discussion clearly says Yes to bureaucrats. — ΛΧΣ21 16:35, 7 April 2013 (UTC)

It has been mentioned that we could use bureaucrats, and I am willing to start a RFC on defining what the role would look like. However, before I do that, does the community have an interest in having crats, or would I just be wasting my time? --Rschen7754 02:37, 5 April 2013 (UTC)

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

It's started

Please comment on Wikidata:Requests for comment/Defining bureaucrats! --Rschen7754 21:36, 5 April 2013 (UTC)

Vandalism again

I was just patrolling the instances of new editors removing sitelinks (the list of 500 edits go back for about two days), and I was really stunned that over 80% of those are vandalism. I guess I know how the removal of this vandalism can be automatized. If a link in the list can be rollbacked, I rollbacked it. In some cases, I got a system message that it can not be rolled back because the sitelink is used by a different item (in 90% of those the initial item is empty, but empty items are found by bots anyway). Those are soft-false-positives. In most cases, the rollback works. Then the edit was either vandalism, or a hard-false-positive. I found two hard-false-positives in my list: once an editor removed a link to a deleted article (those eventually are tagged by a bot), and in another case an editor removed an article which was not really appropriate in the item, but failed to create a dedicated item for the article. In both cases, those were registered users, and the edits of registered users are generally ok (I found only one all contributions of whom were vandalism). Could we let bots doing this? Rolling back items from IP removing sitelinks, and also rolling back items of new registered users removing sitelinks, with creation of a log? The percentage of false positives which can not be repaired automatically (in my example, just one edit) is so tiny that I believe it would be a net gain.--Ymblanter (talk) 13:34, 6 April 2013 (UTC)

I've already thought of this, so I built DataLiveRC to help filtering such edits, but if you feel that a bot could do this task, I can think of it. Possible process:
  • Detect edits with that tag;
  • Exclude edits older than 24 hours (to prevent instant-reverting);
  • Exclude sitelinks that have been added to another item;
  • Exclude edits of trusted users, and patrolled edits;
  • Mass-revert those edits. --Ricordisamoa 15:03, 6 April 2013 (UTC)
    Smth like this, and possibly log reverted edits by registered users, so that those could be checked manually.--Ymblanter (talk) 15:12, 6 April 2013 (UTC)

Tthis will be basically to block newcomers from editing, and I really don't think that is the way to go. The present system is able to block newcomers from editing sitelinks if that is wanted, but I really don't think so. Perhaps if we could split anons from logged ins… Most of the so-called vandalism comes from failing to attach the article to the correct item, and that should be the case we try to solve. Users find that some article are not assigned to the correct item, but fails to find the correct item. Then they stop half-way after they have removed the sitelink but before they have found the correct item to reassign the sitelink. Why does that happen? — Jeblad 15:50, 6 April 2013 (UTC)

I do not propose to block anyone, I only propose to revert. And, at least for the languages I could read, none of the sitelinks removals I have seen today were justified (with the only exception I mentoned).--Ymblanter (talk) 16:30, 6 April 2013 (UTC)
I don't think anyone would be blocked from editing. If the sitelink was removed a day before and not added anywhere else, why should it be left? To give the new editors their fun? I really think that isn't a problem. --Bene* talk 06:47, 7 April 2013 (UTC)
We could think of a help-page to briefly explain the Wikidata system, and how/where to add/remove sitelinks, so that newcomers could be redirected to this when trying to remove sitelinks. --Ricordisamoa 07:16, 7 April 2013 (UTC)

Property:P150

See People Republic of China and Japan... That's crazy!!! tntchn Comment · Contribs 16:57, 6 April 2013 (UTC)

I don't see the problem. If the country subdivides, they should be indicated, isn't it? - Sarilho1 (talk) 16:04, 7 April 2013 (UTC)
It's another great property for playing POV games. But since we don't have transitive properties you can claim the Falkland Islands for GB and Argentina at the same time. --Kolja21 (talk) 22:15, 6 April 2013 (UTC)
China has 34 Provincial-level divisions, Japan has 47 Prefectures. Half of the page are about subdivisions, and you have to scroll to the bottom of the page to look over the link. tntchn Comment · Contribs 02:05, 7 April 2013 (UTC)

Kosovo

Can Q1231 and Q1246 be merged to one item? --2A02:810D:10C0:E1:9107:732E:9B0B:4654 11:51, 7 April 2013 (UTC)

Probably not. Q1231 is about the region called Kosovo in South Europe and Q1246 about the Republic of Kosovo. These are different things. Regards, Vogone talk 12:05, 7 April 2013 (UTC)
Absolutely not. As Vogone said, one is about the geographical region in Southern Europe while one item is about the Republic of Kosovo, the country that declared its independence in 2008. The site links should be checked to see which article on Kosovo is on the right items, but a merge is definitely not a solution. Regards, — Moe Epsilon 13:55, 7 April 2013 (UTC)

Change property proposal template: Highlight the proposal

For me, sometimes it's a little bit hard to see where a property proposal ends and where the discussion begins. Therefore I suggest to highlight the proposal a little bit. Just a little, not that it's in an annoying fashion. So I did a example on Wikidata:Property proposal/Generic. You can compare it with the old version. It's not the question about which background color or how to technical solve it (at moment, pure HTML code, could be an Wikitable or template also), it's just the question if we should highlight the proposal at all and change the proposal template. Please let the background color on the generic proposal till the end of this discussion (just a few days), if we decline the highlight, I'll remove it again. So please vote about it. --Nightwish62 (talk) 09:43, 31 March 2013 (UTC)

  Comment The aqua background (#AFEEEE) makes me feel like I'll be seeing afterimages of proposal discussions.
Some alternatives ordered by decreasing strength:
  • background-color:#EFFFFF; border: 2px solid #CDDDDD
  • border: 2px solid #CDDDDD (no change in background color)
  • Encourage people to indent comments that are in reply to the proposal, and don't make any HTML-based styling changes
I think more strongly visually distinguishing proposals is a good idea, though! Emw (talk) 15:33, 31 March 2013 (UTC)
I like the first one of your alternatives. --Nightwish62 (talk) 17:06, 31 March 2013 (UTC)
One week since my proposal. One pro vote from Emw. I'll change the template therefore now. We still can remove it. --Nightwish62 (talk) 20:01, 6 April 2013 (UTC)
I went ahead and   Done the changes. — ΛΧΣ21 06:01, 8 April 2013 (UTC)

landmark as a "instance of"

Do you think is it correct to use landmark as a "instance of" (P31) especially for historical buildings Amir (talk) 20:54, 7 April 2013 (UTC)

Yes, but it could become a POV battleground (what's well-known enough to be a landmark), so it should be cited to a reliable external source (not just Wikipedia) calling it that. Superm401 - Talk 01:10, 8 April 2013 (UTC)

Properties for creation and deletion

We now have a new userright that is required in order to create properties. See Wikidata:Property creators for details. To request it, you can go to WD:RFP, but you are generally expected to have experience working with and discussing creation of properties.

We also now have a separate Wikidata:Properties for deletion to reduce the size of the Wikidata:Requests for deletion page. --Rschen7754 22:38, 7 April 2013 (UTC)

Phase 2 soon on English Wikipedia

Heya :)

I just wanted to give you a heads-up: We are planning to roll out phase 2 on English Wikipedia next Wednesday (April 3). We're currently carefully monitoring performance on the first 11 Wikipedias that got phase 2 two days ago. If there are issues we might have to postpone this a bit but this is the current plan. --Lydia Pintscher (WMDE) (talk) 14:50, 29 March 2013 (UTC)

Update: We'll have to postpone this for a bit because we ran into some issues. I'll keep you posted. --Lydia Pintscher (WMDE) (talk) 20:42, 29 March 2013 (UTC)
The new date for this is April 8 (if there are no other issues as usual). --Lydia Pintscher (WMDE) (talk) 12:05, 5 April 2013 (UTC)
All remaining Wikipedias are planned for April 10 if there are no issues. Let's see if we can make it :) --Lydia Pintscher (WMDE) (talk) 17:06, 5 April 2013 (UTC)

Update: We decided to delay this. I'll let you know as soon as I know more about a new date. --Lydia Pintscher (WMDE) (talk) 18:31, 8 April 2013 (UTC)

Anzeige bei Versionsunterschied kaputt

Kann da jemand einen Bug zu anlegen bitte: http://www.wikidata.org/w/index.php?title=Q10534895&diff=prev&oldid=22201139 --2A02:810D:10C0:E1:9107:732E:9B0B:4654 15:35, 7 April 2013 (UTC)

re: "Display with diff broken" (google)
Actually, the edit summary is misleading when removing aliases, see Q61. --  Docu  at 23:12, 7 April 2013 (UTC)
Ooops, this was meant to be posted on the German project chat. Anyway, the broken diff was fixed as it seems. --2A02:810D:10C0:E1:8D42:BCBA:1986:F040 17:56, 8 April 2013 (UTC)

Question about {{noexternallanglinks}}

Just a small question I am wondering about. Is censorship of interwiki links for violation of a certain censorship policy on a Wikipedia what this tag was meant for? I am talking specifically about its usage on the Japanese Wikipedia in order to censor the names of people who have been convicted of crimes (which are often the titles of the other language Wikipedias). The Japanese Wikipedia has a strict censorship policy regarding such things, which is why it is currently being done. Is this what the tag is meant to be used for?--192.5.110.4 23:13, 7 April 2013 (UTC)

Per meta:Wikidata/Status updates/2012 04 13 this was implemented about a year ago (under the early name noexternalinterlang). I don't think anyone was thinking about this ja.wp case at the time it was developed! However, whatever the original expected use, the magic-word exists, and how ja.wp decide to use it to handle reader-visible content is fundamentally up to them. Andrew Gray (talk) 23:34, 7 April 2013 (UTC)
BTW: Do you mean censorship or data privacy? --Kolja21 (talk) 00:27, 8 April 2013 (UTC)
Data privacy may be one way of saying it if the information was not publicly available (especially not the page titles of other language Wikipedia editions), although when it comes to suppression of interwiki links, I can't see how there is any better word than "censorship," especially since this has a big potential of hindering cross-wiki collaboration.--New questions (talk) 01:02, 8 April 2013 (UTC)
The parser function is for those cases where the articles at the different Wikipedias doesn't align. The number is small, but significant. There is an analysis [11] that gives the numbers for each Wikipedia. In the specific cases you refer to the parser function might be used, and it is outside the initial intended use, but such use is entirely up to the community at that Wikipedia. Other language versions might do similar things because naming convicted criminals after their sentence is served is illegal in some countries, with some exceptions like size and type of case. If you think that is censorship, feel free to lobby your local government or any of the involved ones. — Jeblad 11:52, 8 April 2013 (UTC)

English variations.

Is it necessary to have two variations of English settings, English and British English. We only have one English language wikipedia which seems to survive edit warring over the correct spelling of colour. So why two different sets of descriptor strings where you can't see one if you have the other selected. And if nothing is entered as the label in the English variant you are using, but is entered in the other variant you see nothing, even on watched pages etc. Martin451 (talk) 09:50, 8 April 2013 (UTC)

That's a problem, but for Portuguese this is very useful for the two major variants of Portuguese (European Portuguese and Brazilian Portuguese) to avoid war editions and to construct a multilingual project. The only thing you can do is add to your userpage an box with your skills in other variant, as I do to correct some errors in Spanish or Galician. - Sarilho1 (talk) 10:04, 8 April 2013 (UTC)
A good example is: ônibus (bus), that in European Portuguese we called autocarro, in Brazilian Portuguese ônibus, in Angolan and Mozambican Portuguese machimbombo and in Guinean Portuguese toca-toca or otocarro. - Sarilho1 (talk) 10:11, 8 April 2013 (UTC)
I see the point, but with English most 99% of people should be able to read both, with just the odd word causing problems, the biggest issue is spelling. When there is confusion, (cell phone/mobile phone) both phrases could be used. This just seems to up the workload (manual and bot) for no gain. I see we also have Canadian English, but not Australian.Martin451 (talk) 10:34, 8 April 2013 (UTC)
I think that this is coordenate in Meta, but I'm not sure. - Sarilho1 (talk) 11:06, 8 April 2013 (UTC)

Gadgets link in toolbar

Since we're such a tech-heavy project, I think it would be useful if we emphasized our gadgets more. My idea is to put a link in the toolbar. So it would go ... Talk PreferencesGadgets Watchlist .... (The green is just to make it stand out a little bit, and because gadgets are cool and flashy.) Thoughts? — PinkAmpers&(Je vous invite à me parler) 13:10, 8 April 2013 (UTC)

I would rather see a link to Wikidata:Tools in the toolbox and fix up the page so that it shines.  — billinghurst sDrewth 16:00, 8 April 2013 (UTC)

symmetric relations

There are relations likes "shares a border with" (for countries). Is there a way to define such a relation symmetric, or are there always two distinct relations in both directions? Rewireable (talk) 13:51, 8 April 2013 (UTC)

"Shares a border with" should be used symmetrically, though there are problems with the consistemt usage (for example, administrative units in different countries sharing a border).--Ymblanter (talk) 15:45, 8 April 2013 (UTC)

Wikidata is slow in updating the Wikipedias

Tracked in Bugzilla
Bug 45892

I am sure that it has been mentioned, but it is a serious problem: Until Wikidata, it was obvious that any changes were immediately available. Now, with Wikidata, we enter the relevant interwiki link here, and then it takes a long time (more than 10 minutes - I do not know how long it takes) to appear in Wikipedia (yes, I am pressing CTRL+F5 befre checking). This is a serious failure of the system. The expectation is very clear: Immediate updating. Anything else cannot be considered as successful implementation or successful roll-out of Wikidata. --FocalPoint (talk) 11:18, 7 April 2013 (UTC)

This is mentioned in FAQ, question 26. --Michgrig (talk) 11:26, 7 April 2013 (UTC)
Why the need of immediate updates ? There are a lot of changes now because we are starting to build the database, but after some months the number of changes will be reduced and a fast update won't be so important. What is important is the data transfer time when openeing a wikipedia article in order to avoid the quality of an action which will be made thousands times (when updating will be done one hundreds times). Stop asking premium quality when we are speaking about limited ressources project. Snipre (talk) 11:46, 7 April 2013 (UTC)
For the moment this is a real issue as the lag time from change dispatching and the job queue itself is skyrocketing. You can check the stats at Special:DispatchStats, but note that the job queue itself is in addition to these numbers. — Jeblad 12:09, 7 April 2013 (UTC)

Restore bots work

See also Wikidata talk:Bots#Bot speed

As discussed on mailing list, we need bots for the kickstart of the project, we can't wait months for a bright future. The update of local wikis is slow by itself and limiting bots didn't work, so I propose to restore normal operation. Wikidata and WMF devs have to work on the proper solution for this anyway, the backlog will be recovered. If we don't use the wiki at its full, we'll never even know what are its actual needs (for updates and so on) and this will slow down development of the project. --Nemo 19:42, 8 April 2013 (UTC)

No actually we do not need to rush bot imports at all. It's ok to not have the data that is being imported now available right away. I think it is much worse to make the Wikipedians angry by delaying page purges and watchlist notifications than have data imported here a bit slower. We are _not_ in a hurry to import all the data of the world! It is perfectly fine for this to take a while - even a long while.
As for the proper solution: yes it is being worked on as fast as we can. But until then please do the right thing and not just close your eyes to the issues being caused. There are a lot of Wikipedians asking why pages don't purge and wth changes don't show up on their watchlists.
The stop/throddling of one bot did help. The lag would be even worse now otherwise. --Lydia Pintscher (WMDE) (talk) 21:36, 8 April 2013 (UTC)

New property types for numbers, coordinates, multi-lingual text

Is there an estimated date when these will be available? In 1 month? 3 months? 6 months?

To import entire infoboxes, these would need to be available.

If this will still take more/some time, building basic statements on items (as we do now) seems the better short term approach. --  Docu  at 18:48, 8 April 2013 (UTC)

The development will provide a planing of the next deployment, see Wikidata:Contact_the_development_team/Archive/2013/03#Communication_strategy and Wikidata:Contact_the_development_team/Archive/2013/03#Roadmap Snipre (talk) 19:19, 8 April 2013 (UTC)
Yes :) Still on my list. Not done so far because a lot of people were on vacation. This should be better starting this week. --Lydia Pintscher (WMDE) (talk) 21:39, 8 April 2013 (UTC)
Sounds good. BTW I'm looking more for a general idea of time it may take than a detailed road map. --  Docu  at 02:27, 9 April 2013 (UTC)

"Edit button" missing

Hi. I can't edit wrong interwiki links here (or anywhere else). Someone please clean up the mess and tell me what is wrong regarding no edit button. Thanks.TMCk (talk) 19:21, 8 April 2013 (UTC)

If you're looking for an "edit button" at the top of the page like on Wikipedia, then there isn't one on Wikidata. You can correct wrong links by clicking "edit" next to each link. If you cannot see the word edit on the page at all, then it may be down to the browser you are using. I've moved the ones about the chemical to Q10741852. Del♉sion23 (talk) 19:25, 8 April 2013 (UTC)
Oh well, thanks. If wikidata doesn't "allow my very common browser" You've gained and lost a new editor within minutes. Real silly thou...
Also thanks for rm. the wrong links, plenty of good ones are still missing thou. cheers TMCk (talk) 19:36, 8 April 2013 (UTC)
Wikidata is a database and is mainly defined for data storage not for data visualization. Snipre (talk) 19:45, 8 April 2013 (UTC)
That's besides the point. It doesn't mean we can't make the interface for editing the data more intuitive.--Jasper Deng (talk) 19:47, 8 April 2013 (UTC)
We still need to put work into usability, yes. And now that we've done a lot of the groundwork it is probably time to get into that as well. If you have concrete changes that would improve things then reporting them in bugzilla is a good thing. The more concrete the idea the better really. --Lydia Pintscher (WMDE) (talk) 21:46, 8 April 2013 (UTC)
(ec)That's no excuse for an edit button not showing up in a common browser, is it? I usually find my way arround but can't do when something essential is simply missing. Nobody could.TMCk (talk) 19:52, 8 April 2013 (UTC)
Actually, your browser has to be rather modern, and Javascript should be on for the best experience. One thing we might do is to increase the size of the "edit" links to make them more prominent.--Jasper Deng (talk) 19:54, 8 April 2013 (UTC)
Which browser (and version) do you use? Del♉sion23 (talk) 19:55, 8 April 2013 (UTC)
I use IE 8 and 9 (2 different machines). Same problem on both.TMCk (talk) 20:23, 8 April 2013 (UTC)
Internet Explorer is not really the best browser to use for Wikimedia sites.--Jasper Deng (talk) 20:28, 8 April 2013 (UTC)
The problem is not the browser. The problem are sites that don't manage to make their software work on the most common browsers. Don't blame it on the car if the gas is bad.TMCk (talk) 20:45, 8 April 2013 (UTC)
The problem with Internet Explorer is that it puts too much constraints on the design, so we develop primarily for other browsers. If the car isn't designed for a grade of gas, it will choke on it, plain and simple. I'm sympathetic to your view, but support for Internet Explorer just hasn't been the very highest priority.
In any case, IE10 shows the edit link correctly and prominently for me. IE 9 should too.--Jasper Deng (talk) 20:54, 8 April 2013 (UTC)
We are working on fixing the bugs in IE8 and one of the developers made progress there (should be in the next deployment here) but some of them are really nasty and outside our control. So in short: it will get better but probably never really great for all I can tell right now. --Lydia Pintscher (WMDE) (talk) 21:42, 8 April 2013 (UTC)
 
23PowerZ (talk) 21:54, 8 April 2013 (UTC)

Leave the sinking ship as soon as you can, there is a reason it's called internet exploder. 23PowerZ (talk) 21:54, 8 April 2013 (UTC)

Delete Q9683152

And then please loading the entry in Q8883307, the right place.--Kaktus Kid (talk) 21:51, 8 April 2013 (UTC)

  Done Regards, — Moe Epsilon 22:23, 8 April 2013 (UTC)
Please make deletion requests at WD:RfD in future, thanks. Del♉sion23 (talk) 01:42, 9 April 2013 (UTC)

Delete Q10340087

Observatório de Estrasburgo, include in Q947985. Best regards, --Kaktus Kid (talk) 01:35, 9 April 2013 (UTC)

  Done In the future, please make deletion requests at WD:RfD instead, thanks. -- Cheers, Riley Huntley 01:55, 9 April 2013 (UTC)

Merge Q3866264, Q6917966

Both of these cover the same topic

  • Q3866264 - Moto Guzzi Cardellino
  • Q6917966 - Moto Guzzi Cardellino

Cheers -- Brianhe (talk) 03:53, 9 April 2013 (UTC)

  Merged into Q3866264. You can take future requests to WD:RFD. Regards, — Moe Epsilon 04:51, 9 April 2013 (UTC)

A quick question

Hi,

i was asleep the last few months and am therefore pretty new to wikidata. I have a quick ontological question. The Philosopher Immanuel Kant was born in a city known at that time as "Königsberg" Q4120832. "Königsberg" was renamed after WWII and is currently known as Калининград (transcribed: Kaliningrad) Q1829. It makes sense to treat both as "different cities" since the first name refers to a Prussian city whereas the second name to a russion one. On the other Hand they are in a geographical sense the exact same city.

I did not find any guideline as how to wikidata handles these kinds of things (if it states this somewhere clearly: i apologize) and i therefore do not know what the correct value for place-of-birth (and death) for Kant should be. Perhaps someone could enlighten me :) 87.152.36.72 15:31, 6 April 2013 (UTC)

If there is no article for the city (as it was defined at the time), you could always create a new wikidata item for it. Items on wikidata do not have to have matching articles on wikipedia. But, there are policies for creating non-wikipedia items. See, Notability. Places are notable by default. Danrok (talk) 18:51, 6 April 2013 (UTC)
But how can the bot that collects empty items tell if such an items meets the notability criteria or not. In particular, if one creates an empty item with label Königsberg and no sitelinks, will this item be nominated for deletion or not? --Michgrig (talk) 20:08, 6 April 2013 (UTC)
That's the reason why administrators always have to check empty items before deletion. --Bene* talk 06:43, 7 April 2013 (UTC)
Königsberg Q4120832 looks fine to me as place of birth for Immanuel Kant. As far as I know, Wikidata:List_of_properties#Intrapersonal is our current guideline for this case. -- Make (talk) 20:21, 6 April 2013 (UTC)
You're right, Make. So, I'm not sure what the issue is with this item. Danrok (talk) 13:14, 7 April 2013 (UTC)
As I understand it, the original question was whether one should use historical instances of a place for geographical properties. In addition the post hints at the fact that the relevant guideline is difficult to find. And that the guideline could specifically elaborate on the aspect of historical (or other forms of prallel) instances. (Thanks to the original poster for bringing this up.) -- Make (talk) 15:44, 7 April 2013 (UTC)
Soon we will be able to add qualifiers to properties. We can then add the 'Official Name" property to this wikidata page multiple times, once for each name, with dates for each name, all attached to the same wikipedia and wikidata page. This will, in my opinion, be better than having multiple wikidata pages for the same wikipedia page. Filceolaire (talk) 16:05, 7 April 2013 (UTC)
It seems to me that a lot of properties need a time dependent way of filling them in. History has a way of changing facts. Cities, for example, tend to change their name, spoken language, population and nation quite often. Probably best to introduce dates so a query for "Kaliningrad" + "1724" + "primary name of town" returns "Königsberg" while "Kaliningrad" + "2010" + "primary name of town" returns "Kaliningrad" (town was renamed 4. Juli 1946?) --Tobias1984 (talk) 16:52, 9 April 2013 (UTC)

Usefulness of sources for identifiers

I think specifying sources for IMDB, VIAF, etc identifies is useless. Identifier could be correct (point to right person, work, etc) or incorrect (point to wrong one). However verification status could useful: verified manually, by bot using respective sites API and data form Wikidata, etc. --EugeneZelenko (talk) 03:41, 9 April 2013 (UTC)

I agree. The best solution would probably be to make them sitelinks like Wikipedia pages. That would also ensure uniqueness and a clear conceptual and physical separation from more complex and potentially tricky properties. Is is something a technically fit volunteer could take on ? --Zolo (talk) 07:40, 9 April 2013 (UTC)
Not in favor: again putting links in wikidata in not the best idea beacause there are often change in external DBs which imply a lot of updates and check of consistency. By providing only the identifier it is possible to create the link to the external DB on the client side using lua or wikicode. It is a mistake to implement to many features in a db. Think like for infoboxes: in the page code you provide the identifier and you use a template to create a link in the final output of the page. Yes, we can the same in wikidata but in that case you have to maintain the template on the server side. Do we have enough persons able to do that ?
@EugeneZelenko. What is the difference between an initial import of a statement and a verified statement ? The persons or the system which is doing the verification. So unless you give an accreditation to a person or bot for verification there is no difference. Then once you verified you have to fix the value in order to avoid further modification without a verification is unless as everybody can change it. Do we want to delete the editing right for some statements ?
If you really want to give an verification proof of the data quality, you have to change the concept of wikidata and give the right of data import and data edition to only a few number of thrusty users. Snipre (talk) 08:26, 9 April 2013 (UTC)
I do not see why it would be more difficult to maintain the link just once in Wikidata's code than many times in various templates. Sure, that would require access to more core-ish features, but communication between the community and the developers should be sufficiently well organised that it should not be a problem.
Data validation by trusted users may be a useful thing to have, but that is not what sources do. Sources provide an external reference to check the validity of a claim. I do not think any external source would be relevant to show that Wikidata item X == Database Y entry Z. If some day, we implement data validation, that could be done for sitelinks as well as for statements. --Zolo (talk) 08:47, 9 April 2013 (UTC)
You assume that a development team will be always available: that is a correct assumption but this is isn't fixed for eternity. Then if we start with one template which is the limit ? Who will pay for that work in the future ? How are maintained templates in wikipedia after 3 or 5 years ? Again you can expect that you will find always someone but this is not sure. As in industry a good tool is a tool which requires a minimal amount of maintenance. If you can not garantee the fundings in the future don't include them as default in your planning. Right now with a reduced development team there is plenty of work to implement the necessary tools and to fix bugs so it is good to fix priorities and to avoid asking new features. It is only my opinion but I think we are lucky by having a professional IT team paid by external funds so don't put pressure on them by asking new features every week. Snipre (talk) 12:06, 9 April 2013 (UTC)
Yes, we are luck to have a professional team, otherwise we sould not have anything yet :). I did not mean to put any pressure on them. I seem to recall that there was mention of such a feature by members of the delopment team at some point. Jeblad, who is a member of the development team mentionned some time ago that there was some time for volunteers who wanted to add additional features, and I was wondering if that was one of them. It seems like just an generalization of Wikipedia sitelinks, but I really do not know how much effort it would require in practice. Once the feature is in place I do not think that would require much maintenance. Updating an URL once in a while can be done by a volunteer. --Zolo (talk) 12:37, 9 April 2013 (UTC)

The properties of neighbours

Yesterday, I tried to collect some properties from the neighbours of an item. I tried to make a list of the neighbours of Germany, and to include the COA of these neighbours (Denmark etc). But I failed, since mw.wikibase.getEntity() couldn't collect the Entity of any other item than that of the article itself. I could easily collect the Label and SiteLink of the neighbours, but not the entity. Did I made any mistake, or is this how it is supposed to work? -- Lavallen (block) 05:43, 9 April 2013 (UTC)

To make it more obvious why this is a serious problem: Today we add only one level in P131 (is in the administrative unit). To find the next level, you go to next item and look for P131 again. As far as I can see, it's not possible to do anything like that because of the above. -- Lavallen (block) 12:58, 9 April 2013 (UTC)

Empty diff

Why some complex diffs like this looks like empty for me though they are not (at least size differs)? P.S. Opera, if it matters. Infovarius (talk) 12:28, 9 April 2013 (UTC)

After moving to the previous edit and back, I see the change:

property / Commons category / reference
imported from: English Wikipedia

A bug? --Michgrig (talk) 12:31, 9 April 2013 (UTC)

Undeletion?

Is it possible to undelete q5361186, which was an item for the genus Odobenus? This interwiki link-less item satisfied the default notability criteria stated at WD:N. --Wylve (talk) 16:53, 9 April 2013 (UTC)

  Undeleted. --Stryn (talk) 17:16, 9 April 2013 (UTC)
Don't be trigger happy in a Database! In the future this can have unwanted side effects, Items can be part of an calculation and/or is used in an article, script, $project. It is better to let this Items survive and convince / block / stop a user or Bot to make more of the same kind if WD:N is violated. We need a process to delete if necessary with marking as deprecated and find places where this item is used, but don't do it just for fun. --FischX (talk) 22:03, 9 April 2013 (UTC)
Such a place exists, and is called Wikidata:Requests for deletion. — ΛΧΣ21 00:26, 10 April 2013 (UTC)

Wikidata:Bots

Hi, I've started a discussion to mark Wikidata:Bots as an "official policy" at Wikidata_talk:Bots#Mark_as_policy. Legoktm (talk) 00:28, 10 April 2013 (UTC)

Counties of Ireland. Am I doing them right?

I've been going through all the counties of Ireland and adding missing information by following the examples that were already there. i.e. I started adding is in the administrative unit after I saw it used on one of the other counties. I want to make sure I'm using all the properties right before I continue on with the remaining counties. The last county I did was County Waterford and I'm up to 7 properties so far. 86.45.216.32 04:30, 10 April 2013 (UTC)

Looks like you're doing a great job. Why don't you create an account :) — ΛΧΣ21 05:00, 10 April 2013 (UTC)
Really? Administrative division and administrative unit are correct? As in I'm putting in the correct types of values for each one? 86.45.216.32 05:07, 10 April 2013 (UTC)
Ideally, "type of administrative unit" would link to an item about the administrative division of the Republic of Ireland. Not sure if there is one though.
"is in the administrative unit" is correctly used. --  Docu  at 05:34, 10 April 2013 (UTC)

Feedback from the Wikipedias about phase 2

Heya folks :)

I've been having discussions with a lot of Wikipedians about phase 2 deployment on their Wikipedia that is planned for next week. One of the major things that came up all the time is sources and the fear that data is being pushed into Wikidata without a source or with Wikipedia as the source (which seems to be even worse for some). I promised to bring this feedback here. I think we need to have a serious discussion about this - especially those of you operating bots. --Lydia Pintscher (WMDE) (talk) 19:56, 6 April 2013 (UTC)

I agree. I think finalizing how to import references from Empty citation (help), Empty citation (help), etc. is a top priority. If we are planning to just match each of their parameters to a property then let's do that. We have an author property of type item, but many of the authors used in references on Wikipedia are not significant enough to have their own article or item. Do we leave them out and just use titles and ISBNs? We need to make sure we have enough properties to cover web and news sources. I'll look through the existing templates and think about this some more, but I think we should probably put these in their own box on the list of properties page, including the properties that are being reused (which we should when possible). Wakebrdkid (talk) 20:30, 6 April 2013 (UTC)
Absolutely. I think we need a new data type for the references, which would match for example the Empty citation (help) format.--Ymblanter (talk) 15:44, 7 April 2013 (UTC)
Agreed, noting that it's imported from Wikipedia is only a stopgap so when there's better source support you know where to try to get the source from. It should not be allowed long-term, since when infoboxes are using Wikidata it will be inherently circular. Superm401 - Talk 22:00, 6 April 2013 (UTC)
Wikipedia as a source for Wikidata that is used as a source for Wikipedia is self-referentially. But we have the property "imported from" (Property:P143) that is useful. Imho we should change WD:N and allow books that are used as a source to have there own item (see Wikidata:Books task force). --Kolja21 (talk) 22:39, 6 April 2013 (UTC)
I think we're agreeing. 'Self-referentially' is what I meant by circular (which is a bad thing). Just citing a book item is not adequate, though. At the very least, it needs to allow page numbers, and possibly quotes, and various other things. Superm401 - Talk 23:45, 6 April 2013 (UTC)
Property:P304 (page) and Property:P387 (quote) already exist.--Stevenliuyi (talk) 23:57, 6 April 2013 (UTC)
Thanks, I wasn't aware of those. Superm401 - Talk 04:04, 7 April 2013 (UTC)
I could add some other problems that should be solved:
  • We need qualifiers if some items (like place of birth/death's name, country, etc) is used in historical content. Otherwise infoboxes will be inaccurate.
  • We need to provide sample Lua code to create links, lists, etc from properties. Otherwise each project will reinvent a wheel to solve same problem again and again. If such code exist, it should be mentioned in announcement. Same for wrapper template which selects infobox parameter or Wikidata property.
EugeneZelenko (talk) 00:07, 7 April 2013 (UTC)
Code to create list articles is exemplified on the wd:phase III page.Mange01 (talk) 19:14, 7 April 2013 (UTC)
Qualifiers will be available in one or two weeks so it is not a problem. For use of data from wikidata it is not the main task of wikidata community: we have to provide data with references. It would take more time to enter data than programming a lua template for data display in wikipedia articles.
Most properties for books are available: only language of the book is missing and a string property for authors without an item. For web pages we need url datatype property.
The problem is not properties or technical stuff but data with references. People have to take their books, their encyclopedias, and if necessary borrow books in librairies and enter manually data in excel or txt file and finally give those files to bot operators in order to introduce data in wikidata db. We can not skip once the manual work of transfering data from literature to an electronic format. Stop thinking we can extract data from internet or wikipedias infoboxes because often references are not available. Electronic databases will be the good solution but they aren't free or there are copyrights. Snipre (talk) 00:31, 7 April 2013 (UTC)
@Lydia. We already had this discussion one month ago when people were starting massive data import from wikipedia articles but without references. These people did that work because they had the possibility to do it but without thinking of the final goal of wikidata and of the wikipedia rule about sourcing data as a main task (duty?) for contributors. See Wikidata:Project_chat/Archive/2013/02#Proposal:_preventive_control_of_imported_data_correctness Snipre (talk) 00:48, 7 April 2013 (UTC)
"Most properties for books are available"? I doubt. Without naming the year a book has been published we can't start citing books, but this will be solved. The main problem is using items as a source. Items can change. If I cite a book I don't want that the name of the title depends on the language a user has chosen. Even the publisher should stay the same. If a company is bought and is changing its name the source should still remain the same. --Kolja21 (talk) 00:58, 7 April 2013 (UTC)
The solution for the book is not to cite a book, it is to cite an edition of the book, espetially if you are citing a specific page number. There could be an item by edition. This problem is adressed for example in the "Reference" namespace in french Wikipedia. But it's easier to cite a book, and in that case title remain the same. TomT0m (talk) 20:27, 10 April 2013 (UTC)
I've been working some on films. You really have to use both the infoboxes and an external database. IMDb currently has 278,000 films and the film infobox is used on 76,000 articles. I think those are the types of pairings we want to look for. We want the reputable source to have more information than meets the notability criteria of Wikipedia and Wikidata. So I'll download and filter the information from IMDb, and then I'll be able to upload enough information with sources to remove essentially all of the parameters from infobox film transclusions. Wakebrdkid (talk) 01:37, 7 April 2013 (UTC)

Please discuss abour references there: Wikidata:Sources Snipre (talk) 15:20, 7 April 2013 (UTC)

Why the rush? Can we delay phase II deployment on more Wikipedias, or postpone export of more data without source to Wikipedia? Instead I think we should focus on gathering data of better quality to avoid bad reputation of this project. Very few datatypes are still created, and because of that rather few properties. The documentation of properties at their talk pages is still poor, no list of imported data from external sources exists, and still no infobox is fully mapped (at the wd:Phase II pages). The Wikidata repository does still not show which data that is included in which Wikipedia article (like Commons shows where a picture is used). The dbpedia project has created hundreds of datatypes, properties and ontology classes, and mapped hundreds of infoboxes. I think we should reach a similar level as dbpedia, and import citation data from infoboxes, before exporting more data to the Wikipedias. Mange01 (talk) 19:38, 7 April 2013 (UTC)
The deployment is not the problem: the development can't wait on the community to develop and test the system and if the last development was so fast it was to use as much as possible the initial team. The problem is the import of data without any thoughts of the user needs. The simplest solution is to delete all statements without references and to define a limited period before data with "imported from" property (Property:P143) will be deleted too. Snipre (talk) 22:29, 7 April 2013 (UTC)
The data quality problems is partly caused by missing datatypes and qualifiers (for example for references, time series, units, multilingual data, etc). The development do not have to wait for anything, but why not stop further deployment until this is fixed?
Are there any plans on an automatic list of where the data is used in the Wikipedias? That would also be important quality mechanism before further export of data. Mange01 (talk) 15:17, 8 April 2013 (UTC)

A new default notability criteria

On 23 March, user Docu added the following to the list of items that are notable by default:

  • cemeteries, mausoleums and burial locations (if used as item for Property:P119)

The issue is that, while he proposed it on the discussion page of WD:N, nobody discussed the matter with him. Although I understand that this is not his fault, I find that the best course of action is to discuss the matter here, a more visible place, before including the item to the set of default notability criteria. Hence, I am starting this thread to determine if xonsensus exists to add the above consideration to the aforementioned set of criteria. Regards. — ΛΧΣ21 22:04, 7 April 2013 (UTC)

Is this the same user:Hahc21 who just closed a discussion after 26 hours on a weekend at "Who can assign property creation rights?" ?
That was a supplementary discussion, and it needed to be closed given that the implementation of the userright was scheduled for today, and is now deployed. So, keeping the discussion opened was useless, considerind that the changes were already made. — ΛΧΣ21 22:43, 7 April 2013 (UTC)
The criterion was proposed before on page's talk page as many others. As it wasn't controversial, I added it to the page. It hasn't really been a problem up to today either.
If you seek discussion of the criteria or if you want to remove it again, feel free to open such a discussion. --  Docu  at 22:12, 7 April 2013 (UTC)
I don't think that Hahc21 was proposing that your addition was wrong, just that there didn't seem to be any discussion about it at all. If I may, I submit that the entry doesn't need to be added to the list, as it falls under the criterion regarding items used in statements. FrigidNinja 22:18, 7 April 2013 (UTC)
I'm not sure if such a criterion has already been added somewhere (items used in statements).
BTW Hahc21 actually tried to remove it before posting here. --  Docu  at 22:25, 7 April 2013 (UTC)
The second point of the third criterion states that items are notable if they "can be connected to at least one other Wikidata item"; I take that to mean use in statements, as I can't see any other possible method of connection. I don't see a point for including places of burial specifically, as they would fall under the first default notability criterion as a place. FrigidNinja 22:29, 7 April 2013 (UTC)
Yeah, but the first point of the third criterion ends with "and".
Interesting point, I hadn't thought of that. Maybe "places" is sufficient. --  Docu  at 22:34, 7 April 2013 (UTC), edited
  • First. I'm not saying that I'm against your addition, or that it was wrong. The issue is that nobody participated. I know that this is not your fault, Docu, and that's why I brought this here. If consensus is to add it, I'll add it myself, but this kind of actions need discussion. — ΛΧΣ21 22:41, 7 April 2013 (UTC)
    • I guess we disagree on that point. 26 hours on a weekend seems too short to me. --  Docu  at 23:21, 7 April 2013 (UTC)
      • It is short, of course. But, as I explained above, the change was done. Why keeping the discussion opened if the change was already made? is it worth to make people read and vote for something that was already been handled? No. — ΛΧΣ21 23:22, 7 April 2013 (UTC)
        • Still, it's inappropriate to close a discussion after 26 hours on a weekend.
          Proposals are put for discussion to explore issues and gain consensus if in case there are diverging opinions. They are not for idle chat or abrupt closure if someone had acted prematurely (e.g. after 26 hours). It seems odd that I have to explain you such things as you seem to be lecturing others about that. --  Docu  at 17:16, 8 April 2013 (UTC)
          • I still think you do not get the point. I will say it only once more: That was a supplementary discussion that was held after the real RFC, that lasted for a long time, was closed. It was not entitled to last for too much, given that the right was already been handled by then, and we, the administrators, reached the conclusion that consensus there was clear and that having that discussion opened was unnecessary; hence, I closed it. I won't discuss this any further. Also, this discussion is about your undiscussed inclusion of a fourth default criteria that has prevented me and other admins from deleting an item that should have been deleted. You are trying to write policy without consensus, and that's something I won't allow. — ΛΧΣ21 04:05, 9 April 2013 (UTC)
            • We don't expect you to agree with us . Don't worry about it and stop soap-boxing about it. Just think that it's probably for good reason that the future bureaucrat re-open a discussion you closed after 26 hours only. --  Docu  at 06:22, 9 April 2013 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ can we stop bickering and discuss the issue at hand? --Guerillero | Talk 05:17, 11 April 2013 (UTC)

Property:P394 Vs Property:P405

"binomial authority" and "taxon author" are not the same? some can explain me what is the diffrence between them? -- yona b (talk) 16:47, 9 April 2013 (UTC)

You are right but if an expert can confirm before selection and deletion... Snipre (talk) 18:42, 9 April 2013 (UTC)
The explanation of it is held at the talk page of Property:P405. This property was deemed as necessary because some species have several taxon authors, though I'm not an expert and you should consult the users who supported the creation. Cheers. — ΛΧΣ21 13:48, 10 April 2013 (UTC)

Request for UNdelete

At what page I can ask a request for undelete an item (https://www.wikidata.org/w/index.php?title=Q10798782) because it is actually linked to many pages. 145.53.76.181 19:02, 10 April 2013 (UTC)

It has been undeleted, and I do not think we have a dedicated page, so the page for deletion requests should be fine for the time being.--Ymblanter (talk) 20:56, 10 April 2013 (UTC)

Label lister gadget gone

Am I the only one who does not currently see this gadget? I have it installed (which my preferences confirm), and yesterday it was working fine, but today it just does not appear, which prevents me from article deletion. Windows 7 and FF, if this is important.--Ymblanter (talk) 10:22, 9 April 2013 (UTC)

Works fine here. But somehow some of the gadgets (move, main language first, deletionhelper) were turned off (at least for me), so I had to turn them back on in my preferences. --Stryn (talk) 10:37, 9 April 2013 (UTC)
Yes, I also had problems with other gadgets, in particular, deletionhelper does not work for me right now, just this one is critical for what I am doing. May be I should try tu torn them off and then on again in the preferences.--Ymblanter (talk) 10:39, 9 April 2013 (UTC)
You are right, both gadgets were checked out in my preferences. Strange.--Ymblanter (talk) 10:41, 9 April 2013 (UTC)
I just had a similar issue. Numerous gadgets I had enabled were turned off (preview, DeletionHelper, NukeOnContribsLink, and autoEdit, to name a few), while at least one I hadn't had enabled (RequestDeletion) was turned on. I'm on an "Alex" model Chromebook, version 25.0.1364.173. What about you, Ym? — PinkAmpers&(Je vous invite à me parler) 00:58, 10 April 2013 (UTC)
(Interestingly, though, labelLister was working just fine for me.) — PinkAmpers&(Je vous invite à me parler) 00:58, 10 April 2013 (UTC)
I just noticed that all of my RecentChanges preferences are gone too... looks like a Special:Preferences-wide issue. — PinkAmpers&(Je vous invite à me parler) 03:02, 10 April 2013 (UTC)
Hmmm strange. I am not aware of anything that could have caused this. I'll ask Reedy though if he did any core updates or anything like that. --Lydia Pintscher (WMDE) (talk) 07:38, 10 April 2013 (UTC)
Same problem for me: I've lost many of my preferences in Gadgets section, and some have been added (Navigation popups for example). Ayack (talk) 08:42, 10 April 2013 (UTC)
I can confirm this. Some of my gadgets were unchecked when I previously checked them, and my custom signature was wiped. There might have been other removals. E.g. I'm pretty sure I put in my gender, and it's gone. Filed as bugzilla:47109 Superm401 - Talk 03:56, 11 April 2013 (UTC)
We talked about this and can't think of any change that could have caused this really. One thing that might be happening is that a gadget is saving preferences and erroring out doing that. I've been told that this would manifest in this way. Anyone able to have a look at the gadgets that have been enabled recently for any problems? --Lydia Pintscher (WMDE) (talk) 16:34, 11 April 2013 (UTC)

Import task force?

related discussion

I suppose it would be useful to import some (many) properties from DBpedia. Look at this example: France. Most of information was anyway extracted from Wikipedias infoboxes, so we'll just reuse it but in much more structured way. And it can be highly automated. Infovarius (talk) 08:03, 11 April 2013 (UTC)

It may make sense to take a closer look at DBpedia, but the example you provide is not very appealing :|. Most of what I see there is either non-sensical or irrelevant. Actually, I also have strong concerns about the quality of data imported from Wikipedias, but DBpedia seems to be much worse. --Zolo (talk) 08:51, 11 April 2013 (UTC)

Template:ExternalUse

Is there an option somewhere to check whether a specific property is being used in template in some other project (e.g wikipedia)? Use can be with {{#property}} or with Lua API. If there is no such option (yet) I think it be wise to tag properties being used by templates in other projects with Template:ExternalUse. 132.65.16.64 08:11, 11 April 2013 (UTC)

Q5166982

  Done Regards, — Moe Epsilon 08:39, 11 April 2013 (UTC)
Thank you. --тнояsтеn ⇔ 08:41, 11 April 2013 (UTC)
  Done. --Kam Solusar (talk) 09:34, 11 April 2013 (UTC)
  Done. --Stryn (talk) 13:17, 11 April 2013 (UTC)
  Done. --Stryn (talk) 13:17, 11 April 2013 (UTC)

Enough for today on IE8 ;) Thank you --тнояsтеn ⇔ 13:27, 11 April 2013 (UTC)

Google Summer of Code

Hey :)

If you're a student and interested in hacking on Wikidata over the summer please consider applying for Google Summer of Code. There are 3 organisations offering projects related to Wikidata this year. More details at Wikidata:GSoC. --Lydia Pintscher (WMDE) (talk) 08:46, 11 April 2013 (UTC)

Piruvato decarboxilasa-Pyruvate decarboxylase

This page and this page are for the same enzyme. I think it should be only one page, isn't it? Thanks. --Shalbat (talk) 09:47, 11 April 2013 (UTC)

If you are sure that they are the same, please try to merge them by yourself --Michgrig (talk) 10:01, 11 April 2013 (UTC)
Done, thanks. --Shalbat (talk) 10:21, 11 April 2013 (UTC)

RequestDeletion tool no longer available

Also asked here.

Hello,

Any reason why the RequestDeletion gadget disappeared this morning? The "Request deletion" link is no longer available in the interface. It worked just fine only 12 hours ago. Place Clichy (talk) 10:21, 11 April 2013 (UTC)

It should work now, I fixed one thing: [12]. --Stryn (talk) 11:06, 11 April 2013 (UTC)
It's back, many thanks and congratulations for the quick fix! Place Clichy (talk) 11:22, 11 April 2013 (UTC)

Introducing WikidataTrust

It's a simple script that could help reviewing content on Wikidata... --Ricordisamoa 13:32, 11 April 2013 (UTC)

This is a blame algorithm and not an algorithm for calculating trust. [Blame is the first part of the process where a user is identified as responsible for a part of text. Second part is using some method to extract reputation, often due to previous work. Third part is calculating trust for some composite text.] Jeblad (talk) 14:20, 11 April 2013 (UTC)
Of course, I took the name from WikiTrust... --Ricordisamoa 14:24, 11 April 2013 (UTC)

Rename Wikidata:Sources

Can someone rename the page Wikidata:Sources into Help:Sources ? Snipre (talk) 16:40, 11 April 2013 (UTC)

  Done[13] Vogone talk 16:55, 11 April 2013 (UTC)

Wikispecies

How is Wikidata going to affect Wikispecies?

For example, Q3072 is the item for species:Gorilla beringei graueri:

Does Wikispecies still make sense? --Ricordisamoa 16:10, 9 April 2013 (UTC)

For now wikidata have data only for the larger group (in Genus you have data for Familia, kindom but not for Species). this data is in Wikispecies for exemple [[species:Gorilla and Synonyms didn't axist too. -- yona b (talk) 16:52, 9 April 2013 (UTC)
As I see things, wikipscies could become mainly a frontend that displays data available from wikidata in a smart and appealing way (wikidata pages are not really meant to be looked at). TomT0m (talk) 09:43, 11 April 2013 (UTC)
There was a discussion on mailarchive:wikispecies-l/2013-January/thread.html by the way. --Nemo 15:45, 12 April 2013 (UTC)

Q10859555

Please add de:K.u.k. Marinefriedhof Pula and hr:Mornaričko groblje u Puli. Thank you --тнояsтеn ⇔ 11:08, 12 April 2013 (UTC)

  Done--Ymblanter (talk) 11:50, 12 April 2013 (UTC)

And de:Galtabäck-Schiff + sv:Galtabäcksskeppet to Q10860311. Thanks --тнояsтеn ⇔ 14:22, 12 April 2013 (UTC)

  Done, regards. FallingGravity (talk) 16:05, 12 April 2013 (UTC)

Visibility of change

Hi, why is this change not visible in any of the referred projects? --RonaldH (talk) 13:42, 12 April 2013 (UTC)

Purging solved the issue. In German: Die Server hängen gerade mit der Aktualisierung ziemlich hinterher. de:WP:Purge in den betroffenen Artikeln aktualisiert die Interwikilinks dann aber sofort, wenn man ungeduldig ist ;) --тнояsтеn ⇔ 13:46, 12 April 2013 (UTC)
(edit conflict) As far as I see it is visible on all projects now. Updating the wiki pages might take some minutes, but if you cannot wait you can also add ?action=purge to the URL of a wiki page. It should update immediately then. Regards, Vogone talk 13:49, 12 April 2013 (UTC)
Change propagation is extremely slow for the moment, it will hopefully be fixed. Jeblad (talk) 16:11, 12 April 2013 (UTC)

How to manage statements for organisations that are eponymous to a location?

At times there are statements which are eponymous to a location. Things like

  • a winery that has a geographical/locational basis, but is clearly an organisation (primary purpose)
  • a school which can be a location or an organisation

What is the means that people use to make the primary assessment of which of the "main type" properties to assign. I still believe that there is a paucity of direction and clear linking to the right help from active main namespace pages to assist people to 1) know that they should add statements, and 2) how to accurately add the correct statements.  — billinghurst sDrewth 00:04, 8 April 2013 (UTC)

Both, wineries and schools, are main type "organization" (please also see Wikidata:Infoboxes task force). --Kolja21 (talk) 00:33, 8 April 2013 (UTC)
I will say it again, and this time in plain language, .... Expecting anyone who wishes to add statements to have them try and head butt their way through Wikidata:Infoboxes task force YOU HAVE TO BE KIDDING ME!
That is so NOWism as an answer, and does not take into the context of how people use the wikipedias. A school building that is now empty, but is still known as the school is not an organisation. Similarly for old courthouses, ... It may now be part of a tourist attraction (formally or informally). So the direction of "organisation" is neither complete nor helpful.  — billinghurst sDrewth 02:25, 8 April 2013 (UTC)
That seems like a frequent problem. Though that may seem at times heavy-handed, I think the only clean solution is to split the item into two (assuming of course that we have things to say about both the building and the organisation). We already make the distinction for bigger things like the Metropolitan Opera or the Parliament of the UK. --Zolo (talk) 07:34, 8 April 2013 (UTC)
Which was my point. If split/added twice, we can still only link to one article per language WP. Plus in some of the WP articles they concatenate places anyway, eg. PLACENAME is a town and district of … So the advice is slightly an antithesis to the architecture. It needs direction to the punters.  — billinghurst sDrewth 10:15, 8 April 2013 (UTC)
Splitting the item in two is not a "clean solution" because it's dictated by the GND. The question isn't "how many items do we need to split off to make GND happy?". It should be "given that we want to keep this as a single item, how do we deal with GND?" As I've argued in other contexts, omitting the GND type or having more than one GND type should be viewed as acceptable solutions. Pichpich (talk) 13:49, 8 April 2013 (UTC)
Nearly always an entity has a location, but is not the location itself. A good test is if you can move the entitiy and it is still the same. But even if an entity is a location, it can be implemented as if it has a location. In my opinion a school, present or previous, is a profan building that has a location and some administrative and/or organizatoral structure. — Jeblad 13:55, 8 April 2013 (UTC)
Pichich, I would be happy if we got rid of the GND type property altogether, but still, conflating data about the building and those about the organization occupying it may make things confusing. It is not just their GND type that need to be distinguished, it is also the owner, the legal status, the creation date and probably various other things. --Zolo (talk) 14:08, 8 April 2013 (UTC)
Wikipedia pages are about places and, in general, recount the history of the place as it's use changes over time, all on the same page. There is no reason why wikidata pages cannot follow the same pattern with the same place having different uses, each with a start and finish date (using qualifiers when we get them). Can you even find an example of a place with separate wikipedia pages for the winery that was and for the place that is? Where there are two pages one will deal with the organisation and the other page will deal with the place including it's history back to before the organisation arrived. No problem. Filceolaire (talk) 12:25, 12 April 2013 (UTC)
Indeed; I worked at Bow Street Magistrates' Court, which is both a place and an organisation. I happened to work for a different organisation whilst there, so the statement "worked at" is one with geographic, not organisational, domain in this context - but under current policy, we won't have entities for both the place and the organisation, merely because Wikipedia (as is rational for a prose-oriented work for humans) won't have multiple such articles. The solution is to change our policy on entities. James F. (talk) 11:34, 13 April 2013 (UTC)

Classes of orders and decorations

How can I assign different classes of the same order as a statement within Property:P166? For instance: businessman de:Richard Altvater received three different classes of the German Order of Merit Q21164? May I create new datasets for every class, even though they only exist as redirects (de:Bundesverdienstkreuz 1. Klasse)? -- 79.168.51.74 06:49, 10 April 2013 (UTC)

One property, 3 qualifiers Snipre (talk) 21:25, 10 April 2013 (UTC)
  • If there's some hypothetical award Great Ribbon that has three different variants (i.e. classes) Great Ribbon with Star, Great Ribbon with Star and Stripe and Great Silver Ribbon, and some person John were given those three awards, I think using award received (P166) and subclass of (P279) would make sense. For example:
  • Great Ribbon subclass of award
  • Great Ribbon with Star subclass of Great Ribbon
  • Great Ribbon with Star and Stripe subclass of Great Ribbon
  • Great Silver Ribbon subclass of Great Ribbon
  • John instance of person
  • John award received Great Ribbon with Star
  • John award received Great Ribbon with Star and Stripe
  • John award received Great Silver Ribbon
This seems better to me than specifying those subclass relations via qualifiers. Qualifiers seem more appropriate to use when specifying, for example, the date on which the award was given, etc. Emw (talk) 23:59, 10 April 2013 (UTC)
Ok ...don't know, if I really got that. Could you please show me in the given example "Richard Altvater" (Q2148964) what to do? He received the Federal Cross of Merit 1st class in 1978, the Grand Merit Cross with Star in 1988, the Grand Merit Cross with Star and Sash in 1993. How do I add this information to his statements? --79.168.51.74 17:13, 11 April 2013 (UTC)
Qualifiers are not yet available, sadly, so you cannot do all of that right now. However, for the others, you would right now have to:
  1. … create an item for Verdienstkreuz, sub-typed of Bundesverdienstkreuz using property P279 ("subclass of");
  2. … create an item for Verdienstkreuz 1. Klasse, sub-typed of the new Verdienstkreuz using property P279;
  3. … create an item for Großes Verdienstkreuz (the award class not the award itself - you'd want to create that seperately), sub-typed of Bundesverdienstkreuz using property P279;
  4. … create an item for Großes Verdienstkreuz mit Stern, sub-typed of the new Großes Verdienstkreuz using property P279;
  5. … create an item for Großes Verdienstkreuz mit Stern und Schulterband, sub-typed of the new Großes Verdienstkreuz using property P279;
  6. … add to item Richard Altvater the property P166 ("award received") to the new Verdienstkreuz 1. Klasse;
  7. … add to item Richard Altvater the property P166 to the new Großes Verdienstkreuz mit Stern; and
  8. … add to item Richard Altvater the property P166 to the new Großes Verdienstkreuz mit Stern und Schulterband.
Later, come back to the item to add in the qualifiers to say what year in which they were granted.
Hope this helps! Ideally you'd add the entire award tree as you went. :-) James F. (talk) 11:59, 13 April 2013 (UTC)

Statements

I saw in the blink of an eye, a Swedish translation of the word "Statements" in an item. (It dissapeard very fast, maybe lost in the cache.) The word I saw was "Uttalanden", I think. Do we have any other translations of this word into any other languages, so I can see which translation would fit best? "Uttalanden" maybe is good, but I can imagine some alternatives. -- Lavallen (block) 17:49, 12 April 2013 (UTC)

Try Wikidata:Glossary and especially Norwegian (bokmål), Norwegian (nynorsk) and Danish. [Seems like the translations are messed up again.] Jeblad (talk) 15:47, 13 April 2013 (UTC)
  • Statement → utsagn, utsegn, udsagn (something where you usually can give a reference to a previous source, used in our items)
  • Claim → påstand, påstand, påstand (something where you usually can't give a reference to a previous source, would be used in our properties)
This is one of the places were the implementation is a bit odd, as both the statements in the items can be without sources and the claims in the properties should have sources in some cases. So the a statement should be just a claim that has a source through composition, not through inheritances. It has no implications for items, but it could have implications for properties – but we are not there yet! Jeblad (talk) 16:15, 13 April 2013 (UTC)

Q10862947

Q10862947 appears to contain links to somebody's userpage. I thought that was against consensus, no? Jakob Megaphone, Telescope 11:16, 13 April 2013 (UTC)

I have   Deleted the page. The next time you can report it at WD:RFD. -- Bene* talk 11:38, 13 April 2013 (UTC)

Duplicate

Q2152415 and Q7273287.--Kaktus Kid (talk) 15:18, 13 April 2013 (UTC)

New template (again!)

There's still no consensus to move User:Ricordisamoa/Datatype to Template:Datatype? --Ricordisamoa 15:36, 10 April 2013 (UTC)

If I go to the page, I see an error message and nothing else (previously, it was an empty page). May be you could briefly explain what it is about and how it is expected to work.--Ymblanter (talk) 15:40, 10 April 2013 (UTC)
The error message is by default... you can view a testcase here. --Ricordisamoa 15:46, 10 April 2013 (UTC)
Looks good to me, though it could use a {{documentation}} :) Legoktm (talk) 04:16, 11 April 2013 (UTC)
I would rename 'wikibase-item' to 'item' (no need to disambiguate), and 'commonsMedia' to 'commons-media' (consistent casing) but otherwise it looks good. Superm401 - Talk 05:13, 11 April 2013 (UTC)
I think he used those since thats what you get if you use wbgetentities on the property. Legoktm (talk) 06:01, 11 April 2013 (UTC)

I've   Done the move: it's already successfully used in Template:List of properties/Row and Template:Property documentation. --Ricordisamoa 19:46, 13 April 2013 (UTC)

Categories for property creators

Should we have, for example in Property talk:P44, the Category:Properties by Viscontino, etc.? --Ricordisamoa 15:48, 13 April 2013 (UTC)

No, I think it does not make sense. And who started the properties, are not owners of properties. --Stryn (talk) 16:23, 13 April 2013 (UTC)
No, such categories are a lousy idea, and would pretty much serve to indicate page ownership. Courcelles (talk) 18:30, 13 April 2013 (UTC)

Ok, ok. --Ricordisamoa 19:43, 13 April 2013 (UTC)

Not only a bad idea, it's also unnecessary since it's possible to query properties creation by user. However, I don't understand why this works e.g. for me, but not for Viscontino ?? --Nightwish62 (talk) 20:04, 13 April 2013 (UTC)
Things drop out of Special:NewPages after 30 days, I believe. Courcelles (talk) 20:22, 13 April 2013 (UTC)

Mark WD:Tools for translation

I found that many interesting tools are linked only from the English version of the page, and (for example) WD:Strumenti is very outdated. Could we use the Translate extension to keep all of them? --Ricordisamoa 13:52, 11 April 2013 (UTC)

I'd suggest no, until the bugs for Translate extension is worked out over here. As an example, I got motivated one day and translated 99% of the glossary and look at it now. I do a lot of translating at translatewiki using the same extension but I've never seen anything like this than what I've seen here.--Lam-ang (talk) 14:55, 11 April 2013 (UTC)
There should be no problem with marking it as most sections do not change. I will have try. Regards, Vogone talk 15:00, 11 April 2013 (UTC)
FYI...If you are looking for your old translated Wikidata tools they are here.--Lam-ang (talk) 14:11, 14 April 2013 (UTC)

New features

Is there a central place where we can propose and discuss new core features for a potential Phase 4 (and beyond)? Something like the RFC system, but for this purpose specifically. Project chat isn't really appropriate since it's generic and things get archived out of sight. Silver hr (talk) 16:20, 12 April 2013 (UTC)

I wonder if this call for some sort of village pump page? ·Add§hore· Talk To Me! 11:54, 13 April 2013 (UTC)
There is already such a page at meta:Wikidata/Notes/Future, mainly about using Wikidata on other sister project like Commons, Wikisource, Wikiquoute etc. HenkvD (talk) 07:58, 14 April 2013 (UTC)

Gadget MediaWiki:Gadget-markAdmins.js and translation administrators

I think it may be convenient to add translation admins to the gadget. Any objections? --Michgrig (talk) 16:25, 13 April 2013 (UTC)

  • Nt really sure how useful this gadget is, but that would fit with its usefulness. Courcelles (talk) 18:41, 13 April 2013 (UTC)
  • I don't think translation admins should be added to this gadget because IMHO their rights aren't more extensive than those of rollbackers or property creators; they just have some additional buttons within the translation interface. Regards --Iste (D) 19:47, 13 April 2013 (UTC)
    I disagree with "their rights aren't more extensive than those of rollbackers or property creators". Admins, bureaucrats, etc. are important for other community members, as they have some extended permissions for fulfilling some tasks. Translation admins are as well. If someone wants to translate a page which isn't marked he must search for a translation admin first. If a user wants to undo a change, he does not need to contact a rollbacker. That's a difference. Regards, Vogone talk 20:27, 13 April 2013 (UTC)
    To be fair, one can as well just add translate tags and then wait for a translation admin to approve, rather than try and contact one of the dozens members of the group. At least on m:TR this is the recommended process. --Nemo 14:52, 14 April 2013 (UTC)
    But even in this case, at least a translation admin should be able to tell who added the tags: their colleague or an ordinary user. --Michgrig (talk) 15:43, 14 April 2013 (UTC)

Reverse of membership properties

I've just responded to a question on the English Wikipedia Help desk where somebody is trying to insert an interwiki link from en:Cain and Abel to fr:Caïn, and it fails because Q205365 already points there. This is right - he is trying to link to pages which are not on the same topic - but it occurred to me that it might be helpful if Q717996 - "Cain and Abel" had a statement linking it to Q205365; so I was looking for an appropriate Property. I found Help:Basic membership properties - and indeed "Cain" already has a "Part of" (P:P361) relation to "Cain and Abel". But I couldn't find a reverse property. Is this because

I think it's "Nobody's made one yet". We do have "subdivides into", but this is mainly for geography. --  Docu  at 08:10, 14 April 2013 (UTC)

Structural choices

A second topic related to the previous one : Wikidata is currently stucturing, this is a very insteresting time, but also a very dangerous one. Choice we make now might not be easy to change tomorrow if we are not carefull enough. I am the only one to be a little confused about what is happening right now, for example related to GND versus more flexible ontology tools ? What are the implications on theses choices ? Are they robusts ? TomT0m (talk) 10:35, 14 April 2013 (UTC)

Wikidata 1.0 ontology ?

Early adopters already built tools from wikidata datas, but they might broke. At some point will we provide guarantees that built tools might not broke without warnings ? TomT0m (talk) 10:35, 14 April 2013 (UTC)

A stable "Wikidata 1.0"? =) Jeblad (talk) 12:48, 14 April 2013 (UTC)

Disambiguation pages: link same word or same meaning?

Q225822 (West (disambiguation) currently links to Czech Wiki West (disambiguation) (explaining the English word) and not the Czech word for west: Západ (disambiguation). Which is correct? Sorry if I didn't find this in the help pages. --Tobias1984 (talk) 19:15, 14 April 2013 (UTC)

For disambiguation, link the same "string" or sequence of characters,
for all others: the same meaning. --  Docu  at 19:26, 14 April 2013 (UTC)
Thankyou ;) --Tobias1984 (talk) 19:28, 14 April 2013 (UTC)

Exclusion criteria: files (done)

Right now we get a lot of new items like "File:Toy Shop.jpg" (Q10932531). I think we should put files on the exclusion criteria list. --Kolja21 (talk) 22:50, 14 April 2013 (UTC)

See Wikidata:Requests for comment/Inclusion of non-article pages 2. Legoktm (talk) 22:55, 14 April 2013 (UTC)
Thanks! --Kolja21 (talk) 23:10, 14 April 2013 (UTC)
Mais pas de copyright Cover art. Cela n'est pas permis à les Communes. Farrajak (talk) 01:15, 15 April 2013 (UTC)

Duplicate

Q6519525 = Q10317027.--Kaktus Kid (talk) 03:48, 15 April 2013 (UTC)

Notability of items

Hi folks, I've found Q10513452 which is about a version of the GNU General Public Licence. I am not sure if we should keep it because currently it does not meet the notability criterias. However it has four statements and has lots of information we shouldn't waste. An opportunity would be to say that every item is allowed if it has enough good statements, too. What do you think? Regards, -- Bene* talk 14:15, 6 April 2013 (UTC) PS: There are also Q10513450 and Q10513445. --Bene* talk 14:21, 6 April 2013 (UTC)

In general I agree that "every item is allowed if it has enough good statements", but I worry this would frequently provoke arguments on whether some statements are "good" or not. --Stevenliuyi (talk) 14:50, 6 April 2013 (UTC)
Support, current notability rules are no longer suitable, and there does not seem to be any good alternative. However, I think we can keep the rule that items without any sitelink should have at least one sitelink from one other item, that wuold ensure that items are well connected to the main graph, and that they have use. --Zolo (talk) 09:40, 7 April 2013 (UTC)
I agree with Jeblad that the above items are good examples, but we also should think how to prevent bad items. For instance, I can create an item for myself and an item for my brother (both with "enough good statements") and link them together with P:P7. Do these two items fail the criteria? --Stevenliuyi (talk) 19:56, 7 April 2013 (UTC)
That's a good question. In general we should define some notability criteria for persons. We shouldn't have 7 billions of items about every human currently living. ;-) --Bene* talk 19:59, 7 April 2013 (UTC)
I think GPLv1, GPLv2 and GPLv3 are rather good examples of succession and instantiation of the GNU General Public Licence and should be kept. The notability criterias is more or less borken and has been so since the beginning. — Jeblad 12:34, 7 April 2013 (UTC)

I would propose to rewrite WD:Notability thus:


Wikidata has two main goals: centralizing interlanguage links across Wikipedias and serve as a general knowledge base for the World at Large. An item is acceptable iff it fulfills at least one of these two goals, that is if meet at least one of the criteria below:

  • It contains at least one valid sitelink to a Wikipedia page. To be valid a link must not be a talk page, nor a MediaWiki page, nor a special page, nor a user page, nor a subtemplate.
  • It refers to a clearly identifiable conceptual or material entity. The entity must be notable, in the sense that serious and publically available sources have written about the item. If there is no item about you yet, you are probably not notable.
  • It fulfills some structural need, ie, if they did not use it, statements made in other items would be less relevant.

Note that these rules are not engraved in stone. In case of disagreement about the notability of an item, the case can be brought up at Wikidata:Requests for deletion.

Sure, that is a bit fuzzy, but I think it is better than what we currently have. --Zolo (talk) 07:47, 9 April 2013 (UTC)

  Support. The 3rd point is new, and is important. Otherwise we will have an inaccurate properties usage and the "phase 3" of wikidata will be very disappointing. The 2nd point is quite fuzzy, maybe it could be improved. --Gloumouth1 (talk) 08:16, 9 April 2013 (UTC)
  Support, well summarized.--Ymblanter (talk) 08:38, 9 April 2013 (UTC)
  Support best solution so far, and #3 is better than creating a long list of exceptions. --Ch1902 (talk) 10:52, 9 April 2013 (UTC)
  Comment Does "serious and publically available sources have written about the item" include trivial mention? Or there should be significant coverage of the entity?--Stevenliuyi (talk) 21:32, 9 April 2013 (UTC)
Rereading my wording, it actually sounds a bit odd: perhaps something like: "serious and publically available sources have described the item" ? That would imply that at least the subject has been described as such, not just mentionned in passing. That is still a bit vague, but note also that the sourcing policy should prevent non-obvious claims that are not supported by good references. --Zolo (talk) 08:59, 10 April 2013 (UTC)
ok I see your point.--Stevenliuyi (talk) 14:12, 10 April 2013 (UTC)
  Support I think it's the best we can do for now. Maybe there is still a lot of room for improvement, but we need to adopt and use it first to see how it goes.--Stevenliuyi (talk) 14:12, 10 April 2013 (UTC)
  •   Comment I doubt it's a good idea to start creating items for user pages.
    "structural need" is a bit vague .. besides, a statement is not less relevant if one uses multi-lingual text instead of item as datatype, but it would just be "bad structure". --  Docu  at 03:51, 10 April 2013 (UTC)
I had forgotten user pages, added now.
Yes, "useful" would actually be more accurate than "relevant". I suppose we should have a help page to explain why items are more usable than strings.
We should probably have more precise definitions of our "structural needs".But as it seems next to impossible to have general bright-line rules for now. I would rather stay with some very simple, though vague, policy guidelines, and add more details for specific cases when we encounter specific issues. Also, for now it might be more instructive to publically discuss controversial cases in requests for deletion than have precise a priori rules. --Zolo (talk) 08:59, 10 April 2013 (UTC)
In one way or the other, I think we still need a list of items that fit our "structural needs". Obviously, we can't possibly define all of them in advance.
Maybe there is also a better term than "notability" for this. --  Docu  at 18:12, 10 April 2013 (UTC)
Actually there were at least two examples discussed in Wikidata talk:Notability: nobility titles and occupations. I think lists like those can be very numerous and difficult to forecast. On the other hand, those all items in these categories are probably acceptable under criterion 2.
I would prefer something like "WD:Acceptable items", but somehow, there does not seem to be much support for changing the name. --Zolo (talk) 18:41, 10 April 2013 (UTC)

  Done, see WD:N. I really agree that the name "notability" is bad. That really sounds like insiderese for Wikimedians and does not reflect what the page is about. Can I change to WD:Accepted items or something ? --Zolo (talk) 08:26, 15 April 2013 (UTC)

Data imports

I wanted to add too things that does not seems to be taken into account in this proposition :

  • IMHO, wikidata could be used for other needs than to be used as Wikipedia or Mediawiki datasource, for example a BBC employee [14] uses DBpedia as as a reference identifier for things like music peaces. I think just Wikipedia appearance and notability criterium are a bit restrictive for that matter.
  • Wikidata could, with agreements with other organisms or projets (eg. Open Street Map) import massively data from other Databases. In these usecases, there could be a large number of items in those databases that does not require a wikipedia articles yet make sense in other context. Adress notability item by item in these cases should need an undoable amount manual of work ...

I think notability criterium, in those cases, are not on the items (or properties) individually but maybe on the entire dataset ... It seem more praticle and easyer to maintain. Plus their is at least one source of all the claim imported by mass import, or at least one, is easy to find and maintain, which could be also (and actually is also on wikipedias) a very important thing for acceptability of datas.

Yet I think this is an important decision as it is related to the future of the project as a whole : one of the proclaimed goal of wikidata is to be a technical database for Wikipedias and mediawiki, but it is also to be a knowledge database ... should this database restricted to (current) mediawiki projects data or extends the scope ?

TomT0m (talk) 14:41, 9 April 2013 (UTC)

Hello,the purpose of the above proposal is precisely to allow items even when that have no immediate use in other Wikimedia projects. About the specifics of your proposal, we could have rules like the item is considered notable if it has en entry in database X (just like it currently is when it has an article in Volapük Wikipedia). However, I think that such a rule would only make sense if the database is considered a valid source of data, and in this case, the item would be considered notable under criterion 2. --Zolo (talk) 08:59, 10 April 2013 (UTC)
Let's take Open Street Map as an example. It might have a considerable amount of items that are not notable enough to get an article on wikipedia, such as bus stops. Without having an answer in mind, would you think these criterium leave room or exclude mass import of bus stops, or leave the choice to community ? TomT0m (talk) 09:31, 10 April 2013 (UTC)
I think being in OpenStreetMap would not be considered suffcient according to the above criteria, but it might be ok to have items about bus stops if the bus operator publishes a detailed list of them. That said, Openstreetmaps seems to be both an important and a rather special case, I guess it will require some special treatment at some point. --Zolo (talk) 11:25, 10 April 2013 (UTC)
I think Wikidata is a little young to know that, these last years datasources seems to become more and more various, eg. for example there is more and more Open Data portal from public authorities, whose data are more or less not easily usable directly such as excel datasheets. Wikidata could be a great opportunity to effer them in a more structured and usable way. Yet the nature of these data are really diversified, various kind of statistics on various kind of matters for example. I think we do not really know at this point what could happen, so criteria at this point must not take the risk to close too many doors and discourage some usecases that it could manage, that's my concern. Anyway this version does not seem too restrictive, maybe we should add something about public Open Data portals for example ? TomT0m (talk) 11:58, 10 April 2013 (UTC)
The notability criteria can be loosened later on if there is a need. For opendata, I suspect the main issue would not be so much item notability as finding the right properties and mapping large spreadsheets onto Wikidata statements. There are also some licensing issues. --Zolo (talk) 13:20, 10 April 2013 (UTC)
I do not advocate for importing anything of course. I must admit I'm not sure about being too restrictive about property adding either (it is up to community of course, but there is always somebody who will have something against a property, I'm afraid a too restrictive process will tend to the minimal set of usefull for everybody property set, which would drastically limit the potential of wikidata). TomT0m (talk) 13:43, 10 April 2013 (UTC)

"Namespaces" for items ?

Hello, from the vote discussions above, I understand that the database could have a third role : pure technical proxy for mediawiki projects. There is already a mechanism to discriminate "project pages", "discussion pages", and main in wikipedia for example. Would it be possible to do something to discriminate "project items" and "knowledge items", either with a namespace or with a (set of) relation(s) ? This would clarify role of the database and allow some pure techical usecases that might ome up without pollution of the real datas. TomT0m (talk) 09:31, 10 April 2013 (UTC)

One item may contain links to pages from different namespaces (for examples, there is Annexo in es-wiki). How are you going to deal with that? Infovarius (talk) 17:00, 10 April 2013 (UTC)
TomT0m: I think that is more or less supposed to be integrated into P:P107.
It seems a little weird to me, P107 seemed to be a standard preexisting ontology, I did not know it was supposed to evolve at all, and take a central place in Wikidata and Mediawiki. Do tou have pointer to relevant discussions about that ? I would advocate onto having our own system, not trying to twist another standard. Anyway classifying by property would be fine, the main information system just would have to be aware of that. TomT0m (talk) 19:05, 10 April 2013 (UTC)
Classifying by property seems the only technical solution we have, but there are certainly other possibilities than GND. I saw you found one of the relevant threads, good luck with that ;). --Zolo (talk) 21:18, 10 April 2013 (UTC)
Infovarius: I think that a link to a non-mainspace Wikipedia article indicates that the item is not a "knowledge item", but items may have a mainspace link and still be non-knowledge. For instance, I do not think an item linking to lists should contain real data (the list itself should be done by a query, not by writing everything done in a statement). Articles like demographics of the United States should probably not contain substantial data either, as discussed in #Should the big ones get all the info ?. --Zolo (talk) 18:41, 10 April 2013 (UTC)
Or templates and possibly many other things, citation to a particular page on some book maybe, ... TomT0m (talk) 19:05, 10 April 2013 (UTC)

April fool's

It's not important right now, but down the road it might be. I'm wondering whether to permit or ban April Fool's jokes on Wikidata (keeping them out of the article namespace). Jakob Megaphone, Telescope 21:47, 8 April 2013 (UTC)

Technically, there are no April Fool's jokes on (say) en.wiki except in project space where they are tolerated as a mild annoyance. The main page is April Fool's themed in a fun or childish way depending on who you ask. The Wikidata mainpage is not used in this way so there's no issue there. Now, should we let users (of a multilingual project already straining to keep it truly multilingual) write bogus deletion requests and bogus user permission requests to mark a tradition celebrated by a handful of people in the world? No. Pichpich (talk) 00:12, 9 April 2013 (UTC)
Which article namespace? — Jeblad 00:38, 9 April 2013 (UTC)
As an alternative proposal, one could contain the jokes within the English-language sections of the project (e.g. the English language project chat, the talk pages of users who have specified willingness to be part of it). Jakob Megaphone, Telescope 12:39, 9 April 2013 (UTC)
A proposal? We're not going to write a guideline codifying what is an acceptable April Fool's joke on Wikidata. Is it really that much to ask people to avoid them? It's not like they're an irrepressible urge. Pichpich (talk) 13:46, 9 April 2013 (UTC)
I guess I wonder what's the harm in people who like them doing something for one day? If they clean up after themselves then I'm fine with pretty much anything (Within common sense, of course...). Ajraddatz (Talk) 23:22, 9 April 2013 (UTC)
We can't go overboard. In particular, BLP violations can't happen.--Jasper Deng (talk) 23:32, 9 April 2013 (UTC)
If they clean up after themselves is a pretty big if. - Soulkeeper (talk) 14:18, 11 April 2013 (UTC)
Possibly we could pre-approve a small number of jokes before-hand and list them somewhere so we can keep track of them, then remove/revert/delete them on April 2. Jakob Megaphone, Telescope 15:02, 11 April 2013 (UTC)
Nope, it's better just not to make any major decisions on April 1.--Snaevar (talk) 17:30, 15 April 2013 (UTC)

Property qualifier release on Wednesday

I had read on the project status update (http://meta.wikimedia.org/wiki/Wikidata/Status_updates/2013_04_12) that property qualifier would be release next Wednesday. How would this effect the property management? If you want to create a property we need to create a property proposal and have it be reviewed for at least a week. Only during a consensus is meet would the property be created by an administrator or a user with a property creation access. Since a qualifier is similar to a sub property definition. Should the qualifier for each property be proposed using same rules? Also I'm assuming that adding qualifier is not via the property page but via the entity page? Any tutorial on how this will work? I think qualifier should also be managed in the Property list page. Is this available in the test environment? --Napoleon.tan (talk) 18:31, 13 April 2013 (UTC)

Yes you can try it out on the test system at http://wikidata-test-repo.wikimedia.de/wiki/Testwiki:Main_Page --Lydia Pintscher (WMDE) (talk) 18:52, 13 April 2013 (UTC)
I agree. Qualifier is nothing other than a sub property. In principle you can use all properties as qualifier. Therefore I don't see any need to handle it other than properties. In my opinion they should be proposed the same way. --Nightwish62 (talk) 19:40, 13 April 2013 (UTC)
Here the property 'place' is used as normal property, while here it is used as qualifier. --Nightwish62 (talk) 19:53, 13 April 2013 (UTC)
While qualifiers add information about a property's use in a claim, I don't think they're quite subproperties (see rdfs:subPropertyOf). Examples of a subproperties include property 'father' being a subproperty of property 'parent', property 'sister' being a subproperty of property 'sibling', etc. Being a subproperty means that all claims involving a property X are also claims involving Y, if X is a subproperty of Y. Built-in support for subproperties would be great, but I don't think qualifiers are it. Emw (talk) 20:59, 13 April 2013 (UTC)
I tried out the qualifiers on the demo system last week and found that {{#property:Pxx}} returns all values regardless the qualifier or not. For the last few days cannot check it as the demo client gives an internal error "Could not acquire 'client:messages:en:status' lock". Is there a way to make sure the clients can handle the properties with qualifiers? I strongly recommend NOT to use qualifiers on Wikidata until the clients (like Russian Wikipedia) are ready for it. HenkvD (talk) 07:50, 14 April 2013 (UTC)
You can filter them with Lua, even if #property cannot handle them yet. -- Lavallen (block) 09:23, 14 April 2013 (UTC)
Can somebody prove this to me? HenkvD (talk) 11:26, 14 April 2013 (UTC)
Wikidata test Client is back online. Jeblad (talk) 13:41, 14 April 2013 (UTC)
Thanks. I continued testing and found the Lua solution here: if entity.claims[var1][i].qualifiers then .. HenkvD (talk) 17:53, 14 April 2013 (UTC)
Qualifiers are like an and on a specific claim (the claim where it is used), or a statement as it would be in an item, and makes it possible to do additional refinements. One of the purposes would be to increase the expressiveness of a limited set of properties. For example to say that a mayor only held the office for a limited time period, where the limited period would go in the qualifier. It is similar to doing a reification (in rdf) of the claim, and then making a new claim (the qualifier) on that reification. Jeblad (talk) 10:51, 14 April 2013 (UTC)

Thanks for the demo system link. I guess since qualifier for property is just another property, the proposal is the same. but i think there should be a rule to which qualifier can be added to a certain property. For example, for the property "release date", there should be a proposal to use "place" and "model" property as qualifier before letting other people use it. it would be good if this constraint can be enforced via the property definition by software. Also, we should also add the valid qualifier for each property in the List_of_properties to manage it properly. --Napoleon.tan (talk) 09:02, 15 April 2013 (UTC)

Generalization of properties

Is there consensus for merging quantity symbol and chemical symbol into a more general property?

Otherwise, we should create also Unicode astronomic symbol... --Ricordisamoa 17:18, 14 April 2013 (UTC)

Against merging with chemical symbol. And why not a new property Unicode symbol ? Snipre (talk) 18:08, 14 April 2013 (UTC)
"Unicode symbol" for "Unicode astronomic symbol" and "quantity symbol"? --Ricordisamoa 19:53, 14 April 2013 (UTC)
I don't see any reason to have different properties for unicode symbols: imho it is a good example of generic property. --Paperoastro (talk) 07:50, 15 April 2013 (UTC)

Help for IE8 user

Once more, please

--тнояsтеn ⇔ 06:49, 15 April 2013 (UTC)

  Done. FallingGravity (talk) 07:50, 15 April 2013 (UTC)
Thank you. Now nah:Coyōtepēc should be added to Q1038679. --тнояsтеn ⇔ 07:55, 15 April 2013 (UTC)
  Even more done :) FallingGravity (talk) 08:15, 15 April 2013 (UTC)

BWV 112

[15], two entries need to be combined, my mistake. Could I do it? how? --Gerda Arendt (talk) 09:08, 15 April 2013 (UTC)

You can find instructions here. --Sixsi6ma (talk) 09:44, 15 April 2013 (UTC)
Thank you, --Gerda Arendt (talk) 11:41, 15 April 2013 (UTC)

Q11000000

Q10999998→Q11000001.... no Q10999999 and Q11000000.... Special:NewPages tntchn Comment · Contribs 10:58, 15 April 2013 (UTC)

See Wikidata:Contact_the_development_team#Q11000000_missing.3F. --Stryn (talk) 11:04, 15 April 2013 (UTC)

Item

How'll I know whether an item is in Wikidata or not?--Pratyya (Hello!) 11:01, 15 April 2013 (UTC)

You can search via Special:ItemByTitle or just use the search box. --тнояsтеn ⇔ 11:03, 15 April 2013 (UTC)
There's also a bookmarklet... --Ricordisamoa 11:23, 15 April 2013 (UTC)

Soxred93

Can i get a site like http://toolserver.org/~tparis/pages/ to found out how many pages i have make on wikidata?--Trade (talk) 11:38, 15 April 2013 (UTC)

New Gadgets

Do you think that any of the scripts listed in WD:Tools is reliable and useful enough to became a gadget? Please comment below. --Ricordisamoa 11:34, 15 April 2013 (UTC)

The list needs a cleanup. I use(d): LabelLister, DeletionRequest, Display Wikidata Info on Wikipedia, and AuthorityControl.js/authority_control.js. Some of the tools are already gadgets and in the case two tools have the same name they should be numbered. --Kolja21 (talk) 13:49, 15 April 2013 (UTC)
please add
KeyShortcuts
Sitenames in user-language
iwconflict.js
Yamaha5 (talk) 19:59, 15 April 2013 (UTC)

Audrius Beinorius

Trying to create item for Audrius Beinorius on Lithuanian and English Wikipedias. Searching on Special:ItemByTitle turns up nothing, but when I go to create the item, it gives and error and says it already exists. Confused. Please help? Thanks, Renata3 (talk) 23:06, 15 April 2013 (UTC)

Sometimes this error appears even if there is no item. Just try to save your changes again. --Michgrig (talk) 23:27, 15 April 2013 (UTC)

Duplicate

Q8219076 and Q8974116.--Kaktus Kid (talk) 02:52, 16 April 2013 (UTC)

Dates of birth/death

How come they are not among available properties? That would be the first thing that comes into my mind to add to people's items. Was this discussed and rejected? Renata3 (talk) 04:31, 16 April 2013 (UTC)

They were discussed and approved. They're waiting for the TimeValue datatype to become available. 86.45.216.32 04:41, 16 April 2013 (UTC)

Proposal: allow interwiki links to link to redirects located on Wikidata

Although the usage is probably going to be limited, what I have in mind is for usage on the Japanese Wikipedia, which has a problem that they cannot link to other languages of articles that they have due to the fact that they are titled after the names of convicted criminals, and release of those names is not allowed under Japanese law. What is the problem is not so much the link itself, so much as it is that the link contains the name of the convicted criminal due to the title of the page. This problem has been resolved in the past by creating redirects (on the Wikipedia that contained the link) that did not contain the name of the offender, and making interwiki links to those redirects rather than to the articles themselves so that they can link to the article without needing to include the name within the page. Now that we are doing interwiki links with wikidata, I would want to consider how this could be done for wikidata: for example, perhaps by adding a "/(enter language code here)" after the page for the wikidata entry, it would create a page that would redirect to whatever page wikidata currently links to for that entry, and a new parser function could be added that can make the interwiki link go towards that redirect rather than directly to the page, which solves the problem for the current Japanese Wikipedia pages.

What do you guys think?--New questions (talk) 02:29, 9 April 2013 (UTC)

This has been discussed, see this Request for comment. Not sure how or if the stated outcome is being implemented, though. FallingGravity (talk) 05:08, 9 April 2013 (UTC)
It does not seem to be about the same topic at all. The discussion you linked is for the case when there exists section redirects but not whole articles in some language version of the redirect. What I am discussing here is if the complete article exists, but cannot be included as an interwiki due to the fact that the target article has a title that goes against the policies of a certain Wikipedia (for example, if it states the name of a convicted criminal). However, it should be fine to link to a redirect to the article, since that does not go against the policy, and I am proposing that the Wikidata itself can host the redirect page by automatically making pages like "Q####/en" redirect to the en entry linked in Q####. This is a case that should be simpler to implement.--New questions (talk) 07:36, 9 April 2013 (UTC)
Well, we need an alternative solution, as creating redirects in the main namespace of Wikidata does not work. Only items are allowed there, and even the frontpage is in the Wikidata namespace.
Maybe linking the japanese article to an softredirect would be that alternative. You see, Wikidata items do reject links to redirects, but not to softredirects. This does have one downside though. Lets say there is an item with an link to an japanese article and an english page with an softredirect. The link coming from Wikidata to the japanese article would be shown on the english page with the softredirect, but not on the english article.--Snaevar (talk) 17:12, 9 April 2013 (UTC)
I dont know if perhaps depending on the law the problem does not exist anymore because the name of the criminal does not appear anymore in the sourcecode of the Japanese site?--Saehrimnir (talk) 14:01, 16 April 2013 (UTC)

How can I reference a claim with a newspaper?

How can one reference a claim using a newspaper as a source? It's unclear how to do this from Wikidata:Glossary#Reference, and no fitting properties pop out to me from Wikidata:List_of_properties#Table_of_properties. Thanks, Emw (talk) 00:49, 16 April 2013 (UTC)

See Help:Sources and Wikidata:Property proposal/References 141.6.11.13 08:13, 16 April 2013 (UTC)

Q3193957

What does it mean?--GZWDer (talk) 10:50, 16 April 2013 (UTC)

It seems to be a collection of mixed link to w:WP:Village Pump, w:WP:FAQ, and w:WP:Community portal. Requested deletion after merging. FrigidNinja 11:19, 16 April 2013 (UTC)

duplicate Statements with duplicate reference

please make a limitation or an abuse filter to stop users or bot to making duplicate Statements and duplicate reference like this Yamaha5 (talk) 11:29, 16 April 2013 (UTC)

Abuse filter for invalid IMDb identifiers

Since the Property:P345 (IMDb identifier) has been added to quite a few items, is it possible to create an AbuseFilter for invalid IMDb identifiers? A valid identifier conforms to the following regexp: (tt|nm|ch|co)[0-9]{7} Gabbe (talk) 11:46, 16 April 2013 (UTC)

Sorry, I obviously meant ^(tt|nm|ch|co)[0-9]{7}$ Gabbe (talk) 13:04, 16 April 2013 (UTC)

Wikipedia widged status

Hi Devs, I just would like to ask how the developement of the widged which would replace and hopefully improve as I understand the Wikidat info script is going. Since making this capabillity available to all wikipedians on all wikipedias is an core functionality for the wikidata interwiki feature. Do you need more testers or how can we help.--Saehrimnir (talk) 14:14, 16 April 2013 (UTC)

Gadget-Preview Problem

What is the problem with This. because, i didn't see Preview button in the item? --Aftab1995 (talk) 17:30, 16 April 2013 (UTC)

It's Ok Now O.o --Aftab1995 (talk) 18:24, 16 April 2013 (UTC)

Swapping two items

Recently, the item Q2066131 which had the English label "sportsperson" was renamed to "sportsperson in light athletics" (which I've renamed "athletics competitor"). This was done to reflect the fact that in some languages there are distinct articles for people who take part in athletics (track and field, running, walking) and people who take part in sports in general. The item Q10865754 was created to handle the latter and, accordingly, its English label is "sportsperson". The problem is that Q2066131 was used in the fairly popular script "wikidata useful" and there are thousands of items with the statement "occupation is Q2066131" which originally meant "this person is a sportsperson" but now says "this person is a competitor in athletics" which is incorrect. There are a few ways to fix this but the simplest by far would be to simply swap the two items and manually fix the handful of items that currently link to Q10865754. Is that doable? Pichpich (talk) 12:33, 16 April 2013 (UTC)

  Definitely. --Ricordisamoa 17:00, 16 April 2013 (UTC)
I think this illustrates that it's generally a bad idea to change the meaning of an item.
I'd revert Q2066131 to its original use, create a new one for athletics, and delete Q10865754. --  Docu  at 18:26, 16 April 2013 (UTC)
I'll do it in 24h if there aren't objections. --Ricordisamoa 18:28, 16 April 2013 (UTC)
Support, we never should change the original use. --Stryn (talk) 18:36, 16 April 2013 (UTC)
These were my actions, let's discuss. I agree that status of "what links here" should be corrected. I see 2 ways for that: swapping (quite tedious manual work, regarding inconvenient editing) or bot change of values (simple but many edits). It doesn't matter for me what to choose providing the correctness of actions. In the case of "human gender" change from Q43445 to Q6581072 was decided to do with bots though. Other comments. "...it's generally a bad idea to change the meaning of an item" - didn't know that item (Q+number) has a genuine meaning. "revert Q2066131 to its original use" is nasty thing in 2 ways: 1) see previous sentence about "original use" - in original set there were articles about different things; 2) don't revert please to original chaos (interwiki conflict). "create a new one for athletics, and delete Q10865754" - I don't understand this proposal. --Infovarius (talk) 20:26, 16 April 2013 (UTC)

If I understand Docu's plan correctly, the net effect would be similar to a swap of the two items. So I support that and I'm willing to help if help is needed. (I sure hope help is not needed) Pichpich (talk) 04:15, 17 April 2013 (UTC)

Another point: are you sure that a bunch of items linked to en:Athlete (sportsman), not fr:Athlète for example (light athletics)? Infovarius (talk) 07:26, 17 April 2013 (UTC)
I'm not sure I understand the question. The vast majority of links to Q2066131 were added through "wikidata useful". Of course, some of these are sportspeople who happen to be practitioners of athletics but the swap will only mean that they get a slightly less specific occupation. Obviously if you want to go through the thousands of items to check for this, nobody's stopping you. Pichpich (talk) 12:48, 17 April 2013 (UTC)

I'm new

I'm rather new on Wikidata. I noticed several problems with misdirected items. For instance is: česky (cs) Mědiryt relared to English 'engraving' but is in fact what in German is called 'Kupferstich'. It is not easy to correct such missers. I've corected several links in Dutch: 'effectieve waarde' and the related term 'kwadratisch gemiddelde'. BTW: why is the Q-number not indicated at the pages? Nijdam (talk) 22:09, 16 April 2013 (UTC)

You're probably doing everything right. Thanks for helping. The Q-number is only in the address field of your browser. It would probably be helpful if it would be displayed somewhere on the actual page. --Tobias1984 (talk) 19:40, 17 April 2013 (UTC)

Database migration and update today

Hey :)

Just a quick reminder that the database will be migrated to MariaDB later today and Wikidata will be read-only for a short time. More details here. We'll also upgrade the software here to have qualifiers and fix bugs (for example for Internet Explorer 8). --Lydia Pintscher (WMDE) (talk) 10:39, 17 April 2013 (UTC)

Update: We encountered some problems with the deployment and fixed them. However it's best not to deploy shortly before everyone on the team is going to sleep so we'll do this again tomorrow. --Lydia Pintscher (WMDE) (talk) 23:34, 17 April 2013 (UTC)

Census data and software versions

Are you working on how we will store this data? I think the qualifiers can be a solution but it would be interesting to have the whole series. For example: population = (2013 = 125, 2011 = 236, 2010 = 369) and by default display only the most recent date and the rest are hidden to draw graphs or whatever. --Kizar (talk) 18:23, 17 April 2013 (UTC)

We are waiting for numeric datatype for property. Snipre (talk) 20:36, 17 April 2013 (UTC)

New property proposal: national classification code for cities

Hi. Is there some kind of property for the national codes given to cities/villages? I think the Eurostat term is LAU1_NAT_CODE/LAU2_NAT_CODE [16]. If not, where can I propose this new property? Thanks.--Strainu (talk) 19:55, 16 April 2013 (UTC)

See Wikidata:List of properties and Wikidata:Property proposal Snipre (talk) 12:56, 18 April 2013 (UTC)

Capitalization of labels

Hello, for I while now I've had an idea for fixing the English (and is variations (and maybe others?)) labels for items, based on Help:Label#Capitalization. My idea for the longest while has been to only change if the sitelink title is the same as the label and a page with that title exists in Wiktionary starting with a lowercase letter, but none starting with an uppercase letter (for example, if the sitelink for English is. I'm not so sure if that would be the best idea, but I've no others, and was hoping for comments about this (better suggestions, supports of this implementation, or opposition of the task being done by a bot).  Hazard-SJ  ✈  00:59, 18 April 2013 (UTC)

The idea is fine, but many common nouns are also used as proper nouns for titles of books, songs, films etc. May be the item type (Property:P107 and possibly others) can be used. Byrial (talk) 03:10, 18 April 2013 (UTC)
In some cases you can also use Property:P107 and others as hints to what is correct written form, but in general I think the only reliable way to do this is to analyze how the link text is written in in other articles. That is the article pointed to by the sitelink, and how other articles on Wikipedia points to that article. Because not all links has the correct form you must do some statistical analysis to figure out what is the most likely correct form. You must also try to figure out if some grammatical rule might override the correct wtitten form itself, for example if it is placed first in the sentence and therefore is written by a majuscule even if it should be written with a minuscule. Jeblad (talk) 13:40, 18 April 2013 (UTC)

New RFC about references and sources in Wikidata

Following the discussions in the Books task force and the Help:Sources, I have started a new RFC about sources and references in Wikidata to summarize the different options and to gather feedback from the community.--Micru (talk) 22:40, 16 April 2013 (UTC)

I think it is important to define properly some terms related to references, sources, citations, etc. the discussion is here.--Micru (talk) 00:41, 19 April 2013 (UTC)

Idyll - Idílio

There's a problem with the pages of wikipedia:en:Idyll (Q742530) and wikipedia:pt:Idílio (Q10299833). Actually, the segund one is the portuguese version of the first, it should be added as an interwiki to Idyll (Q742530), but I can't do that, since they created a new page for Idílio, as if it were something different. How could this be solved? Outis (talk) 11:39, 18 April 2013 (UTC)

Solved. I moved pt-link from Q10299833 to Q742530. --Stryn (talk) 11:46, 18 April 2013 (UTC)
You have two choices. 1) delete link manually, and then add it to another. 2) use "move" tool, which can be found at Special:Preferences#mw-prefsection-gadgets. It automatically moves the link where you want. --Stryn (talk) 11:48, 18 April 2013 (UTC)
Thanks a lot! As you can see, I'm kind of new around here... Outis (talk) 11:52, 18 April 2013 (UTC)
It would be funnier, if there were truly en:Idílio :) En-wikipedians like to multiplicate articles :) --Infovarius (talk) 19:30, 18 April 2013 (UTC)

Oversight RFC

A new request for comment about the roles and rules of local oversighers has been started. Regards. — ΛΧΣ21 19:54, 18 April 2013 (UTC)

Can't simple administrators delete revisions, can they? --Ricordisamoa 22:49, 18 April 2013 (UTC)
They can, but oversight hides them so that not even admins can see them. --Rschen7754 public (talk) 22:50, 18 April 2013 (UTC)
Wow... ! --Ricordisamoa 22:54, 18 April 2013 (UTC)
Yeah, it's an additional feature of RevDel called suppression, only available to users with the 'oversight' permission. — ΛΧΣ21 23:07, 18 April 2013 (UTC)
Oversight actually precedes revision deletion by several years and is not part of revision deletion or even standard deletion at all. The revision under oversight completely nukes content from the server... --Izno (talk) 02:09, 19 April 2013 (UTC)
Not quite. That's true of the extension originally used, but now it's just a part of normal revision deletion. Oversighters see an additional checkbox to hide edits from admins in revision deletion forms, and these can be reversed.--Jasper Deng (talk) 02:12, 19 April 2013 (UTC)
To clarify, we're talking about suppression here, the new form of oversight. We're not talking about Extension:Oversight which has been turned off on at least some wikis. w:en:Wikipedia:Oversight explains this a bit better. --Rschen7754 02:23, 19 April 2013 (UTC)
Exactly. @Izno: In 2009, the old oversight extension was discontinued in favour of the enhanced suppression tool included in the RevisionDeletion development feature. The advantages were that, unlike the old oversight tool, suppression can be reversed, all actions are logged in the new suppression log, and it's cleaner. — ΛΧΣ21 02:54, 19 April 2013 (UTC)

Add a population database by country to wikidata

Hi there all, Dr. Blofeld from wikipedia here. As you may or may not know, I've done a lot of work on towns and villages on wikipedia. World Gazetteer has been a tremendous source for finding population census data by country. Looking yesterday I sadly see http://world-gazetteer.com/ it is to close in July. It is way too valuable a resource to shut down. I was wondering if you'd be interested in some sort of agreement which transfers all of the site's data to wikidata and we continue to build the most comprehensive population data by country resource on wikidata. I can help out with at least finding sites which list further data (Senegal for example has data on every village), but I would need a technician to help compile and organize it. I feel it is very important to process this data and have it within our project for the various wikipedias to use. I hope to get something done with this as there is limited time to arrange something if it shuts down in July and would hate to see such a tremendous resource wiped out when I think we could copy it and continue to build it into the best gazetteer on the web and eventually strive to have population data for every town and village on the planet. I feel that this is what such a project is made for. Eventually the data could extend to all demographic data I guess. I'm sure I read somewhere that wikidata has an intention to do something like this but rather than starting from scratch I think if we could form some sort of agreement with world gazetteer before it shuts down we'd build something much quicker and not waste obvious hours upon hours of research. (Dr. Blofeld)

I would certainly support this activity. If we can not yet move data to the infoboxes (due to the lack of the corresponding property type), we can just incubate it somewhere and use later.--Ymblanter (talk) 10:31, 14 April 2013 (UTC)

The data is available for download here in this zip. The delimited "downloadable file contains a list of all towns, administrative divisions and agglomerations with their current population, their English name (if not equal to the international name) and parent country. The international names are displayed in UTF-8 format, which means that special characters which are not displayable with a normal text editor, are displayed as two characters." so everything should be there. The copyright states:

This project is to be regarded as a free data provider. Some requests let me precise the copyright information. If you use its data or images the only thing I ask you is to promote this site. If you would like to republish the data presented here, please do not change the data and use a copyright note as described as follows: © by Stefan Helders www.world-gazetteer.com

You should get permission to use the data in its entirety here as it may be changed. I haven't really looked into how to structure it, but an automated upload of the data could certainly be done (I've done a number of batch uploads on commons and may do it). Cheers.Smallman12q (talk) 12:40, 14 April 2013 (UTC)

Thanks. I can't see anywhere where you can contact him, can you? (Dr. Blofeld)
It is under Feedback at the very bottom. So if you can write him that would be great because you are familiar with that matter already.--Saehrimnir (talk) 16:11, 14 April 2013 (UTC)
I've sent him an email.(Dr. Blofeld)
Against import without the sources. From what I see sources are not provided. I am perhaps wrong but don't add data just will be deleted in the future: we have a stong feedback from wikipedias about sources and it becomes clear that nobody will use data without sources. Snipre (talk) 18:30, 14 April 2013 (UTC)

@Snipre, from their FAQ it says

What are the data sources of the World Gazetteer?
If possible, official data sources are used. In many cases however no official figures are available. In that case, secondary sources such as year books, encyclopediae, atlases etc. are used. I have also received data from other stats lovers.

Hope that helps. 64.40.54.118 02:03, 15 April 2013 (UTC)

The development team had some discussions with the different wikipedia communities and the results are that wikipedias want to have accesse to reference data. So a simple mention that your data are coming from different sources is not enough: you should provide for each number the reference of the book or of the report or of the website from which you take the value (title, author, year, organization or publisher,...). The quality of the references is not important right now. If you can not provide those references better avoid to spend some time to transfer the data. Snipre (talk) 07:05, 15 April 2013 (UTC)

I think you've misunderstood the proposal Snipre. We import the data and look to see what is stated as the source for the material. It would need verifying of course, something hopefully Helders could help us with. Then we would continue to build the resource with sources featuring the data.Dr. Blofeld (talk) 08:21, 15 April 2013 (UTC)

Look first if you have the references before any import, that just my advice and do the importation of data and references at the same time. Snipre (talk) 12:42, 15 April 2013 (UTC)
  • The idea is superb, but we need to be a little careful on the sourcing. As best I can fathom, the world-gazetteer data is often based on census data with a percentage increase per annum based on some growth rate. I think only the raw (census or NGO or UN sourced) data is usable. I don't think it appropriate to assume imputed growth rates that are not based on a reliable source. Since there is no transparency on world-gazetteer's bases for its growth rates, I feel it necessary to not use them per WP:RS, which policy I'll assume applies to Wikidata being imported to Wikipedia. Carlossuarez46 (talk) 17:30, 15 April 2013 (UTC)
  • I agree with the above. (Ser Amantio di Nicolao from Wikipedia here.) The idea needs to be carefully worked out, but in and of itself is just the kind of thing we need to be doing. Imagine...a repository of all the world's census information. Perhaps from local census bureaus - WorldGazetteer is a great starting point, but we should look at migrating data from other international sources as well. It exists, more and more. --216.81.81.80 18:44, 15 April 2013 (UTC)
While World Gazetteer may not be a great source it is nevertheless a source and can be used on Wikidata. For many locations we will have other population figures from more reliable sources, such as census data, and these will also be in wikidata and will probably be preferred to the Gazetteer data but both (or more) can be listed. From some locations, however, the Gazetteer data may be the only figure we have.
Before a mass importation I suggest we should create an item for the World Gazetteer and on the attached discussion page we can have a discussion of the reliability of their data. This can then be linked to via a comment in the annotations (do we have a comment property yet?). Filceolaire (talk) 07:02, 18 April 2013 (UTC)

Is there a tech guy reading this who can get the ball rolling with something in regard to starting compiling the data? I can help out manually if necessary.Dr. Blofeld (talk) 21:47, 19 April 2013 (UTC)

Template:Protected on translatable pages

There is a problem with using Template:Protected on translatable pages. The translated pages are not protected (but cannot be edited directly by anyone) and are placed in Category:Pages with incorrect protection templates. The template should be fixed to avoid that. It is beyond my template knowledge to do that. Regards, Byrial (talk) 19:30, 16 April 2013 (UTC)

There are two ways to do that.

  1. Build an whitelist by using an switch. The whitelist would then need to be consistantly updated.
  2. Request the content of the whole page with Lua and then find the information you need.

So, which one is it going to be?--Snaevar (talk) 18:13, 18 April 2013 (UTC)

I thought of a third way: Add a parameter "pagename" to the template. If the parameter is used and does not match the actual pagename, the template should do nothing. Byrial (talk) 14:38, 19 April 2013 (UTC)
I found out how to that myself, and learned something about template syntax in process. It seems to work. Byrial (talk) 23:16, 19 April 2013 (UTC)

Deletion of notable items without notifying the creator

Can someone explain to me why the items Q10387575 and Q10387684 have been deleted? Yes, they didn't contained one valid sitelink to a Wikipedia page but they met the two others criteria of notability :

  • they refered to a clearly identifiable conceptual : the subclasses of national heritage sites of France
  • they fulfilled some structural need : to class them more precisely.

Moreover, I don't understand why I haven't been notified of this deletion while I was the creator of this items...

So I request their undeletion and ask the admins to respect the rules of notability for empty items or at least to discuss their deletion with the creator before doing it. Ayack (talk) 09:43, 19 April 2013 (UTC)

We should indeed improve by (1) in the situations where notability is questioned (and only in these situations) notify the creator; (2) if the item is notable, find some way of showing it on the item (qualifiers?)--Ymblanter (talk) 11:33, 19 April 2013 (UTC)
If you look at these items (I think you can as an admin), you should see that they had qualifiers. Ayack (talk) 11:41, 19 April 2013 (UTC)
I only see statements, and qualifiers were only introduced last night, so they just could not have qualifiers. But this is not my second point. My point is that we should have a way to mark the item which was already checked for notability as notable. Currently we might have means to do it, but I do not think the community ever discussed these means.--Ymblanter (talk) 11:44, 19 April 2013 (UTC)
Right. I got them confused. At which means are you thinking? Talk page? A dedicated property? By the way, since Wikidata:Requests for undeletions doesn't exist, could you undelete this two items please? Thanks. Ayack (talk) 12:42, 19 April 2013 (UTC)
I think Wikidata:Notability needs to be worded better :/ Although I'm not really sure how to do this.. ·Add§hore· Talk To Me! 12:54, 19 April 2013 (UTC)
Wikidata:Notability may need to be made more precise, but I am not sure this is yet possible to do so in a sensible way (see #Notability of items aboce). One of the problem may be that habits were taken to delete items with no sitelinks when it was the guideline to do so. Now that needs have changed, behaviors and tools probably need to be updated as well. I think apart duplicate items and obvious vandalism, we should always make a deletion request before deleting. --Zolo (talk) 13:04, 19 April 2013 (UTC)
The problem is these items were deleted as the RfD result.--Ymblanter (talk) 13:58, 19 April 2013 (UTC)
Yes, I have seen this afterwards :). Actually these RfD are essentially people notifying empty items, and admins chain-deleting them within an hour. I guess that we mostly need to change our habits, but smarter tools may be possible. If item was created by an autopatrolled user and never contained any sitelink, it is much more likely to serve a purpose than a bot-created item that was empties by an autopatrolled-user. --Zolo (talk) 14:39, 19 April 2013 (UTC)
Thanks for the undeletions Zolo! Ayack (talk) 14:55, 19 April 2013 (UTC)
If we all agree, we should write it somewhere.--Ymblanter (talk) 20:40, 19 April 2013 (UTC)

Always redirecting to www now

bugzilla:45005 has finally been closed. This means all traffic should now be redirected to www.wikidata.org and for example wikidata.org and fi.wikidata.org and others are no longer used. A lot of people had issues editing on the other URLs, getting very useless error messages and so on. This should now be solved by this change. --Lydia Pintscher (WMDE) (talk) 12:18, 19 April 2013 (UTC)

Thanks again! --Ricordisamoa 12:21, 19 April 2013 (UTC)
Would it be possible to redirect the user to the language version that issued the redirect? For instance de.wikidata.org redirecting to Wikidata:Hauptseite and not to the English version.--Micru (talk) 15:06, 19 April 2013 (UTC)
Yes this is on the todo list but we had to make sure this is fixed first. --Lydia Pintscher (WMDE) (talk) 18:57, 19 April 2013 (UTC)

Validation of property values

I came across User:Tpt/validator.js... could we have a template to be put on the properties' talkpages, to define some rules in-place? Creating an AbuseFilter for each of them seems [...] to me. --Ricordisamoa 15:45, 19 April 2013 (UTC)

Office hour about sources

Heya :)

Denny offered to do an office hour to help with answering remaining questions you might have about sources for the references and sources RfC. I've set this up. Details are here. I hope you can make it. --Lydia Pintscher (WMDE) (talk) 21:28, 19 April 2013 (UTC)

SamoaBot is ready

I've just rewritten my code as Python, so it runs in PWB now. It's ready to set taxon name, IUCN conservation status, and labels in the desired languages. It also changes extinction with extinct species in the IUCN conservation status.

However, the current RFBOT didn't get commented... Should I consider that as an approval?!? ;-)

PS: you'll get the code soon. --Ricordisamoa 23:10, 19 April 2013 (UTC)

An RFC

I have started an RFC on Wikidata:Vandalism. The RFC can be found at WD:Requests for comment/Wikidata:Vandalism. Comment away! ...Jakob Megaphone, Telescope 23:37, 19 April 2013 (UTC)

Election, Sport and other game results in Wikidata => Timetable?

Hello,

I am eager to help Wikidata build a monumental database of all political elections throughout the world for all kinds of offices, at any level of importance. Political science needs those and the progress to be made from such a database is enormous.

So my question is simple. I have seen on the Queries page that sports results may soon be pooled in order to be up-to-date, but the page seems all but a mere suggestion. I suppose election results and everything that looks like games can be added, since the principle of any game is that all participants accept the rules. There mustn't be much exceptions to discuss.

My question is simple : when will this come? Don't you think this data should be added in the Infoboxes project despite most election tables are written in simple wiki form?

P.S. : I also wonder what new safety will be invented to ensure wikdata integration does not become an Achille's heel. The users' common defence of vandals may not work as well if all barbarians gather on a unique, centralized point in which people can't understand each other.

Thank you all,

Kahlores (talk) 21:25, 16 April 2013 (UTC)

I'm interested in this topic, especially regarding sports results, as well. It would be extremely useful to integrate this into wikidata. As I understand it, it essentially should be possible to enter results as usual data objects at this point already. However their structure must be defined precisely before it makes sense to enter any data. Is there any development going on, any discussion somewhere, anything? --Kompakt (talk) 09:58, 19 April 2013 (UTC)
As for the structure I cannot help, because I haven't added a single thing on Wikidata. However, the data must come from the official source, much like, for instance, the last stable version number, the accepted taxon, etc. So, for instance, go to FIFA.com in oder to know the result of the 1934 World cup final. However... there might be more issues for things such as attendances. Not only are the reports rounding numbers, i.e. 80000 instead of 79840, but they often lie or release the number of tickets sold. So I would not create an "attendances" category, but attendances according to the same source X. The general aim should not be forgotten. Wikidata will be a common base for past and present game data, in order to spare ourselves hours of dirty work each in our own Wikipedia. Kahlores (talk) 17:33, 20 April 2013 (UTC)

Qualifiers

 
Explanation of Wikidata terminology.

Should we create a new page for qualifier proposal or the existing pages for property proposals are sufficient ? Snipre (talk) 22:23, 18 April 2013 (UTC)

Either a subpage, or a parameter in {{Property documentation}}, should work well. --Ricordisamoa 22:47, 18 April 2013 (UTC)
Some regular properties may work as qualifiers as well (and vice versa). I don't think there is a strict distinction. -- Duesentrieb (talk) 07:55, 19 April 2013 (UTC)
  1. I add some existent properties as qualifiers (see for example Moon, Earth and Orvieto). Imho qualifiers are very useful to explain better the "hierarchic" properties as is in the administrative unit, parent body, or children body.
  2. I have also a question: when I add a qualifier, in the "Revision history" of the page appears my revision, but without any specifications. Is it possible to add something as "added a qualifier for a claim" as done for sources? --Paperoastro (talk) 08:44, 19 April 2013 (UTC)
Maybe I'm wrong, but in my opinion the qualifier in Moon is totally unnecessary and not the purpose of a qualifier. It just give an information about earth. It gives no information for the statement. The hierarchy can be queried. For me, qualifier should refine a statement as whole. If someone has multiple occupations, you could use qualifier to outline which occupation was in which time. If someone other agree with me, please remove the 'type of astronomical object' qualifier from the moon statement. --Nightwish62 (talk) 08:57, 19 April 2013 (UTC)
I have to agree with Nightwish62 here. In this case, the qualifiers seem completely redundant with information provided in the items given as values. I don't think this is how qualifiers are supposed to be used. See the image I've added. Gabbe (talk) 09:09, 19 April 2013 (UTC)
out of cron. I removed the qualifier. I'm sorry, but I thought it was a good idea specify the type of the "parent object". --Paperoastro (talk) 09:17, 19 April 2013 (UTC)
The type of the parent object is already speficied in the item about it, and it will be possible to retrieve it for clients. However, I think that the parent body should qualifier for "natural sattelite" (and would be structurally equivalent of "sattelite of: Earth", used in the infobox of en:Moon). --Zolo (talk) 11:08, 19 April 2013 (UTC)
Uh, yeah. Even more bad usage of qualifiers (still in my opinion) here (distribution):
  • System Shock 2 <distribution> digital distribution <platform> GOG.com - fine, it have an information to the relation concerning the statement itself
  • System Shock 2 <distribution> CD-ROM <subclass of> Compact disc - not fine, it have just information about CD-ROM, not the statement
However, I think for 007 Legends the qualifiers are used in a perfect way. --Nightwish62 (talk) 09:13, 19 April 2013 (UTC)
@Nightwish: I don't think that item is a good example of qualifiers, since the property of platform is not used to describe physical distributions (like CD-ROM), but the platform on which the piece of software has been released (such as Xbox, Microsoft Windows, OSX, etc.). A good example of qualifier use in my opinion, would be having ticker symbol as qualifier, under the property of stock exchange. --Wylve (talk) 11:54, 19 April 2013 (UTC)
You're right, I'll change that. --Nightwish62 (talk) 08:01, 20 April 2013 (UTC)
I think managing all the qualifier in one central page may be difficult given the list of properties. How about posting the proposal in a central page and then placing the approved qualifier in the talk page of each property page? --Napoleon.tan (talk) 09:59, 19 April 2013 (UTC)
On the other hand, some qualifiers (I suppose most of them) will be useful for several (or many) properties. Maybe a list of approved/recommended qualifiers for a particular property on a property talk page would be more useful. Linked to the talk pages of the qualifiers with their full description.--Šlomo (talk) 11:57, 19 April 2013 (UTC)

Meanwhile, I posted a qualifier proposal on Wikidata:Property_proposal/Generic#Qualifiers. Is there some more suitable place for it?--Šlomo (talk) 11:57, 19 April 2013 (UTC)


According to me Qualifiers have 2 main uses: 1) for historical values like Prime minister of a country Person A from 2000-2004, Person B from 2005-2010 and Person C currently (without qualifier?). and 2) for POV statements like Taiwan is or is not a province of China. If a wikpedia wants to show the generally accepted POV it shows the statement without qualifier, or if not: use the statement with the POV qualifier. HenkvD (talk) 18:04, 19 April 2013 (UTC)

List of shortcuts

Hi! I took a look at the created redirect pages, and compiled a list of some of the shortcuts: Help:Shortcut. Byrial (talk) 07:47, 20 April 2013 (UTC)

move-log bot source code

Hi, I wrote a code which can update items which their language link in locale wiki are moved. I know it is not new and some bots now they do it! but in some wikis delaying for updating moved links made local users confused like #Moved pages on WP, how many bots are watching them? .the code is here and tested -- Yamaha5 (talk) 14:45, 20 April 2013 (UTC)

Local oversighters

Given the increasing amount of en:WP:BLP violations that often need suppression, I believe it is approaching the time when we need local users to fulfill this role. See here for the technical details about this permission.

Stats

18 Apr. 2013 — 8x (User edited while logged-out, revealing IP)
18 Apr. 2013 — 5x (Potentially defamatory content)
16 Apr. 2013 — 3x (User edited while logged-out, revealing IP)
10 Apr. 2013 — 4x (Inappropriate comment or personal information)
10 Apr. 2013 — 3x (User edited while logged-out, revealing IP)
25 March 2013 — 1x (User edited while logged-out, revealing IP)
24 March 2013 — 1x (User edited while logged-out, revealing IP)
22 March 2013 — 2x (User edited while logged-out, revealing IP)
6 March 2013 — 1x (User edited while logged-out, revealing IP)
4 March 2013 — 2x (Personal information)
3 Nov. 2012 — 1x (Personal info)

For comparison, we have had 23 suppressions already this month. Looking at commons:Commons:Oversighters/Statistics that partial month is already about on par with many full months at Commons (ignoring the 100+ suppression months; it seems that their range of activity varies widely). It's clear that the number is only going to increase from here on out. --Rschen7754 02:49, 18 April 2013 (UTC)
  •   Support needed function. Regards, — Moe Epsilon 02:53, 18 April 2013 (UTC)
  •   Support – The need seems documented -- Byrial (talk) 02:55, 18 April 2013 (UTC)
  •   Stöder -- Lavallen (block) 05:13, 18 April 2013 (UTC)
  •   Weak oppose I don't really see the need. We've only had to request it a handful of times, primarily for IP-revelation issues. Contrast this to bureaucrats, who we didn't appoint until we'd had scores of requests for +sysops and +bots. Furthermore, as one of quite a few administrators under the age of 18, I don't love the idea of installing an age-restricted user right unless it's really necessary. — PinkAmpers&(Je vous invite à me parler) 07:09, 18 April 2013 (UTC)
    • There's sufficient requests to have a local team, and they will definitely be needed in the future - for something like this, a bit too early is better than a bit too late. Also, this is giving a right that the stewards have (which is also age-restricted) to another group that is also age restricted; it's not like those under the age of 18 are losing anything here. --Rschen7754 07:23, 18 April 2013 (UTC)
      • People under the age of 18 can always run to be a bureaucrat. That group is not age-restricted. Techman224Talk 17:54, 21 April 2013 (UTC)
        • ? … What has running for bureaucratship to do with running for oversight? These are 2 completely different tasks. The intention of this age restriction is surely not to make people under 18 think "oh, I'm not 18 yet … well, I'm running for bureaucratship instead … the main thing is that I get a new hat". Regards, Vogone talk 18:13, 21 April 2013 (UTC)
  •   Support might be needed often in future. --Stryn (talk) 07:20, 18 April 2013 (UTC)
  •   Support Stewards are second best to a local team for suppression. It's not a huge problem for English and other widely available languages but, unlike CheckUser, for rarer languages a Steward often has to rely on the opinion of other users on whether the content should be suppressed or just revdel'd (or neither). Much better to have trusted local users who understand the language, and the content, to make these decisions QuiteUnusual (talk) 08:29, 18 April 2013 (UTC)
    • Well, but since this is a "multilingual" wiki (with much stuff in English), that seems rather like an argument for stewards of which there are 40 who know various languages, that a single-digit number of OSes, of which the language knowledge is still unknown ;-) --MF-W 19:53, 18 April 2013 (UTC)
      • Familiarity with local policies and content is still a major argument in favor of local users.--Jasper Deng (talk) 21:41, 18 April 2013 (UTC)
        • And I didn't realise at the time that the proposed policy was to have "2 or 3 oversighters"! I had assumed something like 5% of the admins, so 3 / 4 now, but increasing as the community grew. I would have thought this would end up as the most multi-lingual project in the end with those with the very rare language skills from the small Wikis also contributing here. QuiteUnusual (talk) 21:56, 18 April 2013 (UTC)
  • Support, with the intention that the current RfC passes meaning that stewards can still act as local oversighters when doing crosswiki work. Ajraddatz (Talk) 12:27, 19 April 2013 (UTC)

Proposal to make Help:Description an official guideline (just the English version)

I proposed this a few weeks ago, but the discussion was derailed by discussions about what should be translated and what should be language specific. My proposal is to make only Help:Description an official guideline, not any of the other language versions, such as Help:Description/de. The other language versions should be discussed separately since each language has its own conventions for writing descriptions. There has only been one substantive change to the English page in the past month and the page seems to accurately reflect current best practices. We should either make it a guideline or remove the {{proposed}} tag that's been sitting on it since November. Kaldari (talk) 20:27, 18 April 2013 (UTC)

  • I propose that the first part before the headline "Guidelines" + the section named "Statements to avoid" (which should be moved up) become official for all languages. The rest of the guidelines become official for the descriptions in English only. Speakers of other languages can define guidelines for their language, but until that happen the guidelines for English can be used as a starting point to be extent that they make sense and are useful for the individuel language. Byrial (talk) 21:20, 18 April 2013 (UTC)
  Support, certainly fine with me.--Ymblanter (talk) 11:29, 19 April 2013 (UTC)

No one opposed so far, so I wrote my changed proposal into the text. Please check if the phrasing is OK. Byrial (talk) 06:44, 21 April 2013 (UTC)

Problem with Kepler-69

I looked for a wikidata entry for en:Kepler-69, and did not find one. But if I try to create a new entry, I get the error message "Could not create a new page. It already exists. ". What is wrong? And if that entry exists, where can I find it? Unfortunately, the error message does not link to the existing entry. --Mfb (talk) 12:17, 19 April 2013 (UTC)

When you open the create Item page a item ID is assingned to the entry but the Item is only created when you hit create if a bot created a item in the same moment the ID is already taken. In short most of the time you just have to hit create again.--Saehrimnir (talk) 12:32, 19 April 2013 (UTC)
It's a known problem: see this discussion. I am still searching for the bugzilla ID... --Ricordisamoa 12:33, 19 April 2013 (UTC)
  created --Ricordisamoa 12:35, 19 April 2013 (UTC)

Thanks. --mfb (talk) 13:55, 19 April 2013 (UTC)

If you mean the same error which I know (with Special:NewItem, wasn't there a second page where a sitelink could be entered directly?}: You bypass it by repeat pressing the button, it should work after 2~3 clicks. --#Reaper (talk) 15:36, 21 April 2013 (UTC)

Wrong interwikis

The correct interwiki for Q10336806 is Q3341217. How do I solve this? Jbribeiro1 (talk) 11:58, 21 April 2013 (UTC)

Hi, you can use the gadget move which can be enabled in your preferences. After moving the sitelink, you can request the deletion of this item at WD:RFD. Regards, -- Bene* talk 12:01, 21 April 2013 (UTC)

Qualifiers, bug fixes, improved search - all in one night!

Heya folks :)

We have just deployed qualifiers and bug fixes. Qualifiers! Bug fixes are especially for Internet Explorer 8. Please let me know how it is working for you now if you're using IE8 and if there are still any major problems when using Wikidata with it.

In addition the script we ran on the database to make search case-insensitive has finished running. This should be another huge step towards a nice search here. (This change also affects the autocomplete for items and properties.)

As usual please let me know what you think and tell me if there are any issues.

Cheers --Lydia Pintscher (WMDE) (talk) 21:39, 18 April 2013 (UTC)

Greeeeeeeeeeaaaaaaaaaat! I'll test qualifiers soon... --Ricordisamoa 22:48, 18 April 2013 (UTC)
Yeah! Thanks! Case-sensitive properties was – for me – way more disturbing than search & all other bugs. And Qualifier will be great. -- MichaelSchoenitzer (talk) 23:14, 18 April 2013 (UTC)

When clicking notice: "By clicking "save", you agree to the terms of use, and you irrevocably agree to release your contribution under the CC0 license. I accept these terms for my future edits. Do not show this message again.", it does not go away. Nothing happens when I click it. So it pop every time when I'm going to edit something. --Stryn (talk) 08:57, 19 April 2013 (UTC)

Confirm this behavior. Have the same issue yesterday at home (IE10). --Nightwish62 (talk) 08:59, 19 April 2013 (UTC)
Same for me with Firefox 20. --FelGru (talk) 22:09, 19 April 2013 (UTC)
Not cool -.- I've filed it as bugzilla:47435. --Lydia Pintscher (WMDE) (talk) 22:18, 19 April 2013 (UTC)
Can I have old good case-sensitive search please? There are so many work introduced in separation of proper and common nouns that it's dumb to lose this information. Infovarius (talk) 04:03, 22 April 2013 (UTC)

Previous versions

Since recent time I can't preview any previous version - neither properties nor links. Is it known bug? Infovarius (talk) 19:18, 19 April 2013 (UTC)

What does it say?--Ymblanter (talk) 20:40, 19 April 2013 (UTC)
Nothing. Just showed actual state as if it is previous. Infovarius (talk) 03:43, 22 April 2013 (UTC)
Same here too. I always only see the actual links and properties, only the differences are correct - more or less, sometimes not correct working. I think that someone has already reported this bug, I hope. --#Reaper (talk) 15:32, 21 April 2013 (UTC)
Actually all right now. Infovarius (talk) 04:37, 22 April 2013 (UTC)

Values of properties

Hi, folks! We have over 400 properties, so it's hard to keep track of which values are valid for a specific property.

The description of P21 (sex) says that there are four different valid values, and mentions the item number of each. But what about P106 (occupation)? There is an infinite number of possible options, and it's hard to find the right one.

So, my proposal is to create a page where we list the possible values of each property, and link it from the relevant places (most importantly the property page itself). What do you think about that? Thanks! --NaBUru38 (talk) 19:36, 18 April 2013 (UTC)

Strong   Oppose for a additional site. But strong   Support we should give such information you mentioned. But this has to happen as near as possible to the property page, meaning ON the property page itself. See here. I don't understand why this would be that complicate, that Lydia said it wouldn't be realized in the next time. In my opinion this is an important part to promote properties are used the right way. It couldn't be that complicate to make the bottom of the property page editable. Meanwhile the proposal is archived, don't know if or when something will happen. --Nightwish62 (talk) 20:46, 18 April 2013 (UTC)
There is the Discussion page for each property. We could put this info there. We can move it to the Property page later when this is enabled. In the meantime maybe the devs could add a comment to the property page pointing to the discussion page for this info. Filceolaire (talk) 06:47, 19 April 2013 (UTC)
Keep the information about the use of property at the top of the talk page: we can think to use a frame in order to present the rules in the best way. Can some one propose something ? Snipre (talk) 07:24, 19 April 2013 (UTC)

I agree with using discussion pages of properties. I'll start with the basic properties, ok? --NaBUru38 (talk) 16:11, 22 April 2013 (UTC)

I just found this: Help:Basic membership properties. Seems a great piece! --NaBUru38 (talk) 16:17, 22 April 2013 (UTC)

New deployment dates for phase 2 on the remaining Wikipedias

Hey :)

Just a quick heads-up: The new plan is to deploy phase 2 on English Wikipedia on 22nd and the rest on 24th if there are no issues. --Lydia Pintscher (WMDE) (talk) 13:05, 19 April 2013 (UTC)

Perfect, thanks. --Ricordisamoa 13:28, 19 April 2013 (UTC)
What preparation is needed on local wiki?  Ę-oиė  >>>  ™ 15:08, 19 April 2013 (UTC)
Duck and cover! Actually no, nothings changes, if you don't change anything on the wiki. -- Lavallen (block) 15:42, 19 April 2013 (UTC)
Aye aye Sir! I will duck and cover. :)  Ę-oиė  >>>  ™ 16:03, 19 April 2013 (UTC)
It might be a good idea to talk about where and under what circumstances adding Wikidata data is ok - especially at the beginning. One of the early Wikipedias for example decided to only use it in templates, discuss any such template changes on a dedicated page and not use it for controversial values for now. Specific guidelines are of course up to each wiki and up to each wiki to see if they are needed. --Lydia Pintscher (WMDE) (talk) 18:56, 19 April 2013 (UTC)
Will it be activated at midnight UTC, or tomorrow morning? - Ypnypn (talk) 23:01, 21 April 2013 (UTC)
As usual later evening UTC I'd say. --Lydia Pintscher (WMDE) (talk) 07:30, 22 April 2013 (UTC)

Problem with edit message

When I try to edit an entry, and click the button that says, "I accept these terms for my future edits. Do not show this message again.", it doesn't go away. I can still save, but its getting annoying. Could someone fix this? Techman224Talk 01:50, 22 April 2013 (UTC)

+1. Some times ago it went away, but now it sticked. Infovarius (talk) 03:17, 22 April 2013 (UTC)
Bugzilla:47435. --Stryn (talk) 06:48, 22 April 2013 (UTC)
And now it's back to how it was before, forgetting the "Do not show this message again" every few minutes. Oh well. Better than nothing, I suppose. - Soulkeeper (talk) 18:01, 22 April 2013 (UTC)

problem with Help:Label for machine learning

As you know most of the wikidata's edits are done by bots or automatic JS tools. Help:Label has many problems with bot developing and machine learning

  1. This help is not for all languages so if a bot developer want to generate a label he/she should read all the local help versions!! (if he knows all languages :) )
  2. How we can define for bot cases like Cambridge, Massachusetts which should be Cambridge when we have many pages which has , inside their title!
  3. How we can define for bot cases like Chamerion angustifolium which should be fireweed !!
  4. Gadget MediaWiki:Gadget-slurpInterwiki.js and other gadgets don't do some of these rules!
  5. These rule are defined in 3-4 langues and the English version mentioned These guidelines are proposed for application on the English labels of Wikidata, and thus they can't be taken as a general rule for all languages. what about others? also what about Arabic and east Asia languages which don't have lower case?

why we don't define a general rules for all languages or add some minor exceptions for some languages which can recognize by bot and JS tools to simplifying the users works. doing these rules by normal user and hand for 11,000,000 items is impossible! so this help will be useless! for more discussions see here Yamaha5 (talk) 11:18, 22 April 2013 (UTC)

Yes we should indeed have some general rules for this and then some additional notes about how those rules are to be interpreted in a specific language. A situation with 250-300 different rulesets will simply not work. Jeblad (talk) 16:09, 22 April 2013 (UTC)

Interwiki for numbered lists

There are some lists which have very much members - hundrets or thousands. These lists are usually split in subpages. But in any languages are groups of 20, ian any 50 and in any 100. And there is problem with interwiki and with correct groups in wikidata. I found two lists of this type:

  • List of minor planet (136563 members)
    • 1-100: an, ar, as, be, en ...
    • 1-250 cs, sk, sl
    • 1-499 sr
    • 1-500 hu, sv

and e.g. pokémon 203 is on list:

    • 201-220 cs, ko, pl, fi
    • 202-250 en, da

Different languages have differend splitting and there is question how to correlctly interlink them. Since Wikidata there was usually the longest list linked with all which have same starting number (see firs list of planets) and the shorter were linked only with other shorter (planets cs:251-500 linked with sk and sl only).

Any ideas?

Since the various wikipedias are free to organise themselves as they see fit this problem will never have a perfect solution. The least worst is probably to link together all the first pages, all the second pages etc.
Remember that we will want to make statements about each minor planet and about each Pokemon and this will mean that each will eventually have it's own WikiData page. We will still need to find a way to link to the list pages though. Filceolaire (talk) 12:13, 12 April 2013 (UTC)
Interwiki should link to an index page (as en:List of minerals (complete)), and the index page links to all subpages. --Chris.urs-o (talk) 04:08, 23 April 2013 (UTC)

Moved pages on WP, how many bots are watching them?

I see a lot of irritation on WP that iw-links do not follow automatically when pages are moved. Users are now starting to talk about removing the movepage-right for ordinary users, to prevent damage. How many robots do we have who are watching WP's move-log to update the sitelinks? This is especially critical in pages where the content of a title is replaced with a new subject. Suprisingly many users has complainted that it is impossble for them to edit WD at all. It took me some time to realise that they are not using javascript in their browser. I do not think I can change that behavior, and also I am very frustrated with the javascript-interface. -- Lavallen (block) 09:21, 17 April 2013 (UTC)

We need a bot in Wikipedia correcting Lemma in Wikidata after moving. Conny (talk) 12:24, 17 April 2013 (UTC).

I am currently developing a bot to do this globally ·Add§hore· Talk To Me! 12:30, 17 April 2013 (UTC)
Currently BetaBot and Hazard-Bot. Also bugzilla:36729 is open. --β16 - (talk) 12:45, 17 April 2013 (UTC)

Great guys, this is so wonderful :) . Thank you for your work! For german Wikipedia there is one bot doing updates :) . Conny (talk) 12:56, 17 April 2013 (UTC).

I added a code in #move-log bot source code Yamaha5 (talk) 14:48, 20 April 2013 (UTC)
Please see my request for approval Wikidata:Requests_for_permissions/Bot/Addbot_2#Addbot_2 ·Add§hore· Talk To Me! 14:45, 23 April 2013 (UTC)

Categories

There is more complex problem: how to update interwiki to moved categories? Usually, when it is necessary to rename category "A" to a new name "B" this is done using the following algorythm:

  1. Create new category "B"
  2. Move a content of "A" to "B"
  3. Delete "A"

As result, there is no records in a move log and I don't know any assured way to detect whether the new category "B" has the correspondent item on Wikidata and update it. Yes, I know that sometimes there is used a category redirect template, but it is not the common practice.

Does anyone have any ideas on this? --Emaus (talk) 16:24, 17 April 2013 (UTC)

Add a link to the new category in the editsummary, when deleting the old. -- Lavallen (block) 16:42, 17 April 2013 (UTC)
see this Wikidata:Requests for permissions/Bot/JYBot 2 --Javad|Talk (28 Farvardin 1392) 16:47, 17 April 2013 (UTC)
or:
  1. Go to Wikidata and update the item for the category
Just my suggestion. Compared to the work involved in point 2 that should be no big deal. Byrial (talk) 16:46, 17 April 2013 (UTC)
I don't know about other wikis, but in RU-WP in most cases step 2 is done by bots. --Michgrig (talk) 07:17, 18 April 2013 (UTC)

singles vs. songs and similar problems

On a number of wikis (en.wiki in particular), many articles about individual songs use an infobox that labels them as singles. For instance the article en:The Scientist (song) is not really about a single. (Note that even the disambiguator is "song" not "single") Most of the article is about the inspiration and composition of the song and its reception. There are a few details about the single but even the music video of the song receives more attention than the single. Some bots have started tagging these items as "instance of single". So my question is do we

  1. replace "instance of single" by "instance of song"
  2. add "instance of song" but also keep "instance of song"
  3. split the item in two so that we have an item for the song and one for the single

More generally, is there a policy or at least a common practice regarding articles that fuse information about two distinct but related objects? Pichpich (talk) 16:58, 17 April 2013 (UTC)

If it's not a single, but a song, then we have to add "instance of song". If it's a single, then we should add both, "instance of single" and "instance of song". Strongly disagree to split the item in two just because of this. --Stryn (talk) 18:27, 17 April 2013 (UTC)
Btw, your example article, "The Scientist" is a single. But en-wiki always has "song" as a disambiguation per naming conventions. --Stryn (talk) 18:32, 17 April 2013 (UTC)
The issue is that every single is a song, but not every song is a single. Therefore, I agree with Stryn. If the song is a single, add the value to the 'instance of' property. No need to have two separate items for the same subject. — ΛΧΣ21 18:40, 17 April 2013 (UTC)
Since, as ΛΧΣ said, a single is a kind of song, there is no need to add "instance of: song" as well as "instance of: single" because "instance of: single" implies song. However, if there were a property to indicate certain songs were published as singles or in albums (e.g. "published in"), then I think the item should have "instance of: song", and the special purpose property should be used to specify the song was published as a single. Silver hr (talk) 16:33, 18 April 2013 (UTC)
A single is not a kind of song. A single is a recording that includes one or more versions of a song (and often other songs as well). These are 2 completely different things and should be represented by 2 different entries. Kaldari (talk) 20:35, 18 April 2013 (UTC)
Not exactly. A single is a song that has been commercially released as (i) digital download (or as a physical disc) and/or to (ii) radio stations to promote the album from which it belongs. We also have an additional definition of single: A disc set that includes one or more versions of one or more songs that is released to promote the album from which any or all of those songs belong. Therefore, all songs must have instance of set to song, and all singles must have it additionally set to single, per the first description of what a single is, and because Wikipedia has no individual articles for each of the disc sets that a song can have while being released as a single. — ΛΧΣ21 20:44, 18 April 2013 (UTC)
So, there are two different concepts that have the name "single". One is "a kind of song", another is "a kind of album/release". This means that we need to have two items that describe what "single" is, not one (let's say "single (song)" and "single (album)"). Then, if "single (song)" is a subset of "song", for those songs which are singles we only need to have "instance of: single (song)" because that also implies "instance of: song". Having multiple "instance of" statements along a single hierarchy is redundant and should be avoided. Silver hr (talk) 23:04, 22 April 2013 (UTC)

en:templates

I like en:Template:Chem and en:Template:Cite journal. I'd like to use template:Chem on 'property' 'chemical formula'. Does anybody like these ideas? --Chris.urs-o (talk) 11:47, 20 April 2013 (UTC)

I'm not sure that it is possible to use templates in those items. --Izno (talk) 15:56, 20 April 2013 (UTC)
See Help:Sources and Wikidata:Chemistry task force. Snipre (talk) 20:20, 20 April 2013 (UTC)
Ok, I'll follow, let's see where we go. --Chris.urs-o (talk) 04:22, 21 April 2013 (UTC)
Question: is it possible to save the property 'chemical formula' as 'H|2|S' and add {{Chem| ... }} of the en:Template:Chem in the 'infobox mineral' on en.wikipedia? --Chris.urs-o (talk) 05:12, 22 April 2013 (UTC)
No, nor would we want to, because that's not the chemical formula. en:Template:Chem will probably need to be written, and it might be a good idea to do it as a Lua module. --Izno (talk) 02:59, 23 April 2013 (UTC)
Ok, thx --Chris.urs-o (talk) 04:12, 23 April 2013 (UTC)

language group "de-formal"

It looks like translation activity for de-formal has stopped some time ago after very little work done. Is there a way to suppress "de-formal" translations from the Wikidata interface? (I really wouldn't mind if these pages were deleted, as the language variant "de-formal" looks completely superfluous to me: Even documentation in de:wikipedia makes no use of it. Or am I missing something?) -- Make (talk) 20:12, 20 April 2013 (UTC)

  Support Please delete it, I don't see any use for it. --Pyfisch (talk) 06:03, 21 April 2013 (UTC)
  Support I see no use of it, too. Should be deleted. --#Reaper (talk) 14:39, 21 April 2013 (UTC)
  Comment As far as I know, a fallback for de-formal here on Wikidata is implemented. Thus, de and de-formal are the same here (except some differences in the interface ). Regards, Vogone talk 14:44, 21 April 2013 (UTC)
Why should they be suppressed from the interface? It is an option for interface language in all Wikimedia wikis AFAIK. It harms noone, and supposedly someone uses it. Byrial (talk) 15:03, 21 April 2013 (UTC)
Indeed. There are still people using it and there is a difference between de and de-formal. But theoretically all translations to de-formal could be done by bot if a translation to de exists. Regards, Vogone talk 15:14, 21 April 2013 (UTC)
But the interface has nothing to do with the content of the items. "de-formal" claims that there is language de-formal, but that is definitely wrong. --#Reaper (talk) 15:23, 21 April 2013 (UTC)
There is no de-formal for items anymore. Only for the interface … Vogone talk 15:25, 21 April 2013 (UTC)
Whoops.. sorry, wrong context. But: there is. It's possible (e.g. with labelLister) to create "de-formal" for items. autoEdit uses this sometimes (till now, I've posted this there). --#Reaper (talk) 15:55, 21 April 2013 (UTC)
Yep, you can edit labels for all interface "languages". But you don't need to. You can try to set your interface language to de-formal. In your watchlist, recent changes etc. the labels are still the same as in "de". That's the fallback. Regards, Vogone talk 16:04, 21 April 2013 (UTC)

Obviously this should not be discussed here, but on Wikidata:Forum and in German with the people concerned by this issue. Ljubinka (discussion) 11:21, 22 April 2013 (UTC)

OK, I did some googling on "de-formal" and get the idea. – What would be the best way to deal with Help:Sitelinks/de-formal? (Other than finishing that translation because I have only started to update Help:Sitelinks/de.) – Leaving it as is leads to a somewhat broken user experience. -- Make (talk) 21:13, 23 April 2013 (UTC)

Q11317656

can we create these items?--GZWDer (talk) 04:57, 22 April 2013 (UTC)

I don't think so! --Viscontino (talk) 09:36, 22 April 2013 (UTC)
Why not? All of them are not in the user namespace, and leaving out just those pages does not make sense. --Stryn (talk) 09:48, 22 April 2013 (UTC)
I don't see how this fits into Wikidata's scope, especially since the existence of this page in multiple languages is artificial. (why do we need an fr.wiki page on the most active accounts on en.wiki?) It's a userpage and it should be deleted. Pichpich (talk) 14:18, 22 April 2013 (UTC)
I've deleted the item because it does not meet the notability policy. If you want to get it undeleted, please propose a change of the policy. -- Bene* talk 16:50, 22 April 2013 (UTC)
So we count that pages in the Wikipedia namespace are also user pages? That's clear then... --Stryn (talk) 18:37, 22 April 2013 (UTC)
I think that you want to delete also this: Q4655354. --Stryn (talk) 18:41, 22 April 2013 (UTC)
See also Q11478627. -- Lavallen (block) 18:46, 22 April 2013 (UTC)
I hope the sysop doing that will not forget to restore local interwikis. Infovarius (talk) 10:44, 23 April 2013 (UTC)
Noticing your reactions, I would think that we can restore this item as an exception. I only deleted it because technically it is in the user namespace on Wikipedia. However, we can call this an "almost Wikipedia namespace page" and keep it with this reason. -- Bene* talk 16:10, 23 April 2013 (UTC)

Adding generic descriptions on a massive scale ?

Hi,

I saw that some bots are adding generic descriptions in several languages (same description for all Wikidata items linked to categories on Wikipedia, or to templates, ...). Is there any point in doing that ? I find this rather useless (what's the point of describing a category as being a "category" ?) and rather counter productive (difficult to find items with missing descriptions). What do you think?

Examples of bots doing this: KLBot2, ThieolBot, ... --NicoV (talk) 12:10, 22 April 2013 (UTC)

In my opinion some of these descriptions should fill by wikibase by default.Yamaha5 (talk) 12:19, 22 April 2013 (UTC)
I am not exactly sure what it is, that are observed. But it make sense to me to describe items after what wikipedia categories the linked articles belong to. If, for example, an article is in the category "German actors" somewhere, the description "German actor" is probably appropiate. However maybe it should be checked if the article also belongs to other categories. If the article is also in, say, the category "German painters", a better description would be "German actor and painter". I see no problem if a bot does soemthing like that. Byrial (talk) 14:59, 22 April 2013 (UTC)
@Byrial, there are issues with your proposal. As an example, George Washington is in the category 1799 deaths, but "1799 death" is hardly an appropriate description. Also, articles are sometimes in loads of categories, so how would a bot know which one makes sense....Jakob Megaphone, Telescope 15:07, 22 April 2013 (UTC)
Well, the bot should of course be told which categories (including their subcats) can used to form descriptions and also what to do if the article is in several of them (merge the description according to some rules or give up). It is not simple, but I think it could be done. Byrial (talk) 15:17, 22 April 2013 (UTC)
PS. I see now that there are links to two of these bot edits. Such a description without a label make no much sense. However if also a label was added giving the category's name (impossible for a bot to do automatically if the category doesn't exist in the Wikipedia of the language) I guess it would be OK. Byrial (talk) 15:17, 22 April 2013 (UTC)


I have reported this because both bots are adding a generic description, not anything really meaningful (see the examples given as link in my first post):

  • KLBot2 is adding "categoría de Wikipedia" (Wikipedia category) as Spanish description for items linked to a category in some wiki, but that don't have a link to Spanish wiki. It's also doing the same kind of things for templates ("plantilla de Wikipedia" = Wikipedia template)
  • ThieolBot is adding "page de catégorie de Wikipédia" (Wikipedia category) as French description for items linked to a category in some wiki, but that don't have a link to French wiki.

So, many Wikidata items end up with a completely meaningless description. What's the point ? --NicoV (talk) 15:47, 22 April 2013 (UTC)

I really do not get what is wrong with that. The bot add "category page" in the description of category pages and "template page in the description of template pages. Sitelinks for templates and categories should not really contain real data-ish content, apart from a few maintenance oriented statements, and should not be used in other items, so it does not seem very useful to have very precise description. The descriptions that added by the bots may not be very useful, but they do not hurt and they help cleanup the list of items needing description.
As for the point that it adds description in languages for which there is no sitelink, how exactly is it a problem ? --Zolo (talk) 16:00, 22 April 2013 (UTC)
I just think that adding meaningless description is worse than having no description at all: no description is a clear statement, easily understood by bots and humans; when you have a meaningless description, it's more difficult for a bot to know that the description has no real meaning.
Usually, a category in itself has a meaning, it groups things that have something in common, so I thought that the description should reflect the purpose of the category, not just be "Category". I think that if categories/templates/... don't really need a description, then the "list of items needing description" should take this into account, not needing to put a dummy description everywhere...
Just my opinion. --NicoV (talk) 16:35, 22 April 2013 (UTC)
The reason I see this as necessary is because the descriptions are used when going to add claims (via the autofill). When you have 10 people named George Washington, or 2 categories named Category:Birds (for example), or a disambiguation page named Volcano and a page based on an actual volcano, the description is absolutely necessary. It is obviously less useful for templates and categories, but I'm not sure that it shouldn't be added in those cases anyway. --Izno (talk) 18:20, 22 April 2013 (UTC)
I guess we could have generate better descriptons for category and list pages using P:P301 and P:P360 if that seems necessary. --Zolo (talk) 08:45, 23 April 2013 (UTC)

Phase 2 live on English Wikipedia

Heya folks :)

Phase 2 is now live on English Wikipedia. For more details please see the blog entry. The remaining Wikipedias are still scheduled for Wednesday.

At the same time we have enabled a new feature that lets the user add another language link to an article without having to go to Wikidata if there is no other language listed on Wikidata yet. --Lydia Pintscher (WMDE) (talk) 19:23, 22 April 2013 (UTC)

In the meanwhile, we are trying to set an RFC on the issue (in particular, see the talk page).--Ymblanter (talk) 19:27, 22 April 2013 (UTC)
Sorry but it is not the role of wikidata to decide how data will be implemented in wikipedia. Don't mix the roles of the different communities. Snipre (talk) 08:39, 23 April 2013 (UTC)
#Sitelinks-id works only if both, Wikipedia and Wikidata are used in the English language. --Stryn (talk) 20:30, 22 April 2013 (UTC)
Can you please clarify what you mean? I don't see a change when using it in English as opposed to German currently. --Lydia Pintscher (WMDE) (talk) 20:49, 22 April 2013 (UTC)
Oh, sorry. I thought that #Sitelinks means that it links to section where is all sitelinks, but now I see that I was wrong. --Stryn (talk) 20:59, 22 April 2013 (UTC)
Yes that is how it is supposed to work. But I think it still needs the corresponding update here to actually work as expected. I'll ask to make sure. --Lydia Pintscher (WMDE) (talk) 21:00, 22 April 2013 (UTC)

Problem linking de:WP

I cannot seem to link the de:WP article "Schlagstein (Archäologie)" to its counterparts in other languages ("en:Hammerstone", "fr:Percuteur", etc.). When I try, I get an error message telling me "Site link [[dewiki:Schlagstein (Archäologie)]] already used by item [[Q2238201]]." I have no idea what the problem is. Can someone help with this? Kelisi (talk) 02:37, 23 April 2013 (UTC)

It means that the de article you are trying to link to Hammerstone etc. is already linked on the item in the error message, Q2238201. To fix that, you should read up on Help:Merge. --Izno (talk) 02:54, 23 April 2013 (UTC)
  Fixed I've just merged it into Q2915181.  Hazard-SJ  ✈  03:04, 23 April 2013 (UTC)

How are base URL being defined for property referring to URL

I noticed that for some property like P345 (IMDb identifier) are automatically integrated to use the "http://imdb.com?Name=" as base uri. Where is the base url defined? is it hardcoded in the wikidata client app? I cannot seem to find it in the property page http://www.wikidata.org/wiki/Property:P345 How are new property similar to imdb defined to use their respective base url? --Napoleon.tan (talk) 08:16, 23 April 2013 (UTC)

I think the gadget AuthorityControl adds these links. — Ltrl G☎, 09:14, 23 April 2013 (UTC)
Thanks --Napoleon.tan (talk) 03:21, 24 April 2013 (UTC)

Item without article

I was just adding statements to Félix Iñurrategi and I wanted to say "occupation = mountaineer". But I couldn't find an item "mountaineer" (in that sense) - there isn't such an article in enwiki (or frwiki) - the term redirects to "mountaineering". So I created a new item Q11774620, without links to any articles. Is this the right way to proceed? My belief is that Wikidata is meant to be a repository independent of the Wikipedias, but I'm not sure. --ColinFine (talk) 12:59, 23 April 2013 (UTC)

yes, thats fine as long the item is used and translated for some languages. (I believe occupations will be on the notability inclusion list next round automatically). 139.63.60.122 13:06, 23 April 2013 (UTC)
There is no longer any notability inclusion list, as it was really unmaintainable, but I think it clearly fits WD:Notability.
We need a lot of new items for the "occupation" property. I think the most efficient way to do it is to complete https://docs.google.com/spreadsheet/ccc?key=0AgsXIfkI03IzdHRyQ0RwRTZQeUtiTmxxaUE1UVNtT1E#gid=0 and the have a bot create the items. --Zolo (talk) 13:16, 23 April 2013 (UTC)
Thanks. Mountaineer is in that spreadsheet. --ColinFine (talk) 13:44, 23 April 2013 (UTC)
I think we really missed something here; the problem seems to me that we chose the wrong name for the type. Creating a new item in this case (in general, with respect to "occupation") seems to fail at saying how these people are connected to the overall activities. So it seems to me that using the word "occupation" was wrong. Just musing. --Izno (talk) 16:35, 23 April 2013 (UTC)
Yes, I am still wondering if there is any point in distinguishing it from Property:P101, I tend to think no, except that "occupation" is probably more intuitive. --Zolo (talk) 16:56, 23 April 2013 (UTC)
Ah, that's the one we should probably be using. Why aren't we using it? Would you be up for an PfD? :) --Izno (talk) 16:58, 23 April 2013 (UTC)
The current scheme has its problems but using only P101 would be worse (I think). For queries, it's good to have a natural way of distinguishing players from, say, coaches and you can't achieve that if you delete the "occupation" property. Pichpich (talk) 17:21, 23 April 2013 (UTC)
Couldn't you set to parameters on field of work "sports coaching" and "baseball"? --Izno (talk) 18:52, 23 April 2013 (UTC)
The way I look at it is that we could have occupation=historian but field-of-work=Ancient Egypt or occupation=mathematician but field-of-work=Number theory -- Ypnypn (talk) 18:07, 23 April 2013 (UTC)
See my above example. There's nothing stopping us from using field of work to define "mathematics" and "number theory" (or in that case, inference is trivial so we could just say number theory if the person is actually a mathematician and not a computer scientist interested in it), and in the case of ancient Egypt, make the claims of both "ancient history" (or "history", or whatever), and of "ancient Egypt". --Izno (talk) 18:52, 23 April 2013 (UTC)
This is mostly a question of taste and both schemes can work. I still prefer the current one and it might also be easier to use because it most resembles the current use of infoboxes for people. Pichpich (talk) 19:05, 23 April 2013 (UTC)
  • I think that now that we will allow redirects, you can add the "mountaineer" redirect link to that item. However, i also think it fits the purposes of notability. — ΛΧΣ21 15:04, 24 April 2013 (UTC)

Problem of interwiki

These 2 wikidata pages contains 3 corresponding interwiki, but I can not modify them: It is necessary to have only one wikidata with the corresponding 3 pages.

http://www.wikidata.org/wiki/Q4938723

http://www.wikidata.org/wiki/Q3200875

Thanks for your help. --Rédacteur Tibet (talk) 21:07, 23 April 2013 (UTC)

Consider reading Help:Merge for the future. I took care of this one. --Izno (talk) 21:17, 23 April 2013 (UTC)
Many thanks for all this. --Rédacteur Tibet (talk) 21:24, 23 April 2013 (UTC)

New RFC on possible interproject link interfaces

Hi, I have started on Meta an RFC on possible interproject link interfaces for Wikipedia, Commons, Wikisource, Wikivoyage, etc. --Micru (talk) 02:01, 24 April 2013 (UTC)

Request for comment on inactive administrators

(Please consider translating this message for the benefit of your fellow Wikimedians. Please also consider translating the proposal.)

Read this message in English / Lleer esti mensaxe n'asturianu / বাংলায় এই বার্তাটি পড়ুন / Llegiu aquest missatge en català / Læs denne besked på dansk / Lies diese Nachricht auf Deutsch / Leś cal mesag' chè in Emiliàn / Leer este mensaje en español / Lue tämä viesti suomeksi / Lire ce message en français / Ler esta mensaxe en galego / हिन्दी / Pročitajte ovu poruku na hrvatskom / Baca pesan ini dalam Bahasa Indonesia / Leggi questo messaggio in italiano / ಈ ಸಂದೇಶವನ್ನು ಕನ್ನಡದಲ್ಲಿ ಓದಿ / Aqra dan il-messaġġ bil-Malti / このメッセージを日本語で読む / norsk (bokmål) / Lees dit bericht in het Nederlands / Przeczytaj tę wiadomość po polsku / Citiți acest mesaj în română / Прочитать это сообщение на русском / Farriintaan ku aqri Af-Soomaali / Pročitaj ovu poruku na srpskom (Прочитај ову поруку на српском) / อ่านข้อความนี้ในภาษาไทย / Прочитати це повідомлення українською мовою / Đọc thông báo bằng tiếng Việt / 使用中文阅读本信息。

Hello!

There is a new request for comment on Meta-Wiki concerning the removal of administrative rights from long-term inactive Wikimedians. Generally, this proposal from stewards would apply to wikis without an administrators' review process.

We are also compiling a list of projects with procedures for removing inactive administrators on the talk page of the request for comment. Feel free to add your project(s) to the list if you have a policy on administrator inactivity.

All input is appreciated. The discussion may close as soon as 21 May 2013 (2013-05-21), but this will be extended if needed.

Thanks, Billinghurst (thanks to all the translators!) 05:30, 24 April 2013 (UTC)

Distributed via Global message delivery (Wrong page? You can fix it.)

Semitic alphabets

In the English Wikipedia, there is one article for each letter in all Semitic languages. So there is just one "Aleph", one "Qoph", and so on. But in other Wikipedias, there are several articles – one for the Hebrew letter, one for the Arabic letter, and so on. Therefore, the English articles don't have an exact counterpart in the other articles. But it still seems right to include some interwiki links. What should we do? – Ypnypn (talk) 17:53, 24 April 2013 (UTC)

Given the English description of aleph, at least, I would include all of other wikis' articles on the Hebrew letter, or where possible, other wikis with "combination" articles. A separate item should be set up for the Arabic letter.

This will only work where the description clearly says "this is the Hebrew letter". What about others? Are there other items with the en item linked on which the item does not have a description of "this is about the Hewbrew letter"? --Izno (talk) 20:03, 24 April 2013 (UTC)

The description says it's about the Hebrew letter, but the article is clearly about all alphabets. – Ypnypn (talk) 22:32, 24 April 2013 (UTC)
I think they should be split into separate items. Interwikis from the articles about specific letters can have interwikis to the less specific articles directly through wikitext interwikis. --Yair rand (talk) 22:42, 24 April 2013 (UTC)
Okay, I split aleph. Only twenty-something left to go ;-) -- Ypnypn (talk) 23:40, 24 April 2013 (UTC)
I did the remaining twenty-one. Phew! Ypnypn (talk) 00:06, 25 April 2013 (UTC)

Phase 2 live on all Wikipedias

Heya folks :)

The start of phase 2 has just been deployed on all 274 remaining Wikipedias \o/ http://blog.wikimedia.de/2013/04/24/wikidata-all-around-the-world/ --Lydia Pintscher (WMDE) (talk) 19:00, 24 April 2013 (UTC)

thank you. the First bug :)
for fa.wikipedia it doesn't switch to the Farsi and it still English! here look in the box.
  • بنیانگذار(ها) Larry Page، Sergey Brin
  • شعبه مرکزی Googleplex، Mountain View

their name should be like this. Yamaha5 (talk) 19:25, 24 April 2013 (UTC)

Hmm yes you are right. This is a bug. We will have a closer look. THanks for reporting it. --Lydia Pintscher (WMDE) (talk) 19:30, 24 April 2013 (UTC)
While we're at it, nowiki is anomalously supposed to get "nb" labels, not "no" labels, when the bug is fixed. Do you know if this (and similar situations) has been taken into consideration? - Soulkeeper (talk) 20:04, 24 April 2013 (UTC)
Soulkeeper: I am not sure tbh. When the bug is fixed (hopefully very soon - people are working on it right now) can you check what is happening? If it is different from what it is supposed to be please let me know and I'll bring it up with the team. --Lydia Pintscher (WMDE) (talk) 20:48, 24 April 2013 (UTC)
Thanks, I will. Here's to hoping the language is pulled from a variable on the individual Wikipedia project, not "hard coded" to be extracted from the subdomain name. We'll see. :) - Soulkeeper (talk) 21:01, 24 April 2013 (UTC)
The bug should now be fixed. --Lydia Pintscher (WMDE) (talk) 21:52, 24 April 2013 (UTC)
Q134862 for everybody. HenkvD (talk) 21:39, 24 April 2013 (UTC)
Wohooo! ;-) --Lydia Pintscher (WMDE) (talk) 21:52, 24 April 2013 (UTC)
Q154168 :) - Soulkeeper (talk) 22:00, 24 April 2013 (UTC)

Problem with Q11777556

Looks like a bug: there is no links from English and Polish Wikis despite the pages purging. And ItemByTitle cannot find this item too, for example see [18]. --Emaus (talk) 20:17, 24 April 2013 (UTC)

amazing point I made Q11846722 :) now we have two item with the same language links ! 131.175.62.218 20:37, 24 April 2013 (UTC)

Importing data from Persondata templates

On English Wikipedia, there is a template called Persondata (and according to the interwiki links, it's present on several different languages; I know enwp borrowed the idea from dewp) that provides machine-readable data concerning a person's birth name, aliases, date of birth, place of birth, date of death, place of death, and a short description. Obviously, this would be a fantastic data source for Wikidata, if we could get a bot to snag all the data (with the caveat that it might have to wait until we get a date datatype to actually run). Any takers? —Scott5114↗ [EXACT CHANGE ONLY] 20:53, 24 April 2013 (UTC)

See Wikidata:Bot requests#descriptions for persons from persondata. Also, we're still working on date datatypes, so that'll have to wait. FallingGravity (talk) 21:04, 24 April 2013 (UTC)
Ah, figures I'd be late to the party on this one. Thanks! —Scott5114↗ [EXACT CHANGE ONLY] 21:12, 24 April 2013 (UTC)

and / or

Hello, I want to state that according to one source, Q540716 has been made by the Andokides painter alone and that according to another source, it has been made by both the Andokides and the Lysippides painter, and perhaps also that a third source states that it has been made by either the Andokides or the Lysippides painter. Any idea how to do that in a non-ambiguous way ? --Zolo (talk) 22:16, 17 April 2013 (UTC)

From what you say the sources give ambiguous information. We should reflect that ambiguity.
Add three statements, each one giving the info from one of the three sources with qualifiers giving the source exactly as has been done on this page.
If one source is considered 'preferred' then add a qualifier to mark it so (Do we have a property for this yet?). Filceolaire (talk) 11:25, 18 April 2013 (UTC)
Yes, things would be rather simple if we would allowed to have:

creator

source A
painter X
source B
painter X and painter Y
source C
painter X or painter Y

However, Wikidata's structure is property -> value + qualifiers -> sources, which gives something like

creator:

painter X
source A
source B
painter X
qualifier: maybe not
source C
painter Y
source B
painter Y
qualifier: maybe not
source C

I am not sure that it is really intellegible. --Zolo (talk) 14:26, 18 April 2013 (UTC)

Imho the Or relation could be expressed by a property or an Item, with alternatives as qualifiers :
<painting> Creator <uncertained>
Qualified by:
  • Alternative <X>
  • Alternative <Y>
  • source <C>
 – The preceding unsigned comment was added by TomT0m (talk • contribs) at 15:46, April 18, 2013‎ (UTC).
Something in this spirit ? I would be satisfied with it, except that we would probably need to add qualifieres to qualifiers (in this case, one author states that Painter X made side A and Painter Y made side B). --Zolo (talk) 06:45, 19 April 2013 (UTC)
I assume you meant to put Q1145523, and not Q2982709? If so, it seems like a reasonable solution, however we need to make sure that this solution is used consistently in all instances of this problem, and I'm not sure how to do that. An entry in help documentation somewhere? Does that need a discussion and consensus first?
Will qualifiers for qualifiers be possible? You could also express that information by creating items for side A and side B and relating them via part of.
Silver hr (talk) 01:37, 23 April 2013 (UTC)
Yes, Q1145523 thanks :). It seems indeed that the only way we can compeletely describe this in a reasonably straightforward way would be with subitems for side A and side B (or alternatively for painter X + painter Y), though that that makes things a bit unintuitive. -Zolo (talk) 07:11, 25 April 2013 (UTC)
This is an interesting problem, and it reveals a weakness in Wikidata's design. The basic question is "What do multiple statements with the same property mean?". In some cases, such as book authors, the meaning is obviously "and". In others, such as with conflicting statements, the meaning is "or". Unfortunately, I haven't found anything in the data model to resolve this, and the solution is not obvious to me. I'll try to break the problem down.
  1. Same-property statements with the same source.
    • Example:
      • Historical book A is authored by X according to source S.
      • Historical book A is authored by Y according to source S.
    • Options:
    1. Do we combine them with a logical "and"? (Let's call these an and-group.)
      • Meaning: both X and Y are authors.
      • Problem. How do we state that a single source stated a logical "or" (one or the other or both)?
      • Problem. How do we state that a single source stated a logical "exclusive or" (English "or", one or the other but not both)?
    2. Do we combine them with a logical "or" (one or the other or both)?
      • Meaning: either X is the sole author, or Y is the sole author, or both X and Y are authors.
      • Problem. How do we state that a single source stated a logical "and"?
        • Potential solution. Create an item for the and-group (e.g. Q219937) so you only have to make one statement.
      • Problem. How do we state that a single source stated a logical "exclusive or"?
    3. Do we combine them with a logical "exclusive or" (English "or", one or the other but not both)?
      • Meaning: either X is the sole author, or Y is the sole author.
      • Problem. How do we state that a single source stated a logical "and"?
        • Potential solution. Create an item for the and-group (e.g. Q219937) so you only have to make one statement.
      • Problem. How do we state that a single source stated a logical "or"?
        • Potential solution. Create an item for the and-group and make 3 statements: one for each item and one for the and-group. X xor Y xor (X and Y) <=> X or Y.
  2. Same-property statements with different sources.
    • No need to interpret this. Different sources can make different claims, and all, some or none could be true.
Well, after my analysis, it seems only option 1.3 is a satisfactory solution. If anyone sees any solutions for the problems in the options, please add them. Also, if we were to adopt option 1.3, there would still be the problem of contributor awareness--how do we make sure everyone adding such statements knows about the intended meaning without the interface telling them so? Also, TomT0m's solution is interesting, but it feels too much like a workaround. I think the best solution would ultimately be if Wikidata had native support for explicitly stating either "and", "or", or "xor".
Silver hr (talk) 21:09, 18 April 2013 (UTC)
An additional qualifier to the statements could solve some of these problems by marking a statement as "Co-author" or "Disputed" for instance. This would need a general "Comment" property. To make the relationship machine readable a separate qualifier with values "XOR", "AND" xor "OR" could also be used. This would need a new property. In my opinion both of these properties would be useful as qualifiers and should be used together. Filceolaire (talk) 06:42, 19 April 2013 (UTC)
I agree, a way (standard) way of expressing boolean expression could be use usefull, but is overkill in this case. TomT0m (talk) 09:29, 19 April 2013 (UTC)
@ Silver hr I had a similar reasoning. If it was possible to group a set of claims with a qualifier it would possible to annotate them with a or and or xor qualifier. Or maybe reverse the stuff : add to the statements we want to regroup a qualifier part of group <group item>. <group item> would be an instance of or/and/xor group item. This would work, but it is complex. I think the two choices would be : one "author" statement for this source in the work item, with the boolean expression Item attached to it, and the author statement, or N author statement attached by a group item qualifier. The second would pro bably be cleaner, but we would have to cleanly define a boolean expression and say that the author statement might receive a Boolean expression, which seems overkill.
But having boolean expressions does not need a special treatment of Wikidata, just "Boolean" types to type items, a several items to make a composite statement. TomT0m (talk) 09:29, 19 April 2013 (UTC)

Meta data for creative works — schema.org

At English Wikisource, someone has popped their head in to talk about metadata s:en:Wikisource:Scriptorium#Implementing Easier and More Efficient Metadata Tags on Wikisource which is something that we can look to do. In that post the person links to http://schema.org/CreativeWork and I wondered what the WD community was doing with regard to that project. Obviously in the long term (shorter to medium term????) the Wikisources would love to be have metadata stored and applied to our transcluded works (now over 1000 fully validated works, 280k pages in main ns). We would like to be working towards both goals, and I would hope that WD would be looking at the schema too. Is there feedback that I can give the WS community? Thanks.  — billinghurst sDrewth 15:20, 23 April 2013 (UTC)

Hi Billinghurst, yes I know about that. Aubrey and me discussed with Max that it wouldn't make sense to support the whole specification, because some items seem too specific (intended age), but about book properties we are having the discussions in the books task force. Feel free to come and discuss about the suggested properties. This Thursday we are also having an IRC conversation about the RFC about sources and references which can also be relevant to Wikisource. Summing up, there is a lot going on about the interactions between Wikidata and Wikisource, on the next days we will post the first document with proposals to the Wikisources as we are expected to do following the Elaborate Wikisource strategic vision project. Keep tuned! --Micru (talk) 22:34, 23 April 2013 (UTC)
Moreover, I'd like to add that I tried to map some of the http://schema.org/Book metadata with our templates and properties. It's not complete nor perfect, but at least for books it's a start. Of course, it would be awesome if other users would take the CreativeWork schema and do the same. --Aubrey (talk) 11:11, 24 April 2013 (UTC)
Yes Aubrey, I agree that if others adapt the CreativeWork schema as well, it would be all the more effective and helpful. In regards to being too specific, I think that now that the LRMI metatags are included as part of Schema, there are more possibilities to use other Schema tags more specified to the needs of both WS and WD. Another benefit of now being part of Schema is that all mediawiki installations are sanctioned by default and will undoubtedly increase the assurance of LRMI as a long-term standard. I will be following the IRC conversation, and hope to discuss more. Maximilian.Klein.LRMI (talk) 15:41, 25 April 2013 (UTC)

Adding labels of proper names to multiple languages

As suggested on User talk:Kizar#Labels in multiple languages, I think that it would be a good thing if the user's bot would not just add the label of proper names for Spanish, but for all languages that use the Latin alphabet.

  • Any opinions or suggestion on this?
  • Do you have a list of such languages ready?

Thank you. --Leyo 11:30, 25 April 2013 (UTC)

Agree with the generally idea, but I can understand that the bot's operator would start with one version he is comfortable with. --  Docu  at 11:35, 25 April 2013 (UTC)
A possible problem are transcribed names from other alphabets. Even though Pyotr Ilyich Tchaikovsky is a proper name in English, it cannot be copied 1:1 to other languages with latin alphabet (e.g., it's Piotr Ilich Chaikovski in Spanish, Pjotr Iljitsch Tschaikowski in German, etc.). But of course, my example already shows that this is problem exists no matter if you handle all languages at once or only one single language. --YMS (talk) 12:25, 25 April 2013 (UTC)
(EC) An intermediate option would be to leave out Slavic languages (and Latin). They often have -ova or similar. Maybe just the following Western European languages: en, de, als, fr, it, es, pt, nl, da, no, sv? --Leyo 12:27, 25 April 2013 (UTC)
Doesn't work consistently as proper names may be translated (e.g., en:London = fr:Londres). How do you determine if it is a translated or non-translated proper name? QuiteUnusual (talk) 12:38, 25 April 2013 (UTC)
I think having a starting ground would be good, as the name can be translated "further" if necessary. In other words, the bot should only run on labels which are currently undefined. This is actually something I wanted to request, because it is common for video games to take the same name in multiple languages (c.f. Dota 2).
What would be better is if this could be done via the autoEdit gadget. --Izno (talk) 13:10, 25 April 2013 (UTC)
I disagree with "having a starting ground would be good". Now we have two states for labels: "Is empty" or "Is probably correct". When filled with something that might be correct or not we'll end up only having one state for labels: "Could be correct or just some foreign gibberish". --YMS (talk) 13:58, 25 April 2013 (UTC)
And I disagree with that because Wikidata is completely unusable without those states being filled. Try searching for any odd item and you'll see what I mean. You won't be able to see heads from tails of all the different usages of "mercy". I imagine it's much worse in non-English languages.
If we had language fallbacks, this wouldn't be a problem. But we don't for some reason, and we really should. --Izno (talk) 14:18, 25 April 2013 (UTC)
Is a language fallback what I just wrote above (under ""en-us" locale or "en-us" = "en"?")? QuiteUnusual (talk) 15:32, 25 April 2013 (UTC)
"wikidata_useful" prefills empty labels with other languages. "reasonator" does even more ( http://toolserver.org/~magnus/ts2/reasonator/?q=Q1339 ). --  Docu  at 15:34, 25 April 2013 (UTC)

Abusefilter-view-private and abusefilter-log-private for rollbackers

I suggest to add the abusefilter-view-private and abusefilter-log-private permissions to the rollbacker usergroup, because I think this would make private filter settings more transparent to constructive editors, as the only reason why some filters are private is to prevent them from being seen by vandals, trolls or spammers. IMHO rollbackers could also help more effectively monitoring the abuse filter log, recognizing spambots etc. if they had these rights. Regards --Iste (D) 20:28, 11 April 2013 (UTC)

  • I agree. But also autopatrollers have the required level of trust, in my opinion. Do you suggest it only for rollbackers just because they may have more need for these permissions as they are active in vandal fighting, anyway? Regards, Vogone talk 20:33, 11 April 2013 (UTC)
  •   Support abusefilter-log-private only, for rollbackers. Autopatrollers may be very new users who may not be necessarily trustworthy enough to view those filter log entries, while I feel that the viewing of private filters themselves should be left to sysops only (or to a new "abusefilter" group as used on Wikipedia and some other wikis).--Jasper Deng (talk) 20:46, 11 April 2013 (UTC)
  •   Support for rollbackers per Jasper. Autopatrollers are often new users and rollbackers would have more need for this permission. Lukas²³ talk in German Contribs 21:00, 11 April 2013 (UTC)
  •   Support for rollbackers only. --Rschen7754 22:37, 11 April 2013 (UTC)
  • Support giving rollbackers both. Being able to view what a private filter does is hardly dangerous for trusted users like our rollbackers, and who knows, one might even find a mistake in a filter someday which could then be fixed. Ajraddatz (Talk) 23:14, 11 April 2013 (UTC)

Hold your horses: I cannot see that any (local) usergroup here today have access to any of these permissions: See Special:ListGroupRights. And according to meta:Abuse filter abusefilter-view-private is something that isn't used as a default settings, and is today only used of four wmf-projects. Do we really need these rights at all, for any group? I also know that some of the abusefilter-settings are not allowed on wmf-wikis, since they give access to private information. I would like to see some more research before any of this go any step further. -- Lavallen (block) 06:14, 12 April 2013 (UTC)

  • Anyone with abusefilter-modify has both of those (i.e. that right allows anyone with it to do the same things as these proposed rights).--Jasper Deng (talk) 06:17, 12 April 2013 (UTC)
    • Jasper is right, see mw:Extension:Abusefilter#User rights – abusefilter-view-private and abusefilter-log-private are included in abusefilter-modify. The only right which is not allowed on WMF projects is abusefilter-private, as this gives access to IP addresses of registered users, but that's not what I proposed. Regards --Iste (D) 09:37, 12 April 2013 (UTC)

There have been no more comments about this proposal for one week. From my point of view, there seems to exist consensus about giving abusefilter-log-private to rollbackers, as the only objection is due to a misunderstanding about the current user rights constellation. I'm going to file a bug for adding abusefilter-log-private to the rollbacker flag if no one comments on this with in the next few days. Regards --Iste (D) 18:14, 19 April 2013 (UTC)

This has been put on hold until there is wider participation & more consensus as to what you really want to have done. odder (talk) 01:07, 26 April 2013 (UTC)
I've created an RfC here. Go vote! FrigidNinja 01:35, 26 April 2013 (UTC)

Merging Property:P7 and Property:P9

I find it to be highly redundant to have different properties for brother (P7) and sister (P9), we don't have different properties for child (male) and child (female) either. This is already done by Property:P21 (sex). So I propose we merge these together to Sibling. —★PοωερZtalk 13:16, 25 April 2013 (UTC)

Agree, that unnecessary, and possibly annoying (we allow intersex, but we do not have "intersex sibling" property). Even in pureley linguistic terms, I doubt that it sounds very good in all languages. --Zolo (talk) 13:35, 25 April 2013 (UTC)
Ditto for father/mother (parent), uncle/aunt (parent's sibling), and stepfather/stepmother (stepparent). - Soulkeeper (talk) 14:04, 25 April 2013 (UTC)
Father and mother should be separate properties, because each person has only one mother and only one father. Rsocol (talk) 14:13, 25 April 2013 (UTC)
Depends highly on the definition. Maybe we should consider renaming it to biological mother/father. —★PοωερZtalk 14:19, 25 April 2013 (UTC)
If you really want to go down that path, all that is needed is "parent" (mother/father) or "child"... We had that discussion a while ago. I would personally prefer that we not have all of those parameters and that the relation be simply queryable, but the consensus went the other way (and I wasn't here then! :). --Izno (talk) 14:21, 25 April 2013 (UTC)
  Support. "brother" and "sister" can be left as aliases, though. --Yair rand (talk) 16:02, 25 April 2013 (UTC)
  Support for brother and sister, Rsocol's right about parents. – Ypnypn (talk) 16:51, 25 April 2013 (UTC)
  Oppose someone is a brother or a sister, we don't even have a real word for that in Dutch. Multichill (talk) 18:08, 25 April 2013 (UTC)
So just label it the translation of "brother or sister". --Yair rand (talk) 18:13, 25 April 2013 (UTC)
  Support for merging brother and sister. Tpt (talk) 18:35, 25 April 2013 (UTC)
Above supporters/opposers, could someone please consider submitting these two properties at WD:PFD? Thanks. --Izno (talk) 20:10, 25 April 2013 (UTC)

  Comment We obviously have to talk about this in greater detail. I created User:23PowerZ/Stuff for a more general discussion. —★PοωερZtalk 02:54, 26 April 2013 (UTC)

Maybe you could move the page to WD:RFC? --Stevenliuyi (talk) 03:04, 26 April 2013 (UTC)
  Done —★PοωερZtalk 03:07, 26 April 2013 (UTC)

Property has two values, how to pick one?

I was trying out phase two. Since the example I found gave Q37093, that's what I tried. However, the property "founder" has two values. How do I pick just one of them? I can't find this in the FAQ, is it explained somewhere else? Aurora (talk) 15:56, 25 April 2013 (UTC)

You need a selection tool in your wikipedia based on some criteria: more recent data (used the date of the reference for that), data quality (there is a quality parameter for each statement) or the language of the reference ... But for that you will need to use lua model. Snipre (talk) 16:54, 25 April 2013 (UTC)
That sounds overly complicated. Unless there's an easier way to do separate the values, this kind of multi-valued property won't be very useful. Aurora (talk) 02:27, 26 April 2013 (UTC)

sort order of items with the same name

I know this is an old problem/old news, but just raising it again -- the sort algorithm for suggested items when there are many items sharing the same name really needs work. The worst example I've found so far is trying to add the U.S. state of California, which ranks below dozens of other places, albums, etc with that name when you try to add it as a property value to an item. Pretty irritating, especially for new editors. best, -- Phoebe (talk) 18:31, 25 April 2013 (UTC)

Seconded. Trying to classify different subclasses of of 'cancer' is difficult because of the same issue. Would it be feasible to sort the autocomplete by the number of items that link to the various suggestions? That seems like a crude but roughly effective solution for many (most?) cases. Emw (talk) 04:26, 26 April 2013 (UTC)
I think that would be a fine solution for most cases, yes. -- Phoebe (talk) 04:40, 26 April 2013 (UTC)

Shorthand for ['[Q123456|Title of page]']

Is there a shorthand for links to items that displays the title of the page. If not, there should be. I want to use these links so they get automatically translated. Would be nice if that would also work for properties. Typing for example ['[Property:P138|#]'] (# meaning: put in title of page in the users language) would produce a link named "benannt nach" when the user has set the language to German. --Tobias1984 (talk) 15:42, 19 April 2013 (UTC)

  Support - Very good idea! Sure there must be a fallback language if for the user language no entry exist. --Nightwish62 (talk) 07:48, 20 April 2013 (UTC)
You could use an template for that. Not sure about the fallbacks though. So, where are you going to use this shorthand?--Snaevar (talk) 09:27, 20 April 2013 (UTC)
Do not add any fallback, it's a good reason to add your own language. I'm already frustrated that I see other languages in Special:Allpages and Special:RC, I would prefer to not, to find those who need a translation. Today I cannot always separate those who use a fallback from those who is translated. -- Lavallen (block) 10:54, 20 April 2013 (UTC)
Is it at all possible here? This is the repository wiki, not a client like the Wikipedias. Byrial (talk) 11:05, 20 April 2013 (UTC)
I want to use it in the table Wikidata:Mineralogy_task_force/Properties in the column "creation level". I already used language switches to translate some of the headlines. But it would be convenient if the translation of links could be shorthanded instead of putting in another language switch and copy-pasting all the translations. --Tobias1984 (talk) 11:27, 20 April 2013 (UTC)
@Byrial: Well, I am not going to give an definite answer, since I'm not an Lua coder, but yes, it is possible. That would be done by requesting the content of the whole item with an Lua module and then only use the label (or title as Tobias1984 decides to call it).--Snaevar (talk) 11:37, 20 April 2013 (UTC)
@Snaevar: And what makes you think you have access to these data from Lua modules. It is the client extension which the Wikipedias use that give that access as I understand. Byrial (talk) 12:03, 20 April 2013 (UTC)
@Byrial: Jämför detta med det du ser i RC. Någonstans översätts det ju i http-versionen av RC, så då borde samma sak kunna ske på andra platser. -- Lavallen (block) 12:42, 20 April 2013 (UTC)
Teksterne som er oversat i Seneste Ændringer er alle beskeder fra MediaWiki-navnerummet, ikke etiketter eller andet data fra entiteter. Jeg kan ikke se noget tegn på tilgang til disse nogensteder. Byrial (talk) 13:08, 20 April 2013 (UTC)
And in English: The translated texts in RC are messages from the MediaWiki namespace, not labels or other data from entities. I see no sign of access to these anywhere. Byrial (talk) 13:08, 20 April 2013 (UTC)
Currently the Wikidata information can not be used here (on the repository), not even with Lua. For Lua you would need mw.wikibase, which is only part of the Client extension. Would be great to have it avaliable here though. HenkvD (talk) 14:59, 20 April 2013 (UTC)
I made a few simple parser functions (label, description and link) for this nearly two months ago, but it ended in a full halt and a statement that it should not be done in any other way than with Lua. I still think that is a wrong stance, but I guess it won't ever be included. Its primary purpose was to be used in running text on Wikidata. Jeblad (talk) 15:53, 21 April 2013 (UTC)
So you think it would be possible with Lua? It would be very useful, for instance in the deletion requests, list of properties and many other places. Now only the Q- and P- numbers are mentioned. HenkvD (talk) 17:34, 21 April 2013 (UTC)
I actually thought that this would be more of a back-end (is that the right word?) solution. Ideally it would be just a string of wiki-markup and not a long script of some sort. --Tobias1984 (talk) 13:43, 20 April 2013 (UTC)
  Support We need this, and not only on Wikipedia, but here as well. Silver hr (talk) 21:52, 22 April 2013 (UTC)
  Support HenkvD (talk) 21:57, 24 April 2013 (UTC)
this also would help on pages like Wikidata:List of properties. currently translations in items/properties and on meta-pages are redundant. --Akkakk (talk) 23:42, 24 April 2013 (UTC)
It is nice to see that other people would also like this function. Is there a page where these kind of requests can be made? --Tobias1984 (talk) 06:26, 26 April 2013 (UTC)
I was recently directed to m:Wikidata/Notes/Future, however there seems to be very little activity there, so I don't really know. Silver hr (talk) 18:21, 26 April 2013 (UTC)

Circumstances of phase 2 on respective Wikipedias

Hello. Phase 2 (infoboxes!) has just been deployed on the English Wikipedia, and in my understanding, discussion about how to handle it on their site is still ongoing at w:en:Wikipedia:Requests for comment/Wikidata Phase 2. What I want to know here is the current circumstances in other language versions of Wikipedia including the first 11 Wikipedias, because phase 2 is going to be deployed on all remaining Wikipedias in a day or two (24 April). Are Wikipedians allowed to edit any pages including articles and templates on your language version of Wikipedia with the Wikidata inclusion syntax? (no limitation? under discussion?) Or do we already have a general guideline for this? At this moment there is no guideline on the Japanese Wikipedia. I would just like to exchange information with people from other language versions of Wikipedia. Thank you. --Penn Station (talk) 17:38, 23 April 2013 (UTC)

It's up to each individual Wikipedia on what they want to do... if you want to have a RFC for the Japanese Wikipedia, now would be the time to start it. --Rschen7754 05:07, 24 April 2013 (UTC)
I guess while it is up to each community, it would be good to see if we can collect a concise list of what communities have decided or are discussing. Is there anything like that already? Wikis new to the phase 2 might have something to learn from early adaptors. --whym (talk) 13:57, 24 April 2013 (UTC)
Thanks for following, whym. Yes, it's up to each individual project - that's absolutely right. I thought if there already were known issues, discussions and/or solutions (maybe guideline?) in the precedent projects, they would be case studies as helpful inputs or hints for the other projects. It's not only for the Japanese Wikipedia, but also for any projects. --Penn Station (talk) 14:29, 24 April 2013 (UTC)
I know that the Hebrew Wikipedia already had a discussion, but I don't know where the link is - perhaps Lydia knows? --Rschen7754 17:32, 24 April 2013 (UTC)
The discussion I was involved in there was on http://he.wikipedia.org/wiki/%D7%95%D7%99%D7%A7%D7%99%D7%A4%D7%93%D7%99%D7%94:%D7%9E%D7%96%D7%A0%D7%95%D7%9F and is now somewhere in the archive there (I have a really hard time finding it. Sorry.) I don't know if there were further discussions elsewhere. --Lydia Pintscher (WMDE) (talk) 19:20, 24 April 2013 (UTC)
On it.wiki we started today the discussion. I have copied from en.wiki page --ValterVB (talk) 19:51, 24 April 2013 (UTC)
Is the hebrew discussion at he:ויקיפדיה:מזנון/ארכיון_311#Deployment_of_Wikidata_on_the_Hebrew_Wikipedia ?
Hi all, many thanks. For further information, updating Help:Phase 2 on Wikipedia would be very helpful. --Penn Station (talk) 14:37, 26 April 2013 (UTC)
I have started an help page on this matter at Help:Phase 2 on Wikipedia.--Snaevar (talk) 19:57, 24 April 2013 (UTC)
Thank you for prrepareing this page. If I got related info, I'll add it to the page. --Penn Station (talk) 15:17, 26 April 2013 (UTC)

Non-membership property "has a"

I'm new to wikidata, and I hava a basic question. How to say that a car (Q2298934) has a All-wheel drive (Q214833)? --Avron (talk) 09:55, 24 April 2013 (UTC)

All-wheel drive <part of> car. --Izno (talk) 12:57, 24 April 2013 (UTC)
In Q2 <part of> is used in the contrary direction: Earth is Part of Solar System.--Avron (talk) 14:01, 24 April 2013 (UTC)
Precisely. We do not have a property for "has a". So you have to go the other direction. --Izno (talk) 14:04, 24 April 2013 (UTC)
This doesn't make sense for me. If Earth is defined as part of solar System, I would expect that in Solar System I find the information "Solar Systems consists of Earth"--Avron (talk) 14:12, 24 April 2013 (UTC)
One problem in that, is that the solar system has thousands of "parts". -- Lavallen (block) 15:57, 24 April 2013 (UTC)

Lavallen makes the problem clear. For items which have many hundreds or thousands of parts, it is not practical nor particularly meaningful to mark the relationship on the page with hundreds or thousands of parts. It is useful to do it on the "sub"-parts as compositions of the whole.

As an aside, in your initial question, I'm not sure I would associate 4WD with a particular model of car but instead with cars in general. Just a thought. --Izno (talk) 16:18, 24 April 2013 (UTC)

Lavallen doesn't make the problem clear. It was you that made the suggestion using "part of" but it is not feasible. The car was just example. In case you can't associate 4WD with cars in general, becaus only a subpart of cars have a 4WD.--Avron (talk) 18:51, 24 April 2013 (UTC)
I'm aware of the latter. There was a discussion recently which suggested that we should use qualifiers to say "some but not all". I'm not sure if that's more or less valuable than putting it on each car model's page. --Izno (talk) 20:06, 24 April 2013 (UTC)
How about a property named drivetrain? - Soulkeeper (talk) 20:28, 24 April 2013 (UTC)
I see, there is only a special solution. I asked because I thought threre might be a general solution...--Avron (talk) 07:12, 26 April 2013 (UTC)
The general solution would be creating properties for all the special cases. ;) Yes, that's a lot of work, but I believe specialized properties can be more useful to Wikidata than overly general ones. (I've been wrong before, though...) - Soulkeeper (talk) 08:12, 26 April 2013 (UTC)

Q12000000

Q12000000 is about "high altitude climbing". But I think it might be a duplicate of Q918226. We could move the German link (Höhenbergsteigen) to Q12000000 because the other links in Q918226 are about the effects of high altitude climbing, although I can't confirm that for all languages. --Tobias1984 (talk) 10:35, 26 April 2013 (UTC)

Instance of Lua

Does it make any sense to add "instance of Lua" to random pages in Module-namespace? See here! Technically all pages in ns:828 is an instance of Lua, so I cannot see it makes any sense to add that. All pages in ns-0 is an "instance of article", all pages in ns-14 is an "instance of category" and all pages in ns-10 is an "instance of template", but I cannot see it used anywhere.

Instead, I proposed "Wikipedia:Lua" as a GND-type for ns-828, but I have been reverted this far. - Lavallen (block) 19:21, 17 April 2013 (UTC)

No, it doesn't, since Lua is a programming language, and I don't really see which things can be instances of a programming language (except maybe specific implementations or specifications). Anyway, the pages in the Module namespace are Scribunto modules, so the proper statement would be, I guess, "instance of: Scribunto module". So I created Q11382506 and corrected the "instance of: Lua" statements in the affected items. Silver hr (talk) 17:06, 18 April 2013 (UTC)
I agree with Silver hr on this one. An Lua module cannot be an instance of Lua.--Snaevar (talk) 18:18, 18 April 2013 (UTC)
I think it would make sense to have all templates marked as "instance of template" and all Lua modules marked as "instances of Lua module". Just create an item "Lua module" with the programming language property set to Lua. --Zolo (talk) 18:23, 18 April 2013 (UTC)
Well on a second thought, maybe not. Such items are just for interwiki hubs, they do not describe an external and well defined Lua module, they are just an interwiki hubs for various modules in individual Wikipedias. It may not be the best ideas to conflate data about an external entity and Wikimedian metadata. --Zolo (talk) 18:30, 18 April 2013 (UTC)
I personally have no preference, but maybe data about WM entities is useful to someone. In any case, my answer wasn't meant to be interpreted as pro or con WRT that issue, just as a correction of improper "instance of" use. Silver hr (talk) 23:23, 22 April 2013 (UTC)
  • Instances are tokens. In other words, instances are things that have a unique location in space and time. Instances can't be instances of other instances, because the domain of the 'instance of' property is 'instance' and the range is 'class' (see the P31 talk page and the X and Y columns for 'instance of' in Help:Basic membership properties). So if something can have instances, then it cannot itself be an instance.
Using 'instance of' with items about software is virtually always incorrect. This is quite common, but it ignores the difference between 'instance of' and 'subclass of'. For example, Java (Q251) is about Java as a concept, not a particular instance of Java. What's an instance of Java as it concerns Q251 (i.e. Java the programming language)? It's http://download.oracle.com/otn-pub/java/jdk/7u21-b11/jdk-7u21-windows-i586.exe, installed on a computer. When Oracle says Java is installed on 3 billion devices, it is those installations that are instances of Java. Clearly, the Wikidata item about Java the programming language concerns that subject as a class, and not as an instance.
So, similarly, it seems to me that the things in the module namespace are not instances, but classes, and should thus be classified with P279 rather than P31. P31 and P279 are fundamentally ontological membership properties. Because Wikidata exists to structure all knowledge, I think we should use those properties to classify the world as it actually is -- where the Wikidata items about things like 'Java' and Lua modules are indeed not about instances -- rather than changing the meaning of those properties where their proper usage might be unintuitive at first. Emw (talk) 01:00, 23 April 2013 (UTC)
Although the distinction is clear with physical things, to me it's somewhat less clear with non-physical things. For example, you mentioned that a particular installation of the Java Development Kit is an instance of the Java programming language. Since the JDK includes the Java Runtime Environment plus auxiliary tools for developing, debugging, and monitoring Java applications, why not the JRE? Why a particular installation as opposed to a particular running process (a single installation can be run as multiple processes)? Why a particular installation as opposed to the installer file? After all, an installation is just a copy of the (uncompressed) contents of the installer. And, since the JDK/JRE also contains code that doesn't have much to do with the programming language, but rather deals with the particulars of each of the different platforms Java is made for, why not say that a particular specification or reference implementation is an instance of the language? I don't mean to criticize you, I'm just trying to figure this out. Silver hr (talk) 21:07, 24 April 2013 (UTC)
This distinction can be unintuitive and tricky when applied to some types of non-physical objects, but I think we need to be consistent in how we we apply P31 and P279, and thus what we call an instance and a class. With regard to the question of programming languages, I think installations of, say, Java and computer processes that run that language can be considered instances of Java. Both the installation and the process have a particular location in time and space. The installer file is something that can exist at multiple locations at the same time, thus I don't think it the installer file can be called an instance unless by "installer file" one is referring to a particular copy of the file, at a unique location.
Some types of non-physical objects aren't so tricky. For example, calling square (Q164) a class is both intuitive and correct according to this distinction of classes as types and instances as tokens. (The idea that geometric figures are abstract and not concrete objects has a long history.) Clearly, 'square' as it concerns Q164 is not about some concrete thing with a particular location in space and time, so square (like all other shapes) is not an instance, and thus not a valid subject for 'instance of'. Emw (talk) 17:30, 27 April 2013 (UTC)

New design for RFP

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.

Implemented. — ΛΧΣ21 16:30, 27 April 2013 (UTC)

I have created a new design for the Wikidata:Requests for permissions‎ page. Now, the page looks very bad and unmanageable. Hence, I took a bit of my time and designed this: User:Hahc21/RFP. I would like to know if you think this new design is of your pleasure, and if the community supports its implementation. Cheers. — ΛΧΣ21 18:54, 21 April 2013 (UTC)

Do I understad this right if I think that you want to create an extra page like Wikidata:Requests for Permissions/Other rights? I would gladly support this. -- Bene* talk 18:58, 21 April 2013 (UTC)
+1; I also think that a redesign of RfP is really needed. Thank you for your work! Regards --Iste (D) 19:24, 21 April 2013 (UTC)
  Support Making an index as Hahc21 has allready done was excactly what was needed. Nicely done!--Snaevar (talk) 20:30, 21 April 2013 (UTC)
  Neutral My only concern is that this looks too much like a restaurant menu, which in my opinion would encourage hat collecting. --Jasper Deng (talk) 20:46, 21 April 2013 (UTC)
"Restaurant menu" xD. I understand your concerns, Jasper Deng, but I think that a cleanup is really needed. In my opinion we shouldn't hide the various rights in a hill of disorder. Regards, -- Bene* talk 21:19, 21 April 2013 (UTC)
In my personal opinion, the new design is a bit fussy. I would prefer a plainer solution. Regards, Vogone talk 21:31, 21 April 2013 (UTC)
  Weak support Cleaner, but a bit like a "Restaurant menu" xD --Ricordisamoa 21:55, 21 April 2013 (UTC)
  Support makes it so much tidier, only thing I would change is have a bot update a count of how many requests for permissions there are on each page so we can see without clicking? (I am happy to run the bot ;p) ·Add§hore· Talk To Me! 12:19, 22 April 2013 (UTC)
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

Deleting and moving properties/ Extern used propertys

How is the handling of propertys, which are used in Wikipedia? Users in german Wikipedia telling, that deletion here made trouble in Wikipedia-articles. Thank you for help, Conny (talk) 09:14, 26 April 2013 (UTC).

Can you give some examples? I don't think we delete some properties since Wednesday. Snipre (talk) 11:42, 26 April 2013 (UTC)
To all administrators and users: please be careful when deleting or moving properties from data objects, because they can be referenced in hundreds of language versions of wikipedia.--Sinuhe20 (talk) 21:22, 26 April 2013 (UTC)
See User talk:Zolo#Deletion of Properties --Chris.urs-o (talk) 08:07, 27 April 2013 (UTC)

Prioritizing bugs

Heya :) We're going to try something new and see if that'll work well or not: If fixing a particular bug is especially important for you then please consider voting for it in Bugzilla to help the development team prioritize. A list of all of the currently open ones is here. --Lydia Pintscher (WMDE) (talk) 13:37, 26 April 2013 (UTC)

Wow! 516 bugs... :)  Ę-oиė  >>>  ™ 14:21, 26 April 2013 (UTC)
In my opinion bug 46537 is one of the most important bugs because without summary we can not find abuses!Yamaha5 (talk) 18:27, 26 April 2013 (UTC)
  • Avoid to annoy new users: Bugzilla:47620 - add link widget should not invite to add links for user and talk pages
  • When adding statements, this might help: Bugzilla:45351 - improve sort order in entity selector
  • If it's still open: Bugzilla:46229 - Add a link on the client when a Wikidata item exists, even when no other interwiki is already defined
  • Help re-users of dumps in one way or the other, maybe Bugzilla:44844 - Wikidata specific tables in the dumps
  • Something is borked with property name look-up, maybe one of the following addresses it. More important now that the number of available properties with similar names increases.
    • Bugzilla:46363 - use terms table for property label lookups
    • Bugzilla:46788 - allow linking when property parser function has multiple results
  • Edit conflicts with oneself are annoying. Maybe this addresses it: Bugzilla:47461 - Changing a description and a property causes an edit conflict
  • We used to keep getting duplicate items due to this missing: Bugzilla:36729 - Update repo on page move
  • Wikidata seems invisible there: Bugzilla:46358 - Commons GlobalUsage should include media linked in statements
These are some I found when going through the recently edited bugs .. --  Docu  at 08:32, 27 April 2013 (UTC)
Before I forget: if there could be done something about replication lag of toolserver: it's at 9 days now and reports at Wikidata:Database reports are somewhat old. --  Docu  at 08:50, 27 April 2013 (UTC)
That was quick. Thanks. --  Docu  at 15:23, 27 April 2013 (UTC)

Annoying bug

Hi everybody. I'm trying to add the disambig page Reaper to Q409784, but the Wikidata insists to add REAPER, an older page in the Wikipedia.pt db. Besides, every time we reorganize article's positions by moving them under Wikipedia.pt, the automatic work done by the bots doesn't work anymore (in Wikidata). Finnaly, explain to a novice "how to" add an interwiki nowadays is much more complicate than the older way... So, not pretty happy here, lol... Well, that's it for now. Sorry about my poor English and hope Wikidata can deal with all this. Best wishes from Brazil, Sturm (talk) 23:14, 26 April 2013 (UTC)

RfD

I made some RfDs: Q3966764, Q3966551, Q3792034
Did I get it right? Was it all wrong?
Thx --Chris.urs-o (talk) 07:41, 27 April 2013 (UTC)
The three items have labels, descriptions and aliases in Italian. All of these should also be moved if merged. Byrial (talk) 07:59, 27 April 2013 (UTC)
Besides, the request for deletion is not to be placed in the English language description, but at WD:RFD. Byrial (talk) 08:02, 27 April 2013 (UTC)
Ohh, sorry. How do I see the labels, descriptions and aliases in Italian? --Chris.urs-o (talk) 08:08, 27 April 2013 (UTC)
I have deleted. To see label and description in other language you can use gadget LabelLister. --ValterVB (talk) 08:30, 27 April 2013 (UTC)
Thx --Chris.urs-o (talk) 08:45, 27 April 2013 (UTC)

Featured article

Is there a way to show that an article of a category is featured for example in english wikipedia, that can be mark in all others language as featured article of english wikipedia? Xaris333 (talk) 13:48, 27 April 2013 (UTC)

Q242864

Q242864 does not appear to be working on the English version, Bahraini uprising (2011–present). CambridgeBayWeather (talk) 16:19, 27 April 2013 (UTC)

What do you mean? Could you clarify please? --Stryn (talk) 16:23, 27 April 2013 (UTC)
I can't see any links to other language Wikipedia articles from the English version. CambridgeBayWeather (talk) 16:31, 27 April 2013 (UTC)
Well, I can see all links. Have you purged your cache on the Wikipedia article? --Stryn (talk) 17:11, 27 April 2013 (UTC)
I tried that with no luck. I tried another browser an it was fine there. Now of course purging has worked. Thanks. CambridgeBayWeather (talk) 17:55, 27 April 2013 (UTC)

Strange: I can see the Label

In my Sandbox User:ValterVB/Sandbox/IMDb/Film I have a list of Item [[Qxxxx]], but I don't see the number, I see label +q number and if I change language, label also change. New feature? --ValterVB (talk) 16:44, 27 April 2013 (UTC)

You have been stareing at your screen for too long ;)
It hasn't happened yet .. or one of your gadgets solves it. --  Docu  at 16:52, 27 April 2013 (UTC)
: I have deactivate and activate AuthorityControl gadget (but I don't know if is relevant) and now I can see only Qxxxx --ValterVB (talk) 16:56, 27 April 2013 (UTC)
Actually, now I recall that I did look at that page before and yes: they were there. --  Docu  at 17:11, 27 April 2013 (UTC)

Moeder de Gans

Er is iets mis met de pagina's Q5104112 en Q11631007. De laatste heb ikzelf aangemaakt met een label en een beschrijving zoals gevraagd. Maar je hoort blijkbaar te wachten op de bot alvorens alle interwiki's in nl:Sprookjes van Moeder de Gans verdwenen zijn. Manueel blijk ik niks te kunnen toevoegen aan Q11631007, omdat bijv. de Franse en Engelse pendant al gebruikt zijn in Q5104112... maar die pagina heeft geen label en geen beschrijving. Kunnen die worden samengevoegd? Ik ben hier nieuw, dus ik snap er nog niet veel van. --Dartelaar (talk) 01:55, 22 April 2013 (UTC)

I don't know your language but judging from Google translation I can assume you need to merge the two items. --Michgrig (talk) 03:24, 22 April 2013 (UTC)
See Wikidata:De_kroeg#Moeder_de_Gans. Ljubinka (discussion) 11:28, 22 April 2013 (UTC)
Moved the item. --Joostik (talk) 10:40, 28 April 2013 (UTC)

Qualifier bug?

When I was trying to add the qualifier role to several cast members listed under the property of starring, I was forced to refresh every time I save a change to one value of the property. Otherwise, it would give me an edit conflict pop-up. After refreshing the page, I found the item (cast member) that I just edited to be at the bottom of the list of values under the 'starring' property. Is this a bug? --Wylve (talk) 19:10, 26 April 2013 (UTC)

bugzilla:46722 --β16 - (talk) 22:02, 26 April 2013 (UTC)
false edit conflicts will be fixed in wmf3 (next version). the reordering of claims after modifying is another issue and will most likely not be solved in wmf3. -- Tobias Gritschacher (WMDE) (talk) 12:09, 28 April 2013 (UTC)

check usage in Wikipedia

Tracked in Bugzilla
Bug 47288

Hey,
how to check the usage of Wikidatacontent in Wikipedia (via Zolo). Thank you, Conny (talk) 19:56, 27 April 2013 (UTC).

I can imagine that the use of #property: can easily be tracked by somebody who can create a nice toolserver-tool, but I cannot imagine how such a tool should be created to track every Lua-usage. It looks more or less impossible, but I have underestimated our toolserver-gnomes before. Do we have any toolserver-gnome here? Byrial? -- Lavallen (block) 20:25, 27 April 2013 (UTC)
bugzilla:47288 Legoktm (talk) 20:59, 27 April 2013 (UTC)
Me? I made a few toolserver tools some years ago, but never anything significant. My speciality was analysis of database dumps. But anyway I think that it is impossible for now. The property usage isn't recorded in any database tables available on the toolserver. Not even the statements themselves are. The best you can do, is to track which articles use which templates and which Lua modules, but not how they are used. To really see which properties are used, you have to extract and analyse the data from full databasedumps for each Wikipedia, which is a way too big task for me. But I may be able to do some kind of analysises. You are welcome to ask if have ideas, and I will see what I can do. Byrial (talk) 22:02, 27 April 2013 (UTC)
Det är problemet med att vara duktig, folk minns det, och vill utnyttja de förmågorna igen! :) -- Lavallen (block) 07:18, 28 April 2013 (UTC)

Possible bug: link to label, not to sitelink

Hi, I am not sure, but may have discovered a possible bug for using properties in phase 2. The case: I created a template for the nds-nl wiki, nds-nl:Mal:Wikidata (Mal is template in nds-nl) for retrieving wikidata properties, to be used from an infobox. Then I modified a template nds-nl:Mal:Gemainte to use this template for retrieving the province for municipality. So far everything was ok. Then I modified the page nds-nl:Vlagtwedde_(gemainte). But the result was incorrect: the link was to nds-nl:Grunnen (disambiguation page), not to nds-nl:Grunnen (provìnzie) as it should have been. I did not understand this, as the Q752 item did link to the nds-nl:Grunnen (provìnzie) page. As an experiment, I modified the nds-nl label in Q752 to Grunnen (provìnzie), and then the link on Vlagtwedde was OK. So it seems the retrieval of the property did get the label, not the sitelink. That may be ok in running text, but not in links. I am not sure how to resolve it, as adding the resolve of the disambiguation in the label (with brackets), may not be preferred. Sequence to reproduce:Change the label on Q752 back to Grunnen (as it should be in my opinion, the description should mention the fact it is a provence,not the label, and check the nds-nl:Vlagtwedde_(gemainte) page, and click the Grunnen link. It will go to the disamb page, not the provence page Droadnaegel (talk) 22:54, 27 April 2013 (UTC)

It is not a bug. The #property parser function returns the label. To get the link, you need to use a Lua module in the module namespace. See documentation at mediawikiwiki:Extension:WikibaseClient/Lua. Byrial (talk) 07:59, 28 April 2013 (UTC)
You can copy Module:Wikidata at test2.wikipedia.org. -- Lavallen (block) 08:33, 28 April 2013 (UTC)
Or Module:PropertyLink. - Soulkeeper (talk) 12:20, 28 April 2013 (UTC)
Ok, it's clear now. Thanks for the pointers, I'll give it a try. Droadnaegel (talk) 18:20, 28 April 2013 (UTC)

International Talk pages concerning wikidata

Pls. enable wikidata for the page Wikidata:Project chat crosslinking to http://www.wikidata.org/wiki/Q12136816

Regards --Gruß Tom (talk) 12:57, 28 April 2013 (UTC)

This is covered by Q4847210 from what I can see of the links you made. In addition, that item you linked is out of scope of Wikidata per WD:N. I deleted the new item. --Izno (talk) 13:28, 28 April 2013 (UTC)

Unexpected value for entity type

False positives. Special:AbuseFilter/14 should be updated to be consistent with Property:P107. --Makecat 13:32, 28 April 2013 (UTC)

  Done --Stevenliuyi (talk) 13:40, 28 April 2013 (UTC)

MediaWiki:Gadget-markAdmins.js

For technical reasons, Jon Harald Søby is encoded as "Jon_Harald_S%C3%B8by". What would happen if a user were to create an account with the username "Jon_Harald_S%C3%B8by"? Would they be marked as a steward? FrigidNinja 23:33, 28 April 2013 (UTC)

"%" in user names are invalid. So no account creation would be possible. Regards, Vogone talk 00:00, 29 April 2013 (UTC)

Translation notifications

I'd like to start a notification service for translators here on Wikidata. Could every interested translator sign up at Special:TranslatorSignup, please? Regards, Vogone talk 00:04, 29 April 2013 (UTC)

Wikidata:Requests_for_comment/Items_for_Wikimedia_projects_besides_Wikipedia

Comment away! ...Jakob Megaphone, Telescope 00:55, 29 April 2013 (UTC)


Usability Testing on wikidata.org

Hi there! I am Lukas Benedix, a student of computer science at the Freie Universität Berlin in Germany. In cooperation with the Wikidata developers I’m currently working on my bachelor thesis about usability testing in open source software projects and I’d like to provide the Wikidata community my developed feedback mechanisms (only as a test). Wikidata is a very active, emerging project which is why I think it’s a great platform for my project. For further information about my project, to read how you can help to improve Wikidata please visit... http://www.wikidata.org/wiki/User:Lbenedix/UIFeedback Thank you!

Different years of Constantinus I prohibition of crucifixion in Roman empire

There are two defferent years pibliced on swedish Wikipedia - 315 and 321, it should be one year + http://sv.wikipedia.org/wiki/315 Korsfästning avskaffas som straff inom Romarriket. + http://sv.wikipedia.org/wiki/Kristendom_i_Romerska_riket År 321 blev söndagen arbetsfri dag och korsfästelsen förbjöds som straff

Wikidata : both a Wiki and a Database. What does it imply ?

  1. Wikidata is a Wiki, which means it can change at any moment
  2. Wikidata is a database, which means it will be used to automate things.


Wikidata is also a very flexible tool, in which used ontologies are decided by community an by #1 might change.

This raises the question "How do we handle changes".

  • how do we interract with programmers ? If the structure of datas changes too much, programmers could be pissed of an turn away from the project. Should we handle Wikidata structural changes as an API ? A public API ? Any change in wikidata might need the update a program, which means we will communicate with bot programmers, Infoboxes coders an so on when something they uses in their program changes.

Should we publish a stable API ? How would we handle changes on this public API, property deprecation, time based update ? Just ideas I through as they come, but we should discuss that.

Another way ti see things : Should we recommend that structural choices on wikidata be made with some kind of robustness property in mind ? A structure is robust if when we change something in it we do not have to change everything else ...

3 Items for Linux

There are currently three items for Linux:

The difference of Q14579 to the others is clear. But what should be the difference between Q388 and Q3251801? -- MichaelSchoenitzer (talk) 19:44, 21 April 2013 (UTC)

  • Q14579 is the kernel;
  • Q388 is a family of Unix-like OSs, based on the Linux kernel;
  • Q3251801 is a subset of Q388: a Unix-like OS, based on the Linux kernel and GNU programs, available in several distributions.
--Ricordisamoa 20:16, 21 April 2013 (UTC)
Afaik, there are no OS based on the Linux Kernel and not based on Gnu. -- MichaelSchoenitzer (talk) 21:59, 28 April 2013 (UTC)
If I am not mistaken, this edit changed item Q388's meaning. But sitelinks were not (re-)moved accordingly. For example w:en:Linux is still associated with Q388 and the article clearly states, that it is about an operating system with Linux kernel and GNU userspace tools. What is the right strategy to clean this up? I would say it is not allowed to make changes to a description in a way that afterwards the description no longer matches the subjects of sitelinked Wikipedia-articles. -- Make (talk) 12:10, 29 April 2013 (UTC)
Which is already the case, isn't it? That is, you are not allowed / supposed to make changes to Wikidata labels and descriptions that lead to them no longer matching the linked articles. However, beyond being a procedural guideline or policy, I can't see anyway of enforcing this within the software. QuiteUnusual (talk) 12:15, 29 April 2013 (UTC)
The articles are sometimes edited, and there is often no consensus on whether an article includes GNU-less systems. The English article talks about Android without mentioning that it is not GNU/Linux. Wikidata is not just for interwikis. --AVRS (talk) 12:39, 29 April 2013 (UTC)
@AVRS: I think I understand your rationale. However: this leads to a fragmentation of sitelink clusters, cutting interwiki connections of Wikipedia articles or even isolating articles. Management of sitelinks was transferred to Wikidata with the (implicit) promise that management of sitelinks will get easier, not create additional problems. My opinion is that Wikidata has to deliver on that promise. Otherwise the Wikipedia communities will get upset and support for Wikidata will dissolve (and rightly so). – I kindly ask you to reconsider. -- Make (talk) 15:09, 29 April 2013 (UTC)
Let's wait for the redirect linking support. Wikidata cannot solve all interwiki conflicts anyway. --AVRS (talk) 18:38, 29 April 2013 (UTC)
Android is based on Linux without GNU. --AVRS (talk) 12:39, 29 April 2013 (UTC)

Property:P17

Sovereign state of this item, what about Sahrawi Arab Democratic Republic, South Ossetia, Northern Cyprus and Taiwan? Are they neutral point of view? Wikidata:互助客棧#关于Property:P17的中立性问题--GZWDer (talk) 05:31, 23 April 2013 (UTC)

In my opinion the best solution would be to have two instances of Property:P17 (taking up the Taiwan example, one claiming the item is in Taiwan and one claiming it is in PRC) and add a qualifier to both of them that states that the property's value is disputed, who claims this particular version of the property and perhaps a reference to the item of the Wikipedia article about the underlying dispute (in the case of Taiwan maybe Q706243).
That should fulfil the NPOV. --Slomox (talk) 07:06, 23 April 2013 (UTC)
Using two P17 is not that easy. Though the real political question is indeed whether Taiwan is independent or not, the question is still put differently in legal texts (as user:Stevenliuyi reminded in Wikidata talk:List of properties#Sovereign_state_or...). The official legal question is: is People's Republic of China or the Republic of China (Taiwan) the lawful government of China ? If we take this question seriously, we shouuld both add a "P17: PRC" for things in Taiwan and "P17: ROC" for things in the mainland. Though that me be accurate, I doubt that it will solve POV conflicts.
Actually, P17 essentially means "the highest level in the P:P131 chain". That makes the property logically redundant. If it raises too many concerns I would simply suggest not to use it, so that we can concentrate the controversial claims in places wehere it is most relevant, like the item about the Republic of China as a whole. For places like Western Sahara, where the situtation is not so clear, we need to handle things at a lower level. --Zolo (talk) 07:37, 23 April 2013 (UTC)
Can not we solve it with a qualifier that the state is not recognized?--Ymblanter (talk) 07:56, 23 April 2013 (UTC)
Yes, I think we could also use "P17: Republic of China (state not recognized by the United Nations)" for things in Taiwan. After all it is fairly clear that Taiwan is usually not officially recognized as an indenpendent state, and that it is de facto a sovereign state). But I do not think that we should add a second value "P17: People's Republic of China" to the items then. --Zolo (talk) 08:04, 23 April 2013 (UTC)
I think it is uncontroversiell to use items like Q4628 in P17. So the problem isn't (only) that the country isn't recognized, it's that users has strong opinions in the subject. -- Lavallen (block) 13:10, 23 April 2013 (UTC)
@Zolo: Handling these on a case-by-case basis kinda seems POVy to me. But perhaps I don't understand you correctly. Maybe you could try to describe how you would handle the case South Ossetia? --Slomox (talk) 07:31, 24 April 2013 (UTC)
For Taiwan, we are more or less forced to use the administrative framework defined be the Republic of China (Taiwan) in P:P131, because really it is the only one that exists. That is how it seems to be done by all Wikipedias. However I agree that it is sort of POV to say "country: Republic of China", and that's not very useful either as the country can be inferred from P131. That's why I would suggest to simply not use P17 here. If that seems like ad hoc tinkering, we could probably delete P17. Actually, it does not seem to be a structurally sound property (it is used in two completely different ways, as a subset of P131, where it is logically superfluous, and to mean something like "domain of validity", where a better defined property would be better).
I do not know well South Ossetia, but the major difference is that the de facto situation is much more complex. We probably need three values "part of Georgia", "independent state" and "occupied by Russia", but adding the three of them to every single mountain in South Ossetia seems unnecessary, and hard to maintain properly, so I would say that these three values should only be given for "South Ossetia" as a whole. There may be some parts of South Ossetia where we need to add more specific qualifiers, but otherwise, I do not think is is necessary to copy those three values. --Zolo (talk) 08:29, 24 April 2013 (UTC)

I'm not a lawyer and may law skills are limited. In my view, we should label items as sovereign states if they exist as sovereign. The Republic of China is sovereign in the sense that they have actual state organizations that rule the country. As far as I know, the Sahrawis don't actually rule their region but Morocco, so it should not described as sovereign. --NaBUru38 (talk) 16:51, 29 April 2013 (UTC)

Request limitations on wikidata API ?

Is there anylimitation on http request to wikidata's API ? I have a script doing fine when doing 2 queries/secs, but when I do 5queries/secs, it stale after ~25queries or 5secs. Any idea ? (also on Stackoverflow) Yug (talk) 18:49, 23 April 2013 (UTC)

Check here. Good luck! --NaBUru38 (talk) 16:53, 29 April 2013 (UTC)

Pope Francis the chemist

According to Q450675, the occupation of Pope Francis is 'chemist'. While it is true that Jorge Mario Bergoglio worked as a chemist for a couple years before joining the priesthood, it's a bit confusing to list his only occupation as 'chemist'. Should we create a property for 'former occupation'? Should 'occupation' be changed to 'current occupation' or is that too difficult to keep up to date? Perhaps this is something that would be better handled with Qualifiers, although I'm still not completely sure what those are or how to use them. Kaldari (talk) 06:15, 24 April 2013 (UTC)

Yes, it should be done with qualifiers. They are just statement added to a statement (see Q4115189), but we do not have yet the date datatype. -Zolo (talk) 06:47, 24 April 2013 (UTC)
I don't think he should be listed as a chemist. There should be an implicit understanding that the occupation property is not meant to include every job the person has ever held. Pope Francis also worked as a bouncer in a nightclub. Are we going to add this too? Many people had summer jobs flipping burgers, we shouldn't add that. J.K. Rowling's occupation shouldn't be listed as secretary. This is noise, not data. Pichpich (talk) 12:06, 24 April 2013 (UTC)
That is data. The question is whether it's relevant data. —★PοωερZtalk 12:11, 24 April 2013 (UTC)
That's what noise is: irrelevant data that obscures the signal. Pichpich (talk) 12:41, 24 April 2013 (UTC)
And irrelevant data is still data, your argument is invalid. I didn't want to point out anything beyond that. —★PοωερZtalk 13:12, 24 April 2013 (UTC)
If you want to defend the idea that Wikidata should be flooded with irrelevant data, be my guest but you're not going to get much support. Discriminating noise versus data is a core objective and your "all data is equally good data" is the quickest way to turn Wikidata into a useless mess. Pichpich (talk) 14:13, 24 April 2013 (UTC)
I neither said nor indicated "all data is equally good data"; don't put words in my mouth. —★PοωερZtalk 15:51, 24 April 2013 (UTC)
I am not sure that is is relevant, because someone who needs information about the pope should probably look at a text rather than a database, but I am pretty sure more people are interested in this than in the birthplace of a little-known person, or many other things that are routinely added to database items and encyclopedic articles. --Zolo (talk) 13:22, 24 April 2013 (UTC)
I think that with the right qualifiers more important data can be easily emphasized from "noise". We should bear in mind that one person's noise is another person's signal (e.g. Typos are mistakes "noise", but some linguist study only typos because they reveal how language changes). To me a query like "previous jobs held by popes" doesn't sound too outlandish. --Tobias1984 (talk) 14:24, 24 April 2013 (UTC)
But clearly, we shouldn't litter Wikipedia with typos because they're a fun thing to study for linguists. The problem is that to make room for the query "previous jobs held by popes", we need to accept the fact that the query "20th century chemists" will be littered with popes (ok, littered with one Pope) and other people who briefly had a job otherwise meaningless in their lives. It's a bad trade-off and only a small minority of Wikipedias have chosen to categorize Pope Francis as a chemist. Of course we might be able to solve this with qualifiers but we don't have qualifiers right now, we don't know when they'll be available and we don't know how easy to use they will be. Pichpich (talk) 15:42, 24 April 2013 (UTC)
Oh, really? Are you sure about your claim that qualifiers aren't available? :) --Izno (talk) 21:11, 24 April 2013 (UTC)
Qualifiers are available, but not yet as date related properties, so no qualifier for the pope former job can be added here. HenkvD (talk) 21:34, 24 April 2013 (UTC)
Ideally we could find some qualifiers that prevent him from showing up in a "20th century chemists" query. Chemistry in my opinion can be a job (Qualifier = "worked as"), or as a science. For somebody to be a science chemist he would have to have at least one peer reviewed article or book published. We could have qualifiers like "mayor contributions to the field" and "some contributions to the field". --Tobias1984 (talk) 21:48, 24 April 2013 (UTC)
Is there a "former" qualifier? Where do I find the existing qualifiers that are available? Kaldari (talk) 06:13, 25 April 2013 (UTC)
Qualifiers are just normal properties, and can be found in WD:List of properties. "former" should probably be a value for a yet-to-be-created property, or perhaps the statement should just be marked as deprecated, but I do not think it is possible yet. --Zolo (talk) 07:01, 25 April 2013 (UTC)
This would be done using the "from time" and "to time" qualifiers. However, we don't have the date/time data type yet. (Note that w:Pope Francis is in fact categorized in "Argentine chemists".) --Yair rand (talk) 17:20, 29 April 2013 (UTC)

I agree with Pichpich: we should only include relevant data. Lots of actors have had regular jobs. It shouldn't be our job to list them all. --NaBUru38 (talk) 16:56, 29 April 2013 (UTC)

Why not? How is this not "relevant" data? --Yair rand (talk) 17:22, 29 April 2013 (UTC)

not to make No-interwiki pages

If we check conflicts most of them are caused by single language Items and it makes some difficulty to solving them .In my opinion we should not to make None-Interwiki pages especially after New JS tool which is deployed with second phase at yesterday. if the page has Item in wikidata JS tool doesn't work! please see:

Most of these errors are caused by single lang items! Yamaha5 (talk) 18:55, 25 April 2013 (UTC)

There are many legitimate items that only exist in one language, though, so I'm not sure what you're suggesting we do about those? These lists seem like they need human intervention to fix, and are good to-do lists for those who read the languages a little (for instance, I can work on es: & it:) -- Phoebe (talk) 19:11, 25 April 2013 (UTC)
How we can say that Items are corrects? some of them are duplicate and need to merge and who can merge these amount of single items? Now we have many single language Items which we should maintenance them for correcting move or deleting the article and ...
In my opinion we can Stop making single language Items and leave which are made till now and for future only make items for page which has interwiki.Yamaha5 (talk) 19:28, 25 April 2013 (UTC)
This shoudl be useful if wikidata are still for interwiki links only. But for infoboxes is often need some items witout interwiki.
And it would be weird, if there are all cs.wiki articles from ABCDE, Q and VWXYZ and after Z, but only some from range F-U :-) JAn Dudík (talk) 20:01, 25 April 2013 (UTC)
I'm going to try posting each one of these pages on their respective Project Chats. FallingGravity (talk) 20:53, 25 April 2013 (UTC)
I've come across this too. Instead of creating an item for single article on some wiki, why not have the bots list those articles somewhere, preferably on the local wiki? Or maybe a special page called "Pages with no wikidata association" We have editors who never bothered with interwiki, and of course the newbies doesn't even know what is interwiki/wikidata. Aurora (talk) 02:34, 26 April 2013 (UTC)

2 topics handled in one page

I was working some of these notices above for nl-wiki. For example:

The dutch wiki has one article about both outlook.com and hotmail. In my opinion it's good that foreign articles, for example the english articles on both Hotmail and Outlook.com point to the dutch article on Outlook.com (as it handles both topics) but that does not seem to work in the wikidata system (because it does not know which one to pick for the outgoing links on nl-wiki then?). What is the idea on these conflicts? Basvb (talk) 23:36, 25 April 2013 (UTC)

You can use local interlanguage links to solve this problem. Coincidentally we have the same problem/discussion about this kind of problem here. Probably it's not a coincidence rather than a common problem which needs to be solved. --Knopfkind (talk) 00:41, 26 April 2013 (UTC)
It does seem to work. The redirect page (nl:Hotmail) is in Q3903488 (even though I wonder how that is possible as the page do not use the magic word __STATICREDIRECT__ which I thought was required), so foreign articles on both items link – directly or indirectly – to nl:Outlook.com. Byrial (talk) 07:49, 26 April 2013 (UTC)
Ok, so just leave it as it is now in the current situation. Where one links to the redirect, and one to the page itself. Only problem with that is the project in this topic. All those links to redirects and main article are being looked into, and knowning these kinds of project there will be some day that somebody things, hmm 2 links to one article "click" deleted. But that's a future issue we can't do much about. Basvb (talk) 09:38, 26 April 2013 (UTC)
This has been referred to as the "Bonnie and Clyde problem". Many Wikipedias have one page for the US outlaws Bonnie and Clyde but Wikidata can only have the properties of one item (or person) on a wikidata page. The solution was to get the developers to permit links to Redirect pages for Bonnie Parker and for Clyde Darrow where these redirect to the Bonnie and Clyde page. This means Wikidata can have pages for each of Bonnie Parker and Clyde Darrow (with properties related to each individually) as well as the Bonnie and Clyde Wikidata page (with properties related to this) Hope this helps. Filceolaire (talk) 18:10, 29 April 2013 (UTC)

Sources and changes in time

I do not know exactly how the source-part of our claims will look like, but I guess that, for exampel "publisher", we will use an item, which help us to automaticly make a link and translate from one language to another.

I have been digging deep into the archives of Statistics Sweden and I maybe found a problem. Neither the Swedish name (Statistiska centralbyrån) or the English name (Statistics Sweden) where the same 1910 as it is today. In 1950 the English name was "THE CENTRAL BUREAU OF STATISTICS", in 1910 the Swedish name was "KUNGL. STATISTISKA CENTRALBYRÅN". I cannot see any "international" name from the documents from 1910, but if there is any, it is possibly in French, not in English.

I guess this is also a problem in many other cases. A scientist who has written a paper has changed his/her last name as times goes by. The item will have his/her last known name, not the name from the time when the paper was written.

Are we aware of such problems, and how do we solve them? -- Lavallen (block) 06:28, 27 April 2013 (UTC)

The plan, as I understand it, is that qualifiers would be used to identify when a particular name applied. In practice the Statistics Sweden item would have an Property:Official_Name entry for each official name with a qualifier to identify during which years that name applied. The various source references would all link back to item Q1472511 with it's various names and aliases and official names. Filceolaire (talk) 18:24, 29 April 2013 (UTC)

Template's interwiki

Hi, I found many items of templates that their interwiki is connected to sub-pages like (/doc /document /Doc ,...) instead of main template page. we should solve them by getting query from local wikis I got it for en.wiki and fa.wiki. please help to solve them. some of them has item conflict because they have single link item for main template and they should be merge. 

SELECT page_title FROM page join langlinks ON page_id = ll_from WHERE page_namespace = 10 AND page_is_redirect = 0 AND (page_title LIKE "%/doc%" or page_title LIKE "%/Doc%") GROUP BY page_title ORDER BY count(ll_from) DESC;

This query is not precise because these sub-pages have many names! and so I got any pages which has /doc.. or /Doc.. in their title and interwiki and namspace=10 also it shows any pages which has interwiki (old type or wikidata)

So far I closed the RfD as not done since you removed the templates from the item and did not add them anywhere.--Ymblanter (talk) 11:53, 29 April 2013 (UTC)
why you revert Q7665579? is it True to have /doc page's interwiki when we have the Q5836688 ?
at first I moved them to new item if they had error I removed /doc interwikis Yamaha5 (talk) 13:32, 29 April 2013 (UTC)
If there is consensus to not include these templates I will revert back and delete the items. But for the time being, I do not quite see why the documentation templates should not be included.--Ymblanter (talk) 13:40, 29 April 2013 (UTC)
Documentation pages are not independent pages also two month ago here I mentioned we should exclude sub-pages like /doc or /archive I thought till now it is reasonable! Yamaha5 (talk) 13:58, 29 April 2013 (UTC)
Ymblanter: See WD:N. Subpages of templates are not to be included, which includes /doc and /sandbox pages. I'm not sure about the case where we have an actual subpage but where the template is actually used in a substantial fashion (e.g. en:Template:Infobox animanga/Header). --Izno (talk) 14:04, 29 April 2013 (UTC)
Yes, indeed, it is on WD:N though I can not find it discussed in any of the two RFC's. I will delete the item now.--Ymblanter (talk) 15:27, 29 April 2013 (UTC)
I would ping Zolo about it; he references a project chat topic about it in his overhaul. --Izno (talk) 15:34, 29 April 2013 (UTC)

Cat Scan

Is there a way i can find the articles of a category in english wikipedia that have an article in italian wikipedia? Xaris333 (talk) 16:00, 29 April 2013 (UTC)

I don't know if there's a specific ToolServer program for this... maybe you can do a JavaScript query or use Pywikipediabot... --Ricordisamoa 16:19, 29 April 2013 (UTC)
You can import en:User:Ebraminio/ArticleTranslator.js to your vector.js or common.js in en.wikipedia and go to that category and click on the translation if it translate any page to Italian by sure it has interwiki if not it doesn't have.The tool's documentation is at here Yamaha5 (talk) 18:13, 29 April 2013 (UTC)

Thx!!Xaris333 (talk)

Lua / label of property

To build a template, is there a way to display the label of property based on its number ? e.g. "17" => "country". --  Docu  at 20:09, 29 April 2013 (UTC)

There's a test taking place on Module:Label... --Ricordisamoa 20:33, 29 April 2013 (UTC)

[en] Change to wiki account system and account renaming

Some accounts will soon be renamed due to a technical change that the developer team at Wikimedia are making. More details on Meta.

(Distributed via global message delivery 04:18, 30 April 2013 (UTC). Wrong page? Correct it here.)

Genus vs species

genus of Ansellia africana is Ansellia (deleted). en:Ansellia africana redirected to en:Ansellia.

  • de, fr, it, lt:Ansellia africana
  • other:Ansellia

There're many cases similar to it.--GZWDer (talk) 10:49, 24 April 2013 (UTC)

Well, Q9385262 obviously should not have been deleted. May I suggest that Special:WhatLinksHere/Q9385262 is routinely checked before deleting items to see if they are used in statements. Byrial (talk) 12:48, 24 April 2013 (UTC)
  undeleted. --Ricordisamoa 13:27, 24 April 2013 (UTC)
I started to move links with the genus name only to Q9385262, but reconsidered and undid. As the genus is monotypic (i.e. having only one species), all articles are in reality about the same thing. I suppose the best thing to do, would be to distribute the article links in the items according to the title of the articles, and add links to static redirects where there is no article (that requires that you edit all the redirects, if the Wikipedias will accept that). Byrial (talk) 14:02, 24 April 2013 (UTC)
It's odd if Q20848 (Ansellia africana) contains links to lemmata that actually match the item Q9385262 (Ansellia).
The best solution for this inconsistency would be if all Wikipedias would agree on a common lemma form for monotypic genera.
What would happen, if a new species in a formerly monotypic genus would be discovered? A new article would be written for the new species, the wording of the existing article for the known species would be edited but the article would otherwise stay the same and a new article would be created for the genus describing the common aspects of both species. So the Wikipedias with binomic lemmata (Ansellia africana) seem to be correct because they would be set for this case, while the Wikipedias with genus lemmata (Ansellia) would have to move the existing article to a new lemma.
I would advise that all Wikipedias should use binomic lemmata for monotypic genera. --Slomox (talk) 14:32, 24 April 2013 (UTC)
Fully agreed. As a start, Wikidata could establish this as an internal guideline, and allow linking to non-compliant Wikipedias, i.e. articles with genus names. In the case of a split, the IW links will still need to be splitted, but this solution does not require all WP projects to agree immediately (which seems as an unrealistic hope to me). - Soulkeeper (talk) 19:13, 24 April 2013 (UTC)
hewiki now uses taxonbox that uses Wikidata (example). For monolytic species we assume that an entity for the genus could exist in wikidata, even if it doesn't have sitelinks (for example Q30847 - species, and Q10854011 - genus with no sitelinks). ערן (talk) 21:24, 25 April 2013 (UTC)
Looks very cool! Is there any chance that you could post a version of the templates and modules involved with English texts so I can translate them again and use it on nds.wp? I would love to have such taxonboxes on nds.wp but there's no chance I'll understand the template and module with Hebrew texts. --Slomox (talk) 15:13, 26 April 2013 (UTC)
I've created a modified version in enwiki (the English version support less taxon levels - but it is easy to expand): see en:Module:Taxobox and you should also import en:Module:PropertyLink. Some "documentation" can be found in en:Template_talk:Taxobox#Taxobox_2013. Eran (talk) 20:08, 26 April 2013 (UTC)
Now we have articles like in the portuguese or spanish wikipedias that are labeled "Ansellia" and clearly state "Ansellia is a genus ..." in the first sentence of the articles. But here in wikidata the links are stored in Q20848 (Ansellia africana), the page for the species, with statements like "taxon rank = species". For those wikipedias that decided to write an article about the genus, the statements from wikidata are now useless. --Dietzel (talk) 13:13, 27 April 2013 (UTC)
I think taking Slomox' advice is the only way to go: all Wikipedias should use binomic lemmata for monotypic genera. I'll actually add a little to it: All Wikipedias should use binomic lemmata for all monotypic taxa. It is i simple rule, it gives us orderly data and articles. It works great (speaking from experience with nowiki), except where it gives issues with interwiki linking to Wikipedias where it is not used. So the solution of not not using it seems pretty obvious to me. Then there's of course the work of converting intro sentences in existing articles. - Soulkeeper (talk) 10:14, 30 April 2013 (UTC)

German (Sie-Form)

Sorry, but this is really a bad joke. I can't take this serious. There is no reason to have a "German (Sie-Form)" as language. Tell me one example where this makes sense. And even so, this would be in just 0.00001% of all cases and for the 99.99999% rest we have same data as in German duplicated. It's only a waste of storage. Who came up with this idea? What's next? "English (simple)" for Wikidata?

Someone here with the same opinion? --Nightwish62 (talk) 18:11, 27 April 2013 (UTC)

yes --Akkakk (talk) 18:35, 27 April 2013 (UTC)
Again; we have a fallback de-formal -> de. If you set your language to de-formal you get the same data as for de. Regards, Vogone talk 18:37, 27 April 2013 (UTC)
This changes nothing to the situation it is completely unnecessary and consumes storage filled up with the same information as the German version. --Nightwish62 (talk) 18:47, 27 April 2013 (UTC)
Nightwish62, why rant about it? If some people want to use that form of German, just respect that and let them have it. Only very few translation units for our translated pages have separate translations for de-formal (statistics here). It shouldn't bother you as long you can chose your preferred form of German. Byrial (talk) 19:04, 27 April 2013 (UTC)
I'm not talking about the help or community sites, but the possibility to edit items itself in "German (Sie-Form)". It's okay for me for the rest, the portal and so on. --Nightwish62 (talk) 19:22, 27 April 2013 (UTC)

By the way, de-formal is not wikidatawiki only. All wikis have this interface language, including dewiki itself. I don't see why wikidatawiki should be an exception there. Regards, Vogone talk 19:08, 27 April 2013 (UTC)

The fact that no one can give me answer to my question and tell me an example is prove enough it's unnecessary. Tell me one item with different label or description. --Nightwish62 (talk) 19:18, 27 April 2013 (UTC)
Please read what I've written above. We have a fallback. Regards, Vogone talk 19:43, 27 April 2013 (UTC)
This Question is coming up quite often it doe not make sense to have different labels for en-gb, en-us, de-formal in the main namespace. If we can't deactivate it for the main namespace we could make an community decision not to populate the labels for interface languages that have no corresponding wikipedias. It adds complexity without any benefit because we have the fallback. --Saehrimnir (talk) 00:51, 28 April 2013 (UTC)
en-gb is indeed sometimes a bit different from en. In my opinion, en-gb should always follow a fallback, except in cases, where the language is a bit different (e. g. theater -> theatre; soccer -> football etc.). I completely agree with not using de-formal in items. The fallback is absolutely sufficient in this case. Regards, Vogone talk 01:06, 28 April 2013 (UTC)
Sure the gb-english is different but the labels are generally auto generated from the wikipedia article title anyways so someone would have to add it and it makes another point were the system can break. I think a RFC is in order here. --Saehrimnir (talk) 01:59, 28 April 2013 (UTC)

I made a little statistics using the database dump from 17 April. 769 items have labels in "de-formal" which are different from the label in "de". 22 items have descriptions in "de-formal" which are different from the label in "de". I also made a table over the number of item labels and descriptions per language. Byrial (talk) 14:15, 29 April 2013 (UTC)

Could you show us a few examples, please? Normally, there should never be differences between de and de-formal if no second person is used. Regards, Vogone talk 14:20, 29 April 2013 (UTC)
I would have made a table showing the different texts, but I made a error regarding charsets when I processed the data, so I haven't the labels and description in a printable form at the moment. The 22 items different descriptions are: Q458, Q8880, Q8886, Q8889, Q20728, Q34225, Q34236, Q34264, Q51505, Q212141, Q313530, Q316169, Q873175, Q878959, Q881504, Q882098, Q883713, Q902107, Q948584, Q1006901, Q1006917, Q1007038. 25 of the items with the different labels are: Q188, Q1148, Q1246, Q1281, Q1285, Q1412, Q1417, Q1556, Q1806, Q2370, Q2814, Q3093, Q3131, Q3139, Q3148, Q3945, Q4295, Q4339, Q5266, Q6508, Q6587, Q6961, Q6979, Q6999, Q7039. Byrial (talk) 15:12, 29 April 2013 (UTC)
A number of those are false positives in that they are different but they should not be. --Izno (talk) 15:47, 29 April 2013 (UTC)

I see exactly 0 cases where de and de-formal should be different there. I'm going to remove de-formal from all items. Regards, Vogone talk 18:27, 29 April 2013 (UTC)

A full list with the different label and description texts is now here. If anybody wants the item numbers where de and de-formal texts both exist and are the same, I can post or e-mail the list. Byrial (talk) 05:46, 30 April 2013 (UTC)
For this list, I'd also say: None of these de-formal labels/descriptions should be there (though however, a few are better than the de label/description). --YMS (talk) 07:19, 30 April 2013 (UTC)

Item with two English descriptions

According to the last database dump q3339157 have a description both for language "en" and "En". Appearently the labellister gadget doesn't like that and doesn't show the double description, nor the five labels (fr, zh, en, pt-br, pt). Byrial (talk) 16:10, 30 April 2013 (UTC)

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat/Archive/2013/04&oldid=1209327919"