Wikidata:Project chat: Difference between revisions

Content deleted Content added

Inline

Revision as of 12:48, 8 June 2016

Wikidata project chat

Place used to discuss any and all aspects of Wikidata: the project itself, policy and proposals, individual data items, technical issues, etc.
Please take a look at the frequently asked questions to see if your question has already been answered.
Also see status updates to keep up-to-date on important things around Wikidata.
Requests for deletions can be made here.
Merging instructions can be found here.

IRC channel: #wikidata^connect

Start a new discussion

Edit

On this page, old discussions are archived. An overview of all archives can be found at this page's archive index. The current archive is located at 2024/09.

Research about user participation in Wikidata - call for participation (update)

Dear Wikidata users,
We are a group of researchers of the Web and Internet Science group of the University of Southampton.
We are currently conducting a research aiming to discover how newcomers become full participants into the Wikidata community. We are interested in understanding how the usage of tools, the relationships with the community, and the knowledge and application of policy norms change from users' first approach to Wikidata to their full integration as fully active participants.
This study will take place as an interview, either by videotelephony, e.g. Skype, phone, or e-mail, according to the preference of the interviewees. The time required to answer all the questions will likely be about an hour. Further information can be found on the Research Project Page: Research:Becoming Wikidatians: evolution of participation in a collaborative structured knowledge base..
Any data collected will be treated in the strictest confidentiality, no personal information will be processed for the purpose of the research. The study, which has submission number 20117, has received ethical approval following the University of Southampton guidelines.
We aim at gathering about 20 participants. Users interested in taking part or wishing to receive further information can contact us by writing to the e-mail address ap1a14+wikidata_user_study@ecs.soton.ac.uk

Thank you very much, your help will be much appreciated!

--Alessandro Piscopo

Problems with IE

Hi. Does anyone know why I get permanently the message ”An error occurred while saving. Your changes could not be completed. Details: Forbidden” when I want to save something? I use IE11/Win 8.1 and for a while I can not save anything. Thanks. Haptokar (talk) 06:54, 21 May 2016 (UTC)[reply]

~~You've got an older browser that's no longer fully supported. Some functions might indeed not work. Upgrade to a newer version of IE or use another browser.~~ Mbch331 (talk) 08:10, 21 May 2016 (UTC)[reply]

Maybe this is related to this problem? --Succu (talk) 08:51, 21 May 2016 (UTC)[reply]

@Haptokar: if you are using a computer where you can't install new programs then you can run Firefox portable without installing it --John Cummings (talk) 20:49, 21 May 2016 (UTC)[reply]

Thank you all. There is no problem with Chrome and Firefox, but only with IE, the browser I use for editing Wikipedia pages and it become frustrating to change browsers. Haptokar (talk) 02:46, 22 May 2016 (UTC)[reply]

It probably is the problem linked by Succu. The developers are looking into the problem. Mbch331 (talk) 06:48, 22 May 2016 (UTC)[reply]

when will there be a solution? --Hannes 24 (talk) 16:50, 27 May 2016 (UTC)[reply]

FYI, the same problem was reported by User:Massimo Telò at the Italian village pump, it is quite important to fix it, I guess. So far the only solution we can suggest is to use another browser.--Alexmar983 (talk) 03:59, 1 June 2016 (UTC)[reply]

I believe we have fixed it now. If you still have issues can you please let me know? Thanks! --Lydia Pintscher (WMDE) (talk) 07:56, 1 June 2016 (UTC)[reply]

Wikidata Module:Taxobox

@PhiLiP, FelixReimann, Succu, Infovarius: As I don't seem to get anywhere, (no response since 10 March) with implementing the WD taxobox on cywiki, I've suggested that the only way forward is for me to automatically create all infoboxes similar to one of these here on WD on my userspace. I can then harvest the info onto a csv file and write to cywiki. Any problem with this? Llywelyn2000 (talk) 12:03, 25 May 2016 (UTC)[reply]

@Llywelyn2000: There is just nobody interested in Cymru wiki.--134.169.240.18 13:39, 25 May 2016 (UTC)[reply]

I think you're right. Maybe we should all just work on one wiki - Chinese? Thanks for your positive comments. Llywelyn2000 (talk) 13:53, 25 May 2016 (UTC)[reply]

You're right, Llywelyn2000: China has taken us.--134.169.240.18 14:23, 25 May 2016 (UTC)[reply]

Sorry but I can not help, because I have no LUA knowledge. But maybe the ArticlePlaceholder extension is helpfull for you (taxon 1, taxon 2). See Lydias email. --Succu (talk) 14:59, 25 May 2016 (UTC)[reply]

Thanks Succu. We have a discussion about enabling ArticlePlaceholder, which is still open. I should think we may agree in 3 weeks - when technical issues are sorted. Llywelyn2000 (talk) 15:16, 25 May 2016 (UTC)[reply]

@Llywelyn2000: It would be nice if we could have centralized Modules for the smaller Wikis, so we don't have to create 300 out-of-sync versions of the same code. In any case we really need more people involved in writing easy to understand documentation for the Wikidata modules and Wikidata-powered Templates. Maybe a good Hackathon for Wikimania? --Tobias1984 (talk) 18:19, 31 May 2016 (UTC)[reply]

Hey! Thanks for this! Yes - 100%!!! 279 language wikis have to reinvent the wheel every time they need a new template - or import and translate, which is very time consuming. All templates should be on Meta (imho!) with translation of visible text ONLY on local wiki. And pigs will fly... Llywelyn2000 (talk) 06:20, 5 June 2016 (UTC)[reply]

See Phab:T6547 for exactly this, but it's not exactly happening quickly - it's been open since 2006! Thryduulf (talk: local | en.wp | en.wikt) 09:21, 5 June 2016 (UTC)[reply]

Arm

In the English language the word "arm" is a polynsym. It happens to both refer to the upper arm as well as refering to the full arm.

Q43471 is currently named "arm" and is linked to the FMA concept meaning upper arm.

On the other hand Q43471 has the German name "Arm" which only means get's full arm and doesn't mean upper arm. In German the word can include the hand but doesn't have to. The description suggest that in this case the whole Free upper limb is meant (Hand + Forearm + Upper Arm).

Q43471 is linked to [[ http://www.ontobee.org/ontology/UBERON?iri=http://purl.obolibrary.org/obo/UBERON_0001460%7CUBERON_0001460]] which describes the upperarm + forearm.

There's an existing concept for the upper arm: https://www.wikidata.org/wiki/Q379859

Q43471 also has names in a lot of other languages where I don't know what the word means. Is there a process for untangling the mess? – The preceding unsigned comment was added by ChristianKl (talk • contribs) at 14:04, 26 May 2016‎ (UTC).[reply]

"refer to the upper arm as well as refering to the full arm" Really? Do you have a source for that? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:16, 26 May 2016 (UTC)[reply]

The FMA concept "Arm" means the upper arm. That's easily verificable by seeing that the Free upper limb is made up of the parts "Arm", "Forearm" and "Hand". The Forearm in FMA is not part of the Arm. NBCI's MeSH Controlled Vocabulary defines "Arm" as: "The superior part of the upper extremity between the SHOULDER and the ELBOW."

For extra fun NBCI's MeSH Controlled Vocabulary defines "Arm Bones" as "The bones of the free part of the upper extremity including the HUMERUS; RADIUS; and ULNA." Radius and Ulna are bones in the forearm that are not between shoulder and elbow. That definition of arm bones doesn't include the bones in the hand hand. Webster defines "Arm" as " : a human upper limb; especially : the part between the shoulder and the wrist". The human upper limb includes the hand. Thus we have at least three distinct concepts that are named "Arm" in English.ChristianKl (talk) 11:23, 27 May 2016 (UTC)[reply]

In every day English "arm" means either the upper arm and forearm as a whole, or any part of your upper arm or forearm when specificity does not matter. It may sometimes include the hand, but usually does not. "Arm" is not used to refer only to the upper arm. "Upper arm" is less likely to be used than "fore arm" when specificity does not matter, but "arm" is more likely to be used in all cases. Thryduulf (talk: local | en.wp | en.wikt) 12:26, 27 May 2016 (UTC)[reply]

"fore arm" is no valid term in English. "forearm" is the term. In any case scientific usage of the term "arm" matters. Because "arm" means "upper arm" in the scientific sense of the word, Q43471 is currently linked FMA24890. If we decide that Q43471 refers to forearm + upper arm it shouldn't link to FMA24890. Q43471 should then also get a English description that doesn't just say "body part" but that makes it clear what's meant. Then Q379859 should get FMA24890. The problem is that various interwiki links do think that Q43471 currently describes the upper arm. For example the Occitan Wikipedia (https://oc.wikipedia.org/wiki/Bra%C3%A7_%28anatomia%29) shows an image of the upper arm and talks about Os bones being about the "umèrus" with sounds like humerus (the bone of the upper arm) but with doesn't mention the bones of the forearm.ChristianKl (talk) 18:13, 27 May 2016 (UTC)[reply]

I think there's disagreement here that arm (Q43471) shouldn't refer to upper arm. I would propose to rename Q43471 into `arm (including hand)`, create a new property called `arm (excluding hand)` and move everything related to the upper arm to the existing Q379859. Given that there a mix of various translation on Q43471 I would then delete most of the language links in languages where I'm not sure whether the name is saying upper arm or arm and hope that sooner or later people who speak those languages add terms themselves. I still don't fully understand to what extend I'm supposed to seek consensus over an issue like this before going and making the change. Does anybody disagree with my proposal? ChristianKl (talk) 08:54, 2 June 2016 (UTC)[reply]

Importing Wiki Loves Monuments lists into Wikidata

Dear all

I’m posting this to start a discussion on migrating the information in the Wiki Loves Monuments (WLM) database to Wikidata as part of the Connected Open Heritage project. There have been discussions about the work previously here and here.

Below we have outlined what we can see as the potential approaches and issues with each approach. We would very much appreciate your feedback including the following questions:

Which approach do you think would work and why?
Are there ways to mitigate or overcome the issues highlighted?
Are there any other approaches or issues that have been missed?
If the combination of the two approaches outlined is a good approach, what combination would work best? Would this depend on the country or could universal guidelines be used?

Approaches

Wiki Loves Monuments lists for different different countries are constructed in different ways:

Some countries use only official immovable heritage lists
Some countries add sites onto official national and regional registries e.g South Africa.
Some countries do not have official lists.

In addition some countries have lists with unclear documentations as to the source of the data and methodology used to select objects.

There appear to be three approaches to importing the WLM database into Wikidata taking the construction of existing databases into consideration:

Import directly from the WLM database into Wikidata.
Import from the reference sources used to created the WLM databases where known.
A combination of the two approaches.

Issues

There seem to be a number of issues with each approach:

Importing directly from the WLM database into Wikidata

By importing straight from the WLM database rather than going back to the sources there might be no referenceable sources connected to the items created, potentially we will be creating up to 1,384,000 unreferenced items. However this could be solved (in the short term) by an "imported from Monument database", wlm-id:<id>, date:<date>" reference. This is not ideal but is essentially how importing from Wikipedia works today.

Even if the data is based on official lists there is no simple way of determining which changes have been made to the data since the time it was imported.

Whilst facts are not copyrightable there might be elements of the Wikipedia lists (and hence the monuments database) which are copyrightable. If so there is an issue in that the Wikidata license is more restrictive than that of Wikipedia.

Import from the reference sources used to created the WLM databases

Some WLM countries add sites onto official national and regional registries e.g South Africa and some countries do not have national registers.

Some of the original official documents/lists might be copyrighted. It is currently unclear what copyright restrictions might apply if we are looking at importing the WLM lists/Monument database vs. the original lists that the WLM lists are derived/ copied from?

This approach would not make use of the work that went into standardising the info from the various sources for the Monuments database.

Again we would very much like your feedback on this

Thanks

John Cummings (talk) 15:43, 26 May 2016 (UTC)[reply]

Discussion

Would it be possible to import the data from both sources, i.e. from the monuments database (corresponding to the data contained in the monuments lists mostly sitting on Wikipedia) and from an official source (where available)? – In a second step, identical double entries could be removed (keeping the one that is referenced to the official source). In a third step, differences in double entries could be dealt with (in some cases we may want to keep them as they are; in some cases we may want to report errors back to the maintainer of the official source; and in some cases we may have to correct the entry on Wikipedia). And in a fourth step, missing data could be added (in Switzerland, for example, the official source contains labels only in the official language of a given territory, while we should be able to generate lists in various languages; translation would most easily be done in a spreadsheet, translating all the monuments labels in one go in order to remain consistent across the lists). --Beat Estermann (talk) 13:34, 27 May 2016 (UTC)[reply]

I assume you already read the past discussions on Wikidata-WLM, right? Nemo 14:04, 27 May 2016 (UTC)[reply]

I have certainly followed some of them. But I'm not following the developments closely enough to judge whether one or the other point made a couple of years ago still holds given the actual development stage of Wikidata. - Are you referring to anything specific? --Beat Estermann (talk) 14:11, 27 May 2016 (UTC)[reply]

Yes, a search page of 81 emails stretching back five years isn't exactly enlightening. A ink to a specific post, or even better a statement here rather thank a link, would be much more helpful. Chase me ladies, I'm the Cavalry (talk) 15:32, 27 May 2016 (UTC)[reply]

I expect people to re-read past discussions before opening new ones on the same topic, but I understand that self-discipline is hard. --Nemo 06:29, 6 June 2016 (UTC)[reply]

I think Beat's option of importing in stages from multiple sources is preferable to create something new and more comprehensive than just the WLM database, with the process of de-duplication, and supplementation built in (We also need to refine some of the UK list to exclude the 500k grade 2 listed buildings as it's just too much for Wikidata). Additionally, referring someone to a massive database of emails isn't much help as it implies that this isn't a place to ask questions until you've trawled archived emails. Asking a question here might get you a better, up to date answer, and maybe a little assistance in executing your idea. Battleofalma (talk) 15:40, 27 May 2016 (UTC)[reply]

Exclude Grade II? Sad face. Chase me ladies, I'm the Cavalry (talk) 15:42, 27 May 2016 (UTC)[reply]

Perhaps not all, but 500k is too much, so there's got to be some sensible way of reducing that to a Wikidata acceptable number. We don't need all those Naval forts... Battleofalma (talk) 15:45, 27 May 2016 (UTC)[reply]

HashtagNotAllGrade2ListedBuildings Battleofalma (talk) 15:46, 27 May 2016 (UTC)[reply]

HashtagSomeMonumentsMatter? --Izno (talk) 15:49, 27 May 2016 (UTC)[reply]

If Wikidata cannot currently ingest 500,000 items for listed buildings in the UK then perhaps we can say 'yes but not yet'? John Cummings (talk) 15:51, 27 May 2016 (UTC)[reply]

Currently we appear to have 2324 Grade II listed buildings, out of a possible 370,337.

@John Cummings: Am I right in thinking that an upload for the UK has already been done? Comparing the results of this query: tinyurl.com/j8kqqma with the numbers in brackets at Wikidata:WikiProject_UK_and_Ireland/monuments we seem to have a reasonably full set of scheduled monuments and Grade I and II* listed buildings, and apparently a somewhat over-supply of Scottish category A and category B listed structures. Jheald (talk) 20:25, 29 May 2016 (UTC)[reply]

How many Dutch streets do we have? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:01, 31 May 2016 (UTC)[reply]

Yes I think a per country approach should be taken, and for the larger countries, a per state approach. Yes to the translation idea and I suggest using street address in native language, and for trees, parks, and other objects, the native title. Later languages that eventually create their own labels can always move the native names to an alias. Yes to the idea of importing referencing the old monumentsdatabase, but only for those regions that can't do it any other (better) way. Yes to the 500k GradeII dataset - I have seen some of these and think they are great. Who said this was too much? If there has been a RFC on this which lost, then let's just make sure all Grade II listed buildings with Wikipedia articles are at least properly referenced and part of the package. Let's do this the wikiway: start small and work from a working pilot per country. Jane023 (talk) 06:50, 28 May 2016 (UTC)[reply]

Just for information, all the 45k monuments of France have already been imported from the database to Wikidata with references to the official Base Mérimée (Q809830) (by Gzen92 thanks again to him and with some help of Wikidata:WikiProject France/Monuments historiques). This could be an issue too: don't import on Wikidata what is already in Wikidata ;) Cdlt, VIGNERON (talk) 09:37, 30 May 2016 (UTC)[reply]

For the UK, all of the higher-grade listed buildings were added to Wikidata in 2014 with help from Magnus Manske, along with the official lists of Scheduled Ancient Monuments. We imported all the official lists for England (I, II* and Ancient Monuments), Scotland (A, B and Ancient Monuments), Wales (I, II* and Ancient Monuments) and Northern Ireland (A, B+). NI Ancient Monuments were I think not done. At that time we decided not to import the English grade II listings due to their numbers (over 500k) and the fact that the vast majority of grade II buildings will never be notable enough for a Wikipedia entry. We didn't use the old WLM lists, but went back and created new ones from the most recent official listings. So for the UK, the main need will be to import any changes to the official lists from 2014. --MichaelMaggs (talk) 10:02, 30 May 2016 (UTC)[reply]

When it comes to Spain, the WLM database is far from being complete. Heritage management competences have been devolved and it means different types of protection, code assignment and the like. Furthermore, the list hasn't been deeply reviewed or kept up-to-date. Many monuments do not have a specific code as such. Such importation could be feasible, but only after an extensive review and harmonization. --Discasto (talk) 10:19, 30 May 2016 (UTC)[reply]

@MichaelMaggs: is there a write up of what was done somewhere? Or maybe just how I could find all the monuments? I assume you mean Grade I and Grade II* have been added? (I could easily be wrong). Also can you tell me what the plan is to use the list? What the benefits you saw of doing this were? Will you use Wikidata as the directory for WLM this year? Any other info very welcome. Thanks John Cummings (talk) 12:31, 30 May 2016 (UTC)[reply]

Please add all Grade II-listed buildings; it would help en:Wikipedia:WikiProject Public Art/London realise its goal of achieving complete coverage of public artworks in London. See this list of what's already on Wikidata. It needs a job lot of Grade II-listed sculptures in order to come anywhere near completion! Ham II (talk) 21:07, 30 May 2016 (UTC)[reply]

For information there is a Wikidata Wiki Loves Monuments campaign tool available. --John Cummings (talk) 15:54, 31 May 2016 (UTC)[reply]

Wikimedia Italia too was thinking of importing its own list of buildings since last year, and that's why I suggested last year the creation of Wiki Loves Monuments ID (P2186). We could possibly coordinate, what do you think? --Sannita - not just another it.wiki sysop 18:05, 31 May 2016 (UTC)[reply]

Wiki Loves Monuments ID (P2186) feels like a good fallback for countries (or sometimes individual entries) which don't have unique ids. The same property could also be used as an import reference (similar to how Wikipedia imports are done today). In the long run the statements should carry references to the official lists but that might be done in a later stage (since not all official sources have been clearly documented).

@VIGNERON, MichaelMaggs: Existing datasets should of course not be imported again. The import would start with an inventory of the existing countries where such things should be picked up/documented. The intention is also to add Wikidata connections to the existing Monument database to make it clear when objects already exist.

(Disclaimer: I'm involved in the Connected Open Heritage project and the Monuments Database) /André Costa (WMSE) (talk) 12:59, 1 June 2016 (UTC)[reply]

@André Costa (WMSE): Do we have any ideas of how we should solve this for the Swedish archaeological sites? I mean, the database is here much larger than the items used in the lists you created. A ListeriaBot-based list can of course include much more than the lists you created, but these lists will still be insanely long in many parishes. And when it comes to the connection to articles, we have articles about archaeological sites like The stoneage settlement in Bjästamon (Q19803444) but that article includes several items in the fmis-dataset. -- Innocent bystander (talk) 13:09, 1 June 2016 (UTC)[reply]

Another reason to use the wlm database would be that mistakes in the original have been corrected (typographical, but also factual sometimes. Another reason to use the original source would be because the wlm database has not always been maintained, and it could be more up to date. But please don't underestimate the effort required to standardise the data to some extent! Effeietsanders (talk) 08:55, 8 June 2016 (UTC)[reply]

Viking River Cruises

Hi

I want to clean up the Viking River Cruises listing: Viking River Cruises (Q2524174) as it is incorrect.

Parent company is Viking Cruises

Subsidiaries: Viking River Cruises and Viking Ocean Cruises

Most of the listings are under Viking River Cruises, however the website reference lists the parent company.

This is my first Wikidata edit and I'm not sure where to do this.

Thanks! – The preceding unsigned comment was added by Cruiser (talk • contribs) at 21:13, 26 May 2016‎ 2015 (UTC).

Sitelink to redirect - need opinions

In centimetre (Q174728) @Cycn: has deleted sitelink from it.wikpedidia because is a redirect, but in notability we say «Currently, the community has chosen to have redirects allowed, although the necessary changes have yet to be deployed on Wikidata.» My opinion is that we can add redirect in sitelink, the second problem specific to this item is that in it.wikidata we need to have the redirect because is used a lot in template, without the sitelink we can't add a link to the unit page using wikidata, so we need some other opinion. (nb. in it.wikipedia we have other redirect used in the same manner) --ValterVB (talk) 19:22, 28 May 2016 (UTC)[reply]

I'd really like feedback on this. As of now adding redirects isn't implemented, but people keep doing it and, appart from Valter, I only get wishful thinking in stead of arguments about it. "It should be possible" isn't a reason to do it, it should be implemented. So I hope this pushes the issue to implement this, as it was decided it should be implemented, or the issue should be re-evaluated to the point that the consensus states that adding redirects is never an option. Either is fine by me, as long as it's clear. - cycŋ - (talk • contribs • logs) 19:29, 28 May 2016 (UTC)[reply]

Hi, I only add an information. There are a lot of these redirects in sitelinks:

picometre (Q192274) has redirects in sitelinks: w:de:Pikometer and w:it:Picometro
millilitre (Q2332346) has redirects in sitelinks: w:de:Milliliter and w:it:Millilitro
nanogram (Q2282906) has redirects in sitelinks: w:de:Nanogramm and w:it:Nanogrammo

If these redirects will be removed, in itwiki we have to re-create the articles, because items like centimetre (Q174728) can't stay without a sitelink. --Rotpunkt (talk) 19:43, 28 May 2016 (UTC)[reply]

They can't? - cycŋ - (talk • contribs • logs) 19:47, 28 May 2016 (UTC)[reply]

Sure, the sitelinks are always fundamental. They are needed (1) when you are reading quantities with units of measurement: in a module I read from Wikidata 30 picometre (datavalue.value.amount=30, datavalue.value.unit=Q192274) and I want to show "30 picometre" I need the sitelink from Q192274 to the article w:en:picometre (2) sitelink are always needed when traversing items used in properties/qualifiers through arbitrary access and you want to show the link to the current traversed item. --Rotpunkt (talk) 20:04, 28 May 2016 (UTC)[reply]

I reverted your reverts User:Fomafix regarding redirects to dewiki. --Succu (talk) 22:06, 28 May 2016 (UTC)[reply]

Why? --Fomafix (talk) 07:05, 29 May 2016 (UTC)[reply]

@Succu: We are discussing, so you can't delete redirect. For now redirect are allowed and you haven't reason to delete them. --ValterVB (talk) 17:47, 30 May 2016 (UTC)[reply]

OK, let's talk. Redirs are a lot less stable than articles. They can point to anything within a current article version. They can point to a whole subsection or to special location within a subsection. They could be a relict of a former article version because the article was moved to another lemma. An not uncommon reason is a (not notable) misspelling of some kind. So on which basis we could decide about the usefullness of an redirect? I think redirects should be only maintained by the local wikimedia project involved, not us. They should not created with tricks some users are advocating here. If a template relies only on or is only used together with a redirect then these constructs are volatile Fomafix, ValterVB. --Succu (talk) 21:29, 30 May 2016 (UTC)[reply]

@Succu: Hi, millimetre (Q174789), centimetre (Q174728) are very important items, but in itwiki millimeter and centimeter (and many other smallest units) are redirects to meter. So in millimetre (Q174789), centimetre (Q174728) we have redirects in sitelinks. We *need* sitelinks for millimetre (Q174789), centimetre (Q174728) because we need to construct wikilinks for these items. If you delete these redirects in sitelinks then what we should do in itwiki? In your opinion should we replace in itwiki all the redirects with new articles? It's possible but it's a big work. --Rotpunkt (talk) 22:00, 30 May 2016 (UTC)[reply]

If this is unit specific, you could try to use the unit from the conversion to SI unit (P2370) statements and link to that page.
--- Jura 22:09, 30 May 2016 (UTC)[reply]

Hi Jura, link to which page? --Rotpunkt (talk) 22:24, 30 May 2016 (UTC).[reply]

Rotpunkt: For millimetre (Q174789) this would be "metro" from "0,001 metro" (see wikisource). Maybe formatted as millimetro.
--- Jura 06:49, 31 May 2016 (UTC)[reply]

Thanks Jura, but no, we need a general solution, that works everytime we need a wikilink from an item. Are redirects allowed as sitelinks? (1) Yes, we will continue in this way (2) No, we will replace redirects with articles. --Rotpunkt (talk) 09:55, 31 May 2016 (UTC)[reply]

My opinion is that we should let each client decide themselves if they should allow redirect as sitelinks or not. The potential problem and the potential gains are all theirs. While I see some redirect-sitelinks here as hazardous, I see others as very helpful. A redirect to a spouse are among those I dislike, while I think redirects in "millimeter" to meter makes sense, even if I personally do not think links in units are important. -- Innocent bystander (talk) 10:08, 31 May 2016 (UTC)[reply]

Technically, you can't currently create wikilinks to redirects. This is a good thing as these create maintenance issues at Wikidata. Sometimes they happen following pagemoves and we end up with two items at Wikidata, one to the redirect, one to the article. This is problematic for the client wiki as well, as the information needed may not be on the item associated with the article. On a side note, we did have a mess with units as some languages imported all redirects leading to an article about a unit as aliases to the associated item (sample)
--- Jura 10:10, 31 May 2016 (UTC)[reply]

I don't understand this part: << you can't currently create wikilinks to redirects >>. Go to itwiki and write in a sandbox: [[millimetro|mm]]. It works and "millimetro" is a redirect. What did you mean? --Rotpunkt (talk) 10:34, 31 May 2016 (UTC)[reply]

It should read "sitelinks on items to redirects at a wiki".
--- Jura 10:42, 31 May 2016 (UTC)[reply]

Ah ok, yes we know, as written in the first post << ... although the necessary changes have yet to be deployed on Wikidata. >> --Rotpunkt (talk) 10:48, 31 May 2016 (UTC)[reply]

It would probably need additional functionalities to ensure that the problems we encounter with pagemoves and the like don't proliferate. Ideally we would already have these tools now, just to ensure that we keep the amount of noise in the database low.
--- Jura 11:25, 31 May 2016 (UTC)[reply]

It is not only after page moves redirects appear. They often appear after an article has "de facto" been deleted or merged. The article has been converted to a redirect to a list-article. Or, the cases I hate the most, when they are redirected to their "father"/"spouse" or company they are related to. I currently work with top-level Internet domain (P78) and sometimes redirects to list-articles are used here in the present Wikipedia-templates. Not all redirect-sitelinks looks bad to me, especially not those who leads to list-articles. -- Innocent bystander (talk) 12:50, 31 May 2016 (UTC)[reply]

Maybe a log of sitelinks that are new redirects would be a good start.
The (probably simpler) problem we have with sitelinks on disambiguation items is somewhat related and I don't think we get closer to solve it.
--- Jura 13:12, 31 May 2016 (UTC)[reply]

Did you read my message Rotpunkt? Good night --Succu (talk) 22:12, 30 May 2016 (UTC)[reply]

@Succu Yes, I read, "these constructs are volatile", I agree, but then which is the solution? Should we have in itwiki articles instead of redirects for "millimeter", "centimeter", ... ? Surely we can, but it's quite a big work. But If this is the solution we will do it, no problem. --Rotpunkt (talk) 22:24, 30 May 2016 (UTC)[reply]

Note we have discussed allowing redirects as sitelinks for a LONG time and nothing has been done - people who care about this should probably go comment at the T54564 phabricator discussion

Tracked in Phabricator
Task T54564

; perhaps we need a new RFC to establish consensus here on this. Personally I think this would be a very useful thing to implement properly (right now it is obviously possible but not well-supported by the wikidata UI and not at all supported by the wikipedias on the other end insofar as they do not see wikidata links to redirects at all). There are a number of other suggested options in the phabricator discussion. ArthurPSmith (talk) 14:03, 31 May 2016 (UTC)[reply]

Recap

At today redirect are allowed (see at Notability policy), technical feasibility is not a problem we can use some workaround (is the same for coordinates not on the Earth, we can add them only with API). I think that to change this situation is necessary a new RfC, and without a new RfC delete redirect isn't allowed (except for vandalism error etc.) --ValterVB (talk) 10:06, 2 June 2016 (UTC)[reply]

Hide some bots from watchlist?

Is there a way to exclude certain bot edits from my watchlist (e.g. based on a partial string match of the edit summary) without hiding all bot edits? For example I'd like to hide the "bot: import label/alias from" edits by user:Cewbot but not the "Added link to" edits by user:S205643bot. Thryduulf (talk: local | en.wp | en.wikt) 23:43, 29 May 2016 (UTC)[reply]

You can use c:MediaWiki:Gadget-rightsfilter.js from Commons. At special:mypage/common.js place this line:

mw.loader.load( '//commons.wikimedia.org/w/index.php?title=MediaWiki:Gadget-rightsfilter.js&action=raw&ctype=text/javascript' );

--Edgars2007 (talk) 15:00, 30 May 2016 (UTC)[reply]

Not perfect, but good enough. Thank you. Thryduulf (talk: local | en.wp | en.wikt) 20:48, 30 May 2016 (UTC)[reply]

But it's flexible, that's why I like it :) --Edgars2007 (talk) 21:11, 30 May 2016 (UTC)[reply]

BTW, w:Wikipedia:Village pump (technical)#Hide one user's edits from watchlist. --Edgars2007 (talk) 05:43, 2 June 2016 (UTC)[reply]

Linking to Wikidata from Meta to show tem names in specific languages

Hi

I'm not sure what this is called (which is why I'm struggling searching for it), I'm looking for if there is a template (or other method) I can use on Meta to link through to Wikidata items that will show their name if I just input the Q number or if I just use the name and use a normal link? If it is possible is it also possible to force it to show in a specific language? I have a big list of numbers I want to link to and adding the names individually is going to take me ages.

Thanks

John Cummings (talk) 09:26, 30 May 2016 (UTC)[reply]

meta:Template:Label ought to help you, but at the moment it doesn't work properly; you can try to adjust Wikidata's version of the template to work with Meta. Mahir256 (talk) 17:30, 30 May 2016 (UTC)[reply]

@Mahir256: thanks, I think I would make it more broken if I changed it..... do you know anyone who might be able to fix it? Also is there some way that the name can be a link to the Wikidata item? John Cummings (talk) 13:08, 1 June 2016 (UTC)[reply]

@John Cummings: It works now. Just do something like {{Q|322155|bn}} on Meta to get a similar result. Note that this second parameter does not work on Wikidata itself just yet. If you want to use the template on other wikis, just copy Meta's version to the appropriate wiki's Template:Label page. Mahir256 (talk) 16:17, 1 June 2016 (UTC)[reply]

@Mahir256: Amazing, thanks so much, I'll share what I've done with it soon :) --John Cummings (talk) 19:54, 1 June 2016 (UTC)[reply]

Data Extractaion

Is it possible, given two job designations, to find the relation between them through Wikidata? For eg: given CEO and VP- Marketing, is it possible for WikiData to give back something like, VP - Marketing reports to CEO or even works under CEO is okay! – The preceding unsigned comment was added by 115.113.175.114 (talk • contribs) at 11:06, 30 May 2016‎ (UTC).[reply]

Pywikibot question

Hello. I am trying to use pywikibot to add relegated (P2882). I am using:

pwb.py harvest_template.py -lang:en -cat:Cypriot_First_Division_seasons -template:"Infobox football league season" -namespace:0 relegated P2882 promoted P2881 -pt:0

(I am using the Greek version of category and template)

Its working. But it only add the first team. Not all of them. For example, en:2014–15 Cypriot Second Division at promoted parameter we have

[[Enosis Neon Paralimni FC|EN Paralimni]] <br/> [[Pafos FC]] <br/> [[Aris Limassol F.C.|Aris]]

It only add Enosis Neon Paralimni FC. What can be done?

Xaris333 (talk) 21:40, 31 May 2016 (UTC)[reply]

It works as it's meant to be working, because in some cases (like place of birth) you would want to add only one value, but at parameter there may be many links (city, adm. division, coutry). But yes, some option for adding all values would be nice, also for other infoboxes (like film infobox for actors etc.). --Edgars2007 (talk) 08:44, 1 June 2016 (UTC)[reply]

So there is no solutions for my problem? Xaris333 (talk) 11:53, 1 June 2016 (UTC)[reply]

Simple solution - no. A little(?) bit harder - include such option in harvest_template.py :) --Edgars2007 (talk) 14:49, 1 June 2016 (UTC)[reply]

How? Xaris333 (talk) 22:14, 1 June 2016 (UTC)[reply]

Verifiability policy

I think it is well beyond time that we created a workable policy regarding sourcing of information. Much of the disputes regarding Wikidata:Notability revolve around the vagueness of the second point regarding "serious" sources, and Wikidata:Living people only makes sense if we define what verifiability means to us. I know that we strive to keep policies at a minimum, but with most of our information lacking sources, our database isn't as reliable as it should be.

I hope to solicit enough ideas to draft an RfC. To start it off, I think that the principle that anyone should be able to verify the information in our database should be applied. In terms of reliability, we have to exclude inherently unreliable sources such as most social networks, forums, blogs, and other wikis (since they can be easily edited).--Jasper Deng (talk) 00:11, 1 June 2016 (UTC)[reply]

Jasper Deng I agree although at the moment I am just observing the debate on the topic. Have you seen m:WikiCite 2016 by the way?--Alexmar983 (talk) 04:15, 1 June 2016 (UTC)[reply]

@Alexmar983: No I haven't. From the looks of it, though, I think WikiCite is putting the cart before the horse - how are we to host citations if we have no concrete policy on them?--Jasper Deng (talk) 04:22, 1 June 2016 (UTC)[reply]

My dear chap I agree, that's why I am trying to learn something here. I also have the feeling the situation is as usually evolving in a very fragmented and non-linear way, so I do the only neutral yet useful thing a wikimedian can do: spreading the information. I also wanted to open a RfC in general about "overall source policy on wikimedia platform", that's why I am trying to learn as much as possible. I don't think I can help you more than that.--Alexmar983 (talk) 04:26, 1 June 2016 (UTC)[reply]

I agree something is needed here. My first thought on your proposal though is that I don't think we can simply social networks, etc. as unreliable for all purposes - e.g. Facebook location ID (P1997) and Facebook username (P2013) can likely be reliably sourced to Facebook, although I certainly agree that it is going to be an unreliable source for most other properties. en.wp allows official blogs to be used as sources for statements made by the blog's author and factual information about themselves - I can also see a potential need here for those sorts of sources for people who are notable only based on the structural need criterion.

The practical upshot of this is that any policy provision about reliable sources is going to have to be nuanced, but I believe that such is possible. Thryduulf (talk: local | en.wp | en.wikt) 11:08, 1 June 2016 (UTC)[reply]

Is it possible to consider a general rule along the lines of: sources should be verifiable either online with a URL, or via widely available printed materials. Online sources should have some guarantee of permanence: the organization hosting the content should be stable and have some commitment to long-term preservation. Statements should attempt to be supported by multiple independent sources where possible. ArthurPSmith (talk) 13:47, 1 June 2016 (UTC)[reply]
- Isn't the permanence aspect covered by archive URL (P1065)? I don't think requiring printed sources to be "widely available" is a good requirement - it's fine for mainstream or contemporary subjects (e.g. it would be a plausible requirement for Tom Daley (Q313885)) but for more obscure or historical ones then far less so (eg. there will be information about 18th Century mining of coal (Q12880211) in County Durham (Q23082) available only in the archives of North of England Institute of Mining and Mechanical Engineers (Q7057517)). Thryduulf (talk: local | en.wp | en.wikt) 15:50, 1 June 2016 (UTC)[reply]

I don't think a policy will help sourcing, actually. More use of Wikidata will. Plus our main clients, wikipedias, already have sourcing policies. In the end those policies should by side effect bring sources when it will be easier to enter and use them from wikidata. This is the key, and certainly not the policy. Please also note that technically Wikidata is potentiallyTemplate:Ref needed way more source intensive as sourcingTemplate:Ref needed a wikidata statement is roughly the sameTemplate:Ref needed as having several referencesTemplate:Ref needed for a single sentence in a wikipedia articleTemplate:Ref needed. author TomT0m / talk page 14:42, 1 June 2016 (UTC)[reply]

@TomT0m: I view it differently. At least with respect to the English Wikipedia, they are hesitant to rely on our data if it doesn't meet their quality control standards, most notably verifiability. --Jasper Deng (talk) 02:15, 2 June 2016 (UTC)[reply]

@Jasper Deng: I doubt that in their policy there is a need to source every field in every infobox. They can know if a statement meets their standards by checking its source, or don't use using software non sourced statements. They can also enforce their policies by working on their wiki through our data. author TomT0m / talk page 06:40, 2 June 2016 (UTC)[reply]

@TomT0m: Strictly speaking, policies like en:WP:BLP (similar to our proposed living persons policy) require it for infobox parameters. It's also not their job to filter our data for verifiability, that is our job to do - the whole premise of our project is that machines can use our data and be assured it's reliable.--Jasper Deng (talk) 17:01, 3 June 2016 (UTC)[reply]

@Jasper Deng: Jasper Deng Wait a minute, how could we make such a frontier beetween wikidata and wikipedia ? When wikipedians filled infoboxes, it was their job, but when they use datas, they are just consumer ? Then this will never work. We're a small community, we can't handle the sum of all wikipedias infoboxes by ourself. author TomT0m / talk page 09:10, 4 June 2016 (UTC)[reply]

It is difficult to identify for a module/template which information needs sources and which does not. The statement "Elvis Presley occupation:Singer" is probably nothing anybody have any objection against, not even unsourced in a BLP. But saying "Barack Obama occupation:Singer" would probably need a source. In the templates I have helped to develop on svwiki, I have tried to always make it optional to opt-out from the Wikidata and add local data instead. People uses that options from time to time. I even do it myself. Sometimes it has been necessary to keep it completely Wikidata-free. The objections against "imported from:Wikipedia"-sources has been fewer than I have expected, but they exists. -- Innocent bystander (talk) 09:23, 4 June 2016 (UTC)[reply]

I think every wikidata infoboxes on frwiki have an opt-out feature (and an opt-in for some fields since very recently is activated and possible) It's been a requirement and a global community agreement to develop infoboxes from the start, and we included it in the infobox-building lua modules and templates from the start by default. As all infoboxes use those they all have opt-out by default as a side effect. On the "imported from: Wikipedia" it's usually enough to recall they definitely and obviously are here for tracability and not for sourcing, (good faith) people usually understand. It would be easy to develop an include a filter to exclude those statement in the templates that pulls data from wikidata if a community became really adverse to those statements. The problem I see with opt-out is that you kind of lose the ability to be updated if the data from wikidata are corrected and finally meets your requirements for example, and that you cut yourself from the possibility to enhance Wikidata for other wikipedias and data consumers. Too much use of it would be very damaging for us. author TomT0m / talk page 09:47, 4 June 2016 (UTC)[reply]

@TomT0m: But this is precisely what Wikidata is for - centralizing all of Wikipedia's and other projects' information into one central location (here), which implies maintaining it. That this requires a lot of manpower is not a valid argument against the need for verifiability of our information.--Jasper Deng (talk) 10:02, 4 June 2016 (UTC)[reply]

@Jasper Deng: Please reread the beginning of the discussion. Our main goal is to have good datas. Will a policy really add good datas and sources ? It remains to be proven. What I think is required actually is manpower. And we won't have manpower if we persuade or acknowledge wikipedians to be mere consumer. What we must make them understand is that as data providers we will just have datas that are as good as our users will enter them. If they have high data standards, then ... it's their responsability to raise the quality of wikidata datas. As it's a centralizing tool, as more wikipedias will create and use the datas, we'll enter a win-win state where the high standard of the sum of wikipedias will make better all the wikipedias datas. We will centralize the manpower. This is how this will be a success. As our consumers already have high data standards, do we have to have it ourself ? This is an open question. This is a social matter : if it's necessary to reach a win-win state, then maybe we need it. BUT : this could be a big headeach and a big fight and as such a very high risk. Some people are probably very proud of their quality standards and could look not as high standards with condescension and dedain "we won't use datas with such poor requirements" or in the opposite "wikidatans are pain in the ass reverting good datas because of their ridiculously high standards impossible to reach". So we could reach a lose lose state where everybody stays at home ... author TomT0m / talk page 10:17, 4 June 2016 (UTC)[reply]

A few users on svwiki sees the presence of bad sources and lack of sources as a reason to come here and edit. Bot they are only a few. One problem here is that sometimes the infobox and the text tells two different things. They may be contradicting but most often one of them are not updated (and not always Wikidata). There are also some disputes about what information we should allow in our articles. How do we see that a spouse, child or other relatives are notable enough to be mentioned in a BLP? And how do we create red links? -- Innocent bystander (talk) 10:32, 4 June 2016 (UTC)[reply]

@TomT0m: You cannot even speak of "good datas and sources" when that has not even been defined. The purpose of a policy is to define this so everyone is in agreement over the meaning of that. Everything else in your comment is irrelevant to that. Innocent bystander's comment above shows the need for it. A policy is in general a record of the community's consensus for everyone to follow. Clearly there's no such agreement yet, so it's time to create one.--Jasper Deng (talk) 10:43, 4 June 2016 (UTC)[reply]

@Jasper Deng: What do you mean ? You want a metric for data quality if I understand. Do we need a policy for that ? How is your argumentation different from "we need a policy because we need a policy" ? The reason for a policy should however be at least discussed, overwise we may reach a dead end and take useless risks. Your rejecting of my arguments are way too harsh. author TomT0m / talk page 10:49, 4 June 2016 (UTC)[reply]

@TomT0m: No you don't understand, please re-read my initial comment at the top of this thread. I don't understand your argument, which seems to be that "writing down what the community expects will not help the community fulfill those expectations". That just doesn't make sense. I want the community to write down its verifiability standard, so we have a clear and consistent standard to apply. The need for that has been exhibited by disputes over the notability policy and other places. In addition, the living persons policy, an important principle of Wikimedia, is hard to write without a verifiability policy. What other projects choose to do about it is simply tangential to that. --Jasper Deng (talk) 10:54, 4 June 2016 (UTC)[reply]

@Jasper Deng: No, it's not tengential, it's the keypoint because we're in an ecosystem. If we don't take into account our clients, we might lose them or not winning them, which is probably the worse that could happen to Wikidata. Is Wikidata here to serve clients ? probably, as it's a data repository, and a data repository is nothing without use of their datas. By keeping us application neutral, we could maximise our potential usage and usefulness and minimise the interprojects conflicts. If it's to the client to check if the data meet their quality standards, then we don't lose any client and potential usecase. Do we want to maximize usefulness or enforce strict policies for the sake of enforcing strict policies, with all conflits and energy waste that comes with that are implied ? That's the key question of Wikidata in my humble own. Do we want to align on the standards of the most inclusive wikipedia or the worst one ? Whatever we do, we might lose both. And it's not an unimportant thing. author TomT0m / talk page 11:25, 4 June 2016 (UTC)[reply]

On the other hand, if we keep it that way, we may be useful for the most inclusive wikipedia by putting their datas to the standards of the most exigent one :) author TomT0m / talk page 11:26, 4 June 2016 (UTC)[reply]

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @TomT0m: Again, it is not their job to check the veracity of our data, and the convention is that they won't; it is not any more appropriate for Wikipedia communities to directly dictate our verifiability standards than it is for Wikipedia to dictate Commons' policies. Your argument remains unconvincing: Wikipedia's do not trust us precisely because we currently do not have a verifiability standard they can easily refer to (a common reason for other wikis not using Wikidata phase 2), and there has been discord in our existing community caused by a lack of a single agreed-upon standard. Also, nowhere did I ever suggest that we should be more biased towards one wiki's standards or another's. The whole point of enacting a verifiability policy is a policy of our own, one that is acceptable to and created by our own community. There's no loss of neutrality there. By your argument we could not create any policies such as WD:N because they somehow violate "neutrality". Complete nonsense, as you can see - we have the ability to independently create policies without prejudice towards a particular wiki's standards. Also, Wikipedia editors do not come here in isolation from one another. Applying different standards from different projects is a recipe for edit warring and other conflicts. You are also plainly incorrect that this policy must somehow be draconian - WD:UCS is still our primary guidance and we make policies only when that fails to stave off conflict (which is the case here). The bottom line is: we need to get agreement on what verifiability means, not just to reduce internal project conflict, but to also make clear to other projects what our community decides to do (others will want to know what information can be included here or not), and to also formulate WD:Living people. There is no arguing with that, period.--Jasper Deng (talk) 12:13, 4 June 2016 (UTC)[reply]

@Jasper Deng: That's kind of dictatorial. Then it may just fail. author TomT0m / talk page 11:06, 5 June 2016 (UTC)[reply]

Wiki-related properties

Hi, I have another general question I would like to know if it has been globally addressed.

The vast majority of our "wiki-related" information for a generic item (I hope you can tolerate my vague terminology) is usually hosted in the interlink section of the items and are the links to other platforms.

There are also some "wiki-centric properties" that we use, such as for example Wikivoyage banner and in the field of Identifiers Wiki Loves Monuments ID or Wikimedia database name ... and so on.

I am not an expert but isn't it potentially misleading to mix up properties and identifiers from external sources with something we use on the wikimedia platforms? If I were more expert I would also say that those properties and IDs should be treated "differently" in some way. they could have a different color background, or be listed physically separated at the end, or something like that.

I don't think we even have a generic property to describe them. For example P2186 is an instance of Wikidata property for authority control for places and multi-source external identifier, both encompassing also IDs from external sources, there is nothing like a "wikimedia internal identifiers" as an instance, am I right?

Any feedback?--Alexmar983 (talk) 04:48, 1 June 2016 (UTC)[reply]

Well, my opinion is that page banner (P948) can be useful for any project who wants to use a panorama-file. It then have the same purpose as image (P18), but for a special kind of picture. Any project who wants to use this property can use it. Maybe we should rename it? -- Innocent bystander (talk) 05:39, 1 June 2016 (UTC)[reply]

I have never said it is useless, my comment was not about removing them at all and I don't have problem with the "name" per se, my question is about a general hierarchy of concepts, which is something I really try to focus when discussing newbies doubts onwiki and offwiki. I can still close an eye on image (P18) because it is an image who can be used outside wikimedia platforms, and if it is on commons it can be hosted anywhere or come from everywhere, so i don't see a strong "wiki nuance" in its presence, I think a generic metadata archive would have an image property taken from the internet with appropriate copyright even if the wiki platform never existed. I am not so expert, we can debate, but it is a feeling that a commons category property (as far as i know the gallery should be in the interlink section, not the commons cat) or a wikivoyage banner could be considered something more wiki-related. I mean we can debate where to cross a line, but there is some sort of line, Commons category (P373) is not the same type of property of a birth date, it is specific of our platforms. And with the identifiers, that's even more clear: "Wiki Loves Monuments ID" exists because of us, it not a third-party ID. it is not formally correct to mix them up. there should be at least a specific subclass for them. Excuse me if my language is not proper, I hope the core of my doubt is at least clear--Alexmar983 (talk) 06:04, 1 June 2016 (UTC)[reply]

Any image property currently requires it to be hosted on Commons. The full list is at https://www.wikidata.org/wiki/Special:ListProperties?datatype=commonsMedia

Properties can have a limited use and don't necessarily combine with others. If they are sorted in one way or the other on some page, this is generally fairly random.
--- Jura 06:41, 1 June 2016 (UTC)[reply]

yes, they have to be on commons, but I don't understand why you're telling me that. The "if it is on commons" is not about a possibility regarding P18, it is about the general concept of images compared to other databases. These are free images, they can therefore be as well on another database even if wiki platform never existed, the image per se is in theory independent of the wiki platforms. On the other side, the commons cat is not, it is a property we, as wikimedians, introduced. that's why I don't have problem with P18 but I do find bizarre we put on the same level wiki-related properties with "thrid-party" properties. BTW forget the P18, I didn't even cite that in my original question, i know it can a little bit on the edge of the concept. But not the other ones.

Again, I agree it is random sorting, the point is that the random sorting is strange for this wiki-related properties (and also IDs), there is an intrinsic difference. or at least it is strange to display them with no difference with the other properties or IDs (for example a specific color background), because they come from us, not from the sources.

If we like to focus on details here, take the detail of the instances of P2186 for example. Shouldn't we have at least an instance specific for wikimedia IDs? I din't find it from its page.--Alexmar983 (talk) 07:17, 1 June 2016 (UTC)[reply]

I think that was made for Wikimedia Italy as they were looking for a way to reference some photography permit. They didn't actually start using it .. You might want to read the property creation discussion. Commons category (P373) would probably be best together with other sitelinks to Wikipedia, but for some reason, this can't be implemented. I don't think these two properties have much in common. You could add "issued by" as statement on the property page (sample at Property:P214#P2378).
--- Jura 08:18, 1 June 2016 (UTC)[reply]

It was done by them and some other chapter can do something similar, fine. it is not the specific case is the lack of framework. I can cite also the properties Commons Creator page, topic's main category, corresponding template... all these examples have in common that they are wiki-related, their value is a "wikimedian object" (something that exists because of our wikimedia system). Should we at least monitor them? For example it is already controversial how "free" a propertycreator user should be, someone says a preliminary discussion is needed by default. Well, if it is the case for a standard property I think that for such "self-referencial" properties we should be a little more careful, if possible. Now I grab them here and there but I feel that a specific definition and a link where to find them all together would be better than nothing.--Alexmar983 (talk) 10:14, 1 June 2016 (UTC)[reply]

Well, despite its usefulness to its users, the overall impact of a "Commons Creator page" property is fairly low if not nil, as the scope is limited. In comparison, the property "cite" changes the scope of the project fundamentally. Property documentation is ongoing work and contributors are always welcome.
--- Jura 11:17, 1 June 2016 (UTC)[reply]

As an idea, separating the two would solve the sourcing problem. "Wikimedia-center" properties such as Commonscat do not require to be sourced reliably, for example, "imported from Wikimedia Commons" is just fine. Then the projects can be more easily convinced to directly import these properties from Wikidata, since the sourcing is usually the main issue blocking the import.--Ymblanter (talk) 15:54, 3 June 2016 (UTC)[reply]

some questions about Properties

Hi, I'm now working in a great review of "Infobox of organization" in catalan versions. It will be shared also by "companies" or "business". It will recover from WD as much as possible data we can use. I did not find or I'm in doubt about real meaning about:

Does exist a property for:
- "Acronym" ?. short name (P1813), perhaps ?
- the (annual) budget of a company ? The budget (P2769) property seems to be restricted to an specific project, not to a regular running.
- "Expenses" or "Operating expenses" ?
What's the difference between member of (P463) and affiliation (P1416) for an organization ?. Initially, both of them were restricted to people, but after discussion, it seems that member of (P463), now is available for organitzations and countries. What should be its opposite ? (ex.:part of (P361) is opposite of has part(s) (P527))

Please, use a {{Ping}} in answer. Thanks--Amadalvarez (talk) 06:24, 1 June 2016 (UTC)[reply]

@Amadalvarez: For "budget", in wikidata you can use a "qualifier" to specify a date of validity of the statement. So if the budget of "whatever company" was "1 million €" in 2010 this can be written

⟨ whatever company ⟩ budget (P2769) ⟨ 1 000 000 € ⟩
point in time (P585) ⟨  2010 ⟩

. The precision of the year will have to be set to "year", I guess. author TomT0m / talk page 08:14, 1 June 2016 (UTC)[reply]

@TomT0m: Thanks, it sounds logic. --Amadalvarez (talk) 12:56, 1 June 2016 (UTC)[reply]

~~But ignore that "Eiichiro Oda" - in the real item you simply will have to input year (you'll see)~~. --Edgars2007 (talk) 08:39, 1 June 2016 (UTC). @Edgars2007: Fixed !, thanks.--Amadalvarez (talk) 12:56, 1 June 2016 (UTC)[reply]

P2769 is the assigned budget? Is that number even public in ordinary private companies? Maybe "sales"/"revenue"-property is more useful here? -- Innocent bystander (talk) 09:07, 1 June 2016 (UTC)[reply]

@Innocent bystander: As I understand, budget is a pourpose of incomes & expenses for next year and total revenue (P2139) is the real figures of the current or previous years. Thanks,--Amadalvarez (talk) 12:56, 1 June 2016 (UTC)[reply]

@Amadalvarez: I can imagine to use short name (P1813) for the acronym, but perhaps acronym should be proposed as its subproperty. Lymantria (talk) 08:14, 2 June 2016 (UTC)[reply]

@Lymantria:, Yes. I'm using short name (P1813) by now. For instance, NATO (Q7184) and European Union (Q458) hold acronym in this property. Thanks,--Amadalvarez (talk) 22:26, 3 June 2016 (UTC)[reply]

Wikidata helpers needed in Edinburgh for Wikidata training event

Hi, are there any Wikidatans based in or around Edinburgh who would be free to help out in a practical Wikidata editing session on 1st August? It's for a Wikidata showcase and training event taking place at the Repository Fringe 2016 event.

The session is basically an introduction to Wikidata, demo of some of the coolest things about Wikidata, then practical session to teach everyone how to edit. We're expecting an audience of up to 64 people, and currently have two people with wikidata knowledge so a couple more would be very handy for practical part. Many thanks, NavinoEvans (talk) 17:29, 1 June 2016 (UTC)[reply]

South East England is not quite close, but I could be up for it. Andreasm ^{háblame / just talk to me} 19:18, 1 June 2016 (UTC)[reply]

Similarly, I'm in London but if you can't find anyone closer I'll happily help in return for some help with expenses. Thryduulf (talk: local | en.wp | en.wikt) 23:26, 1 June 2016 (UTC)[reply]

NavinoEvans I can give you some name of users with interest in Scottish topics and active on wikidata. Maybe some of them are living in the area, if they don't state the opposite in their profile. I just don't know who are the people you contacted. I supposed User:AmaryllisGardener as a sysop was already contacted? Than I guess you have some chances with User:Drchriswilliams and also User:Foxj (not a lot of edits, but enough to teach the basics). All other users related to Scotland I scrolled down have a very limited activity on wikidata, I don't think they can really help. I can also look for users in Northern England if needed.--Alexmar983 (talk) 02:40, 2 June 2016 (UTC)[reply]

I, while interested in Scottish topics, live in North Carolina, and have actually never been to Scotland, or any of the UK for that matter. Sorry I cannot attend. --AmaryllisGardener ^talk 03:24, 2 June 2016 (UTC)[reply]

I should be able to help out with this. Drchriswilliams (talk) 08:44, 2 June 2016 (UTC)[reply]

Wow, thanks so much to all for jumping in with so many offers of help. @Alexmar983: - a really helpful list, much appreciated! @Andreasmperu, Thryduulf:, it is a long way to travel and the session actually starts at around 10am so would ideally need a night in a hotel to avoid ridiculously early travel (I'm coming from London myself). If you're still interested despite this extra hardship I can see if we can get travel and accommodation funds organised? @Drchriswilliams: Much appreciated! How far away are you based in Scotland? would some travel expenses be required? Cheers again :) NavinoEvans (talk) 10:26, 2 June 2016 (UTC)[reply]

AmaryllisGardener i am sorry i did not see your "geographical" category on your enwiki profile, there were a lot to check if I remember. NavinoEvans you're welcome, it took only 10 minutes so if you need another name I can find some more. I can look for some expats living in Scotland but active on other wiki, I guess. Or i can search on enwiki for people living in/related to northern England. Just let me know.--Alexmar983 (talk) 10:31, 2 June 2016 (UTC)[reply]

Superb, thanks again. It looks like we've got the numbers covered now, but will get back in touch if anyone can't make it. NavinoEvans (talk) 10:33, 2 June 2016 (UTC)[reply]

@NavinoEvans: I have no problem getting there a day earlier. Also, it would be extremely convenient it you manage to get travel and accommodation funds. Andreasm ^{háblame / just talk to me} 15:23, 2 June 2016 (UTC)[reply]

@NavinoEvans: Ditto. Thryduulf (talk: local | en.wp | en.wikt) 09:57, 3 June 2016 (UTC)[reply]

@Andreasmperu, Thryduulf: I've send the request for some funding, will let you know how we go as soon as I hear back. Thanks again :) NavinoEvans (talk) 16:02, 5 June 2016 (UTC)[reply]

Adding Language-labels for an Item

How would new language-labels be added to an item? I would like to add Finnish and German to the item labeled The Story of a Candy Rabbit. --Lo Ximiendo (talk) 21:47, 1 June 2016 (UTC)[reply]

If you know or are generally interested in the languages in question, you can add a babel box for the languages in question to your user page. Other than that, unless those languages already have something defined in one of the label/description/alias fields, then you must use the "labelLister" gadget in your preferences and add an arbitrary language definition that way. (There is a task for this at phab:T126510.) --Izno (talk) 22:20, 1 June 2016 (UTC)[reply]

As a temporary workaround, you can also use a url like this: https://www.wikidata.org/wiki/Special:SetLabelDescriptionAliases/Q19099978/en (changing en to the language code (in your case fi and de)). --Hardwigg (talk) 00:18, 2 June 2016 (UTC)[reply]

Thanks for the suggestions; I used the labelLister gadget. --Lo Ximiendo (talk) 05:17, 2 June 2016 (UTC)[reply]

Historical maps properties

Please have a look at the historical map properties page I have been preparing over time. I think it would be time to review it together.

I addition to anything that may come up, I'd like to ask a few specific questions.

Are we ready to start indexing historical maps in Wikidata, and would there be a notability threshold?
Are bounding boxes declared anywhere in Wikidata? I find that the properties for that are not sufficient, but if there is a practice, it should be enough.
I cannot find a property that would be used to define the date the map depicts. Is one used for any other media or should a new property be created?
Is there a property for declaring the coordinate system of a map? I was not able to locate one.

Cheers, Susannaanas (talk) 06:41, 2 June 2016 (UTC)[reply]

I like the idea of a property for a depicts date. For historical paintings this would also be useful (e.g. paintings of naval battles commissioned a few years later etc). If you propose it I will endorse it. Jane023 (talk) 08:43, 2 June 2016 (UTC)[reply]

We have a Wikidata:Notability policy, which will directly allow at least some I think. If it's unclear for your use case I think asking a specific question might get a better answer.
Bounding boxes were discussed very recently in another context but I don't recall seeing the properties (Southwest coordinates and northeast coordinates) proposed just yet.
We don't have a property for "depicts date" (the closest we have is set in period (P2408) but that is not quite right) but like Jane023, I will support one if you propose it (or I'll propose it myself if you prefer)
We don't currently have this, but again I'd support if proposed. For the bounding box I think the coordinates would need to be converted to the ones used by Wikidata if different but I'm not an expert on this. Thryduulf (talk: local | en.wp | en.wikt) 12:16, 2 June 2016 (UTC)[reply]

Better perhaps, for your naval battle example, to have "depicts -> (event item)", and put the date on the latter? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:34, 2 June 2016 (UTC)[reply]

Possibly, but a "depicts date" property would be useful for other examples. e.g. if I draw a map today depicting the Roman Empire (Q2277) as it was in AD 18 I'd want I think

⟨ Thryduulf's map ⟩ instance of (P31) ⟨ map (Q4006)  

 ⟩
depicts (P180) ⟨ Roman Empire (Q2277)  

 ⟩
publication date (P577) ⟨ 2 June 2016 ⟩
depicts date Search ⟨ 23756 ⟩

. Thryduulf (talk: local | en.wp | en.wikt) 12:53, 2 June 2016 (UTC)[reply]

Thanks for the answers, sorry about the delay. I think I will propose the depicts date and the coordinate system (unless you have already beaten me to it!). I hope to catch up with the existing bounding box discussion if I find it! Cheers, Susannaanas (talk) 08:10, 7 June 2016 (UTC)[reply]

Here's my first ever property proposal for depicted date! If you can think of more specific constraints or other necessary info, please add that! /Susannaanas (talk) 09:21, 7 June 2016 (UTC)[reply]

And the proposal for coordinate system aka spatial reference system is now up. @chippyy: @Aude: @Kolossos: The concepts coordinate system (Q11210), spatial reference system (Q161779) and geodetic reference system (Q1502887) and their instances are well mixed up. It would be great if knowledgeable people would look into them and the proposal if it wrong. /Susannaanas (talk) 11:22, 7 June 2016 (UTC)[reply]

Medicines

Hoi, the BMJ has indicated that only 12% of medicines are proven effective other sources indicate even less efficacy. A bot is going around adding substances that are registered as a medicine. In my opinion this is harmful when the efficacy is not noted. Arguably it is better not to include data that is correct in that it has been registered but wrong in that the upside is better than the downside for such substances. Thanks, GerardM (talk) 06:42, 2 June 2016 (UTC)[reply]

I have never heard that it is a requirement for a substance to be effective to be a medicine! -- Innocent bystander (talk) 07:21, 2 June 2016 (UTC)[reply]

So it is ok with you to have homeopathic nonsense? Thanks, GerardM (talk) 11:10, 2 June 2016 (UTC)[reply]

You never said it was homeopathy involved! I have no problems with having items about such substances here, nor about astrology, psychiatry or politics. If it is nonsense or not doesn't matter to this project! -- Innocent bystander (talk) 11:35, 2 June 2016 (UTC)[reply]

I think that you need to think in terms of "medicines proven effective" as being a subclass of "medicine". It's also more complicated than just being effective, e.g. things like an antibiotic that was formerly effective but is now completely resisted, medicines that are effective but have side effects that are (sometimes) judged to be too severe, medicines that are effective for some people but not others (e.g for my allergic rhinitis (Q272436) cetirizine (Q423075)-based tablets work but loratadine (Q424049)-based ones don't), medicines that are of unproven effectiveness, medicines that are effective but less so than others, medicines that are not (very) effective for their intended use but are effective for something else, medicines where the effectiveness is disputed, etc. You also need to be clear what is regarded as "effective" in each case. Thryduulf (talk: local | en.wp | en.wikt) 12:03, 2 June 2016 (UTC)[reply]

Well, I am not sure using P31 is the best approach here, just like "instance of:107,356-population city" probably is less good than "population:107,256", even if both methods technically works.

And I am not a big fan of using one kind of "science" to "prove" another. Which mathematician would judge any medicine as efficient? Therefor, I do not think we should mix modern medicine with homeopathic medicine any less than we should let astrological methods judge astronomical observations.

-- Innocent bystander (talk) 15:33, 2 June 2016 (UTC)[reply]

It has been proven yet again that irony does not work. When a substance is marked as a medicine and research proves that it does not work, it is important to know this. Wikipedia uses our stuff in its infoboxes and we have a duty to properly inform. The notion that something is a medicine because it may have been once registered as such is dangerous. People get killed in that way.. (no irony intended) Thanks, GerardM (talk) 18:07, 2 June 2016 (UTC)[reply]

If it is a homeopathy substance/medicine, it should be stated as such. That far I fully agree. They should not be mixed with modern medicine. But this project is not only devoted to modern science. We have mythology, fiction and there is of course also room here for alternative methods, homeopathy is one of them.

Yes, I have very biased feelings toward modern medicine. If I would have let the Doctors treat me according to their scientific methods, I do not know if I would have still been alive today. I do not trust homeopathy more than you do, but at least, I have never heard of anybody being killed by their methods. But I have friends who are not among us because of bad modern medicine. These scientific methods you trust, kept me unconscious during twelve months, and hospitalised me during another eighteen months. What was wrong? They gave me wrong medicine. All symptoms they found came from the medicine they put in my veins. It was not until I revolted and used my legal rights to refuse treatment I became better. I am marked for life, and that because of these scientific methods. An nescis, mi fili, quantilla prudentia mundus regatur? -- Innocent bystander (talk) 19:01, 2 June 2016 (UTC)[reply]

It seems we're on an english language related ambiguity here. In french "medecine" is about the science for sure, "médicament" is about the (active) substances, and "traitement" is about the classes of way some illness is treated. author TomT0m / talk page 18:54, 2 June 2016 (UTC)[reply]

When a scientist or his papers are controversial we state that as a fact. When it is scientifically proven that a substance has no effect as a medicine but is registered as such, we should mark this in a way that will be exposed strongly. The notion that some institution says that a substance is recognised as a medicine is secondary to the harm that the substance can do. We do have a responsibility in this. Thanks, GerardM (talk) 06:56, 3 June 2016 (UTC)[reply]

I think you are missing the point that "medicine" has much broader meanings than you are using it as, and that "effective" also has several possible definitions and even for a single definition is not a binary. We are not here to make value judgements about what is and is not correct and we must cover the facts of everything that meets our notability criteria, which includes many homeopathic and other substances as well as those used by western medicine (currently and historically). If I have understood what you are trying to do correctly then I think we should:

Be more specific with instance of (P31) statements, e.g. use items for "homeopathic medicine" and "clinical medicine" (or whatever better term) (and subcategories) rather than just "medicine"
Have a property for "approved for use as a medicine by" with items for various national/supranational drug regulation bodies and regulatory bodies for other medicine systems. This should use qualifiers for start time and/or end time, using applies to jurisdiction (P1001) as a qualifier if needed (e.g. for medical cannabis (Q1033379))

I've just discovered product certification (P1389) that might work for this. Thryduulf (talk: local | en.wp | en.wikt) 16:12, 3 June 2016 (UTC)[reply]

Use statement is subject of (P805) (or other suitable property) to link an items to studies about a substance.
Use external links or identifiers to link to clinical information about substances.

Because of the range of possible meanings and values for "effective" (my previous comment) I think that a single property for that is beyond the scope of Wikidata (we are not a resource for health information). Thryduulf (talk: local | en.wp | en.wikt) 10:34, 3 June 2016 (UTC)[reply]

How to add labels in sign languages ?

I would like to promote a mechanism to define item labels in sign languages, that is short videos linked to each item (possibly), for the different sign languages in the world.

Context. FYI, the different sign languages have alpha-3 iso codes and are listed in the following Reasonator query [1]. Besides, they are not linked to the "matching" spoken languages, for example American Sign Language (asl) and British Sign Language (bsl) are different. These videos could be somewhat related to content of the wiktionary, but numerous items have specific signs without being language words, most notably famous people. For example, french former president Jacques Chirac (Q2105) has a (rather funny) specific sign in french SL : [2].

Actually, there is already a mean to record such labels : using property video (P10) + qualifier languages spoken, written or signed (P1412), as I tested on the item word (Q8171). Nevertheless, I think this is too generic and could possibly be mixed with other videos, not directly related to the item label, resulting in confusion. That's why I would like to propose a new specific property "Item label in sign language", with a type "video file" or "image file" (animated gif would be useful too, and even some still pictures in rare cases). The statement could be multivalued, with languages spoken, written or signed (P1412) as qualifier to add the specific sign language used.

I would like to emphasize that sign languages are world living languages, as spoken languages and should be treated as close as possible to "item labels", even if these have not been taken into account while designing canonical labels. I welcome all your comments on this. GAllegre (talk) 22:27, 2 June 2016 (UTC)[reply]

Actual sign language labels will need to use SignWriting in order to be useful on projects like the American Sign Language Wikipedia, once those get launched. That's not to say that direct videos wouldn't also be useful, but they should be considered on the same order as audio pronunciation for spoken languages, in my opinion. --Yair rand (talk) 22:33, 2 June 2016 (UTC)[reply]
- I do not quite agree with you on the need of SignWriting to be useful, as most sign languages educational programs heavily rely on videos, but you are right that my proposal should be related to the pronunciation property, which uses statements of the form pronunciation audio (P443) + qualifier language of work or name (P407). For the sake of consistency, I turn my proposal into : new specific property "Item label in sign language (video) " + qualifier language of work or name (P407) (and not P1412 anymore). GAllegre (talk) 23:11, 2 June 2016 (UTC)[reply]
  - GAllegre Do not take "languages spoken or written" to literally. I see no problem using that property together with a sign language. A person in a wheelchair walks to the post office even if (s)he has no legs, and a death person speaks his sign language, even if (s)he does it with hir whole body instead of only the mouth. But I agree that P407 works best as qualifier to your proposed property. -- Innocent bystander (talk) 07:31, 3 June 2016 (UTC)[reply]
As we had discussed, labels in ase are already possible. For other languages, you can request them per Help:Monolingual_text_languages#Requirements_for_a_new_language_code. To request a new property, use WD:PP.
--- Jura 05:00, 3 June 2016 (UTC)[reply]
- Yes, we already discussed, and ASE labels, while interesting in their own, fullfill another need and are out of scope for my proposal. I do not need neither a new language code for standard labels. So, yes, I plan to request a new property. My effort here is to have a preliminary discussion to have the best proposal for an "official" request. Thanks nevertheless for your elements. GAllegre (talk) 07:50, 3 June 2016 (UTC)[reply]
Thanks a lot for noticing this issue and for this solution proposal. Yes a big interest of Wikidata is to promote multilingualism, and by this way to promote accessibility (ex: qlabel . It's difficult to believe that this could work with Wikipedias in sign languages; no one is officially active and there is ~100 sign languages in alpha-3 iso; complexity; dependency; notability; a mess of reuse... Yes, item labels seems to be logically the best theoretical approcah. But as video is not a string, in practice not sure it would be easy to contribute (need a new interface) and reuse (dealing with request results in same channel string and videos, for same context of reuse; hmm). Il like the idea of using a specific property with language of work or name (P407) because it is coherent, usable, allows to contribute already in any sign language (very important point), corresponds to the current context of edition and could be a great way to encourage contribution; and by this way data could be easily used for those who would. Best regards --Shonagon (talk) 07:04, 3 June 2016 (UTC)[reply]

I have relabelled languages spoken, written or signed (P1412) in English, to "languages spoken, written or signed". Please translate the addition of "signed". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:47, 3 June 2016 (UTC)[reply]

hi, on my talk, enjoy the reading.. --SurdusVII 12:57, 3 June 2016 (UTC)[reply]

I finally sent an official standard request : Wikidata:Property_proposal/Entity_label_in_sign_language_(video) GAllegre (talk) 17:46, 5 June 2016 (UTC)[reply]

I think I followed the guidelines to fill the request, but it's a single page, and it's not listed in Wikidata:Property_proposal/Generic. As it's my first property request, can someone explain to me how to add it to the listing page ? Many thanks. GAllegre (talk) 07:31, 6 June 2016 (UTC)[reply]

You just need to transclude the individual page on the listing page - i.e. {{Wikidata:Property proposal/Entity label in sign language (video)}}. [3] (although preferably without the typo in the edit summary I made!). Thryduulf (talk: local | en.wp | en.wikt) 10:51, 6 June 2016 (UTC)[reply]

FYI, my proposal has been re-categorized under "Term" and is now at Wikidata:Property_proposal/Term. GAllegre (talk) 07:56, 8 June 2016 (UTC)[reply]

Link WP article with different name

Hi. I created a WP Page called Scott Bishop. However there is already a WikiData article with this name so I created a Wikidata article called Scott Bishop (admiral). How do I make the WP page link to the correct wikidata entry?I tried adding wikidata:QXXXX to the WP article but that didn't work Gbawden (talk) 10:05, 3 June 2016 (UTC)[reply]

Your article is en:Scott Bishop and you created Scott Bishop (Q24287188). You need to add the former, to the latter, by editing the latter's "Wikipedia" section, choosing "enwilki" and pasting in "Scott Bishop". Q24287188 should be labelled "Scott Bishop", not "Scott Bishop (admiral)". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:29, 3 June 2016 (UTC)[reply]

Thanks for that. I have renamed Scott Bishop (Q24287188). Took me a while to figure out how to add the page to the Wikipedia section but got it now. Thanks for the help. Gbawden (talk) 12:33, 3 June 2016 (UTC)[reply]

Single value constraint for P856

I've raised a question about the single value constraint for official website (P856) on the talk page but it has so far gathered no response in about 5 days, so I'm posting here for greater visibility. Thryduulf (talk: local | en.wp | en.wikt) 10:45, 3 June 2016 (UTC)[reply]

Q3998906

This may be very basic, but I can't see how to fix the typo in the "label / eo" field (currently "Helraldika Triunuo", should be "Heraldika Triunuo"). I can only see how to edit the English, French, and Latin labels/descriptions... AnonMoos (talk) 03:38, 4 June 2016 (UTC)[reply]

@AnonMoos: There is a link "More languages" right under the frame on this item that lists your default set of language labels and descriptions. Alternately, you may activate gadget "labelLister" in your Preferences. -- LaddΩ chat ;) 04:57, 4 June 2016 (UTC)[reply]

P17 and military bases

We had some discussion some time ago about Embassies and country (P17). I do not remember what conclusion we came to, and I do not find the discussion. Here I have found another potential problematic area. A military base in Canada, which is operated (or at least has been operated by) United States of America (Q30). -- Innocent bystander (talk) 08:26, 4 June 2016 (UTC)[reply]

allegiance (P945)? There was this question and this more or less related discussion... -- LaddΩ chat ;) 13:07, 4 June 2016 (UTC)[reply]

I was sure I had contributed to a discussion on this but couldn't find it in the archive. Sure enough it was there in my contributions history, and it was correctly added to the archive by the bot when discussion was 7 days old, and nothing was deleted from the archive. So I did a bit more investigation and found that a single character typo was obscuring much of the archive. The discussion I was thinking of, and probably you too, is now visible at Wikidata:Project chat/Archive/2016/05#country (P17) of an embassy?. The suggestion that I made which Aude said they agreed with and that nobody else commented on was to use: country (P17) for the host country, operator (P137) for the country whose embassy it is as operating area (P2541) for the area that embassy covers. e.g

⟨ British Embassy Holy See ⟩ country (P17) ⟨ Italy (Q38)  

 ⟩
operator (P137) ⟨ United Kingdom (Q145)  

 ⟩
operating area (P2541) ⟨ Holy See (Q159583)  

 ⟩

.

I'm not sure whether operator (P137) or allegiance (P945) is better for military bases though? Maybe both if operator (P137) is used for the branch of the armed forces that operates the base, e.g.

⟨ RAF Gütersloh (Q278725)  

 ⟩ country (P17) ⟨ Germany (Q183)  

 ⟩
operator (P137) ⟨ Royal Air Force (Q165862)  

 ⟩
allegiance (P945) ⟨ United Kingdom (Q145)  

 ⟩

? Thryduulf (talk: local | en.wp | en.wikt) 21:49, 4 June 2016 (UTC)[reply]

Good find, the issue with the typo! Didn't know the indexing was so sensitive. "Operated by" sounds indeed as good as "allegiance" for embassies, but better for military bases, IMHO. -- LaddΩ chat ;) 23:07, 4 June 2016 (UTC)[reply]

Could you clarify that comment please - I can't decide whether you are saying operator (P137) is better than allegiance (P945) for military bases or the exact opposite! Thryduulf (talk: local | en.wp | en.wikt) 09:10, 5 June 2016 (UTC)[reply]

If "Royal Air Force" is the "operator" for an "military base". (That makes sense) Then it is probably the "Ministry for Foreign Affairs" that operates the embassy. -- Innocent bystander (talk) 09:18, 5 June 2016 (UTC)[reply]

Emergency phone numbers

I tried to add "112" as phone number (P1329) to 112 (Q1061257) but the edit filters didn't let me because of constraint violations. How should this be done? -- T.seppelt (talk) 09:17, 4 June 2016 (UTC)[reply]

click save twice.
--- Jura 09:24, 4 June 2016 (UTC)[reply]

Yes, the abuse filter only warns, and a second attempt should be successful. But to me it looks like this filter's format seems to be way to restrictive regarding telephone numbers internationally. --YMS (talk) 09:27, 4 June 2016 (UTC)[reply]

Well, most of the time, that's what's needed.
--- Jura 09:30, 4 June 2016 (UTC)[reply]

Okay, I misread the filter on a first quick look. Special:AbuseFilter/85 allows phone numbers to be in one of the following formats: "+(7|[2-689]\d{1,2})[\- ][\d(][ ()\.\-\d]+" and "+1[\- ]\d{3}[\- ]\d{3}[\- ]\d{4}", so as soon as there is a correct language code, probably everything is fine. And having those is of course a good idea except for very specific things like exactly T.seppelt's example. --YMS (talk) 09:56, 4 June 2016 (UTC)[reply]

The issue is that there not one area code for it. It's 112 across Europe and many different area codes.ChristianKl (talk) 12:50, 4 June 2016 (UTC)[reply]

A related (but not the same) issue is for telephone numbers that are national only, e.g. for Campaign Against Living Miserably (Q5027741) (a charity that only operates in the UK) - I believe the freephone number wont even connect when dialled from abroad (I don't know if the +44 800 format would work dialled from within the UK or not). Thryduulf (talk: local | en.wp | en.wikt) 20:40, 4 June 2016 (UTC)[reply]

In cases like this I would recommend to add the national prefix anyway. After all it is an UK number, whether or not it can be reached from outside. --Srittau (talk) 22:11, 6 June 2016 (UTC)[reply]

Is there a way to search for a specific value for a specific property?

Hi

Is there a way to search for a specific value for a specific property? I'm checking that all the Biosphere Reserves have been imported properly and I know a couple are missing, I have a couple that I suspect that are missing, is there a way to search by a value for a property? Specifically I would like to search for a value that is a URL of the Property UNESCO Biosphere URL.

Thanks

John Cummings (talk) 16:16, 4 June 2016 (UTC)[reply]

This is how to do it. --John Cummings (talk) 17:49, 4 June 2016 (UTC)[reply]

Add more than one geographic coordinate for Q3440914

Hi. The above is an item page for a dam in Sri Lanka. As you know, there are dams that are built to store water, and there are dams that are built for both: generate power and store water. Victoria Dam (Q3440914) is a hydroelectric dam (the latter type). So, I'd like to add the following to the dam item page:

Add all the statements that a dam item page would usually have
Add two geographic coordinates statements: one for the location of the dam itself, and other for the location of the power station.
Add a statement which says that this dam power/fuels the "Victoria Power Station" (new item page [power station] to be created).

Could someone help me with how I could add the below two statements to Q3440914, please? Or, is adding the second coordinate in the to-be-created "Victoria Power Station" item page the correct way to do this? If so, how do I do #3? Thanks in advance, Reh man 03:11, 5 June 2016 (UTC)[reply]

To make it even more complicated. The dam can in some languages be interpreted as the concrete-wall (or whatever building material) that prevents the water to flow away. Even dams made by non-humans are included here. -- Innocent bystander (talk) 07:36, 5 June 2016 (UTC)[reply]

The power station should have it's own item and the coordinates of it should only appear on that item. I suppose you could link the two with

⟨ Victoria Dam (Q3440914)  

 ⟩ has part(s) (P527) ⟨ Victoria power station ⟩

but there is probably a better way. Thryduulf (talk: local | en.wp | en.wikt) 09:17, 5 June 2016 (UTC)[reply]

Native language property (P103) and P1412

We have

which seems to be a little confusing and have redundancy. Why would we separate a first language from the other languages that a person knows? If we wish to identify a language primacy, then wouldn't it make more sense to have a qualifier label that could be added to the native language within P1413? When items are being added by users will generally type language and see what property appears, and for people it is not P103, it is always P1413. At a practical level, from my viewing, the native language is not highly used, so should we migrate to where the users are adding data, or are we going to be stuck in a system that it is not obvious, and not particularly useful, or used? — billinghurst sDrewth 04:17, 5 June 2016 (UTC)[reply]

I think you're assuming too much intentional action here. The one property is much older than the second, and could probably be deleted with the values replaced in the method you described. I imagine this just went under the radar, as opposed to being explicitly chosen at some stage. Ajraddatz (talk) 06:21, 5 June 2016 (UTC)[reply]

It's easier to determine P1412 than P103. Which is why the former is used.
The information can more easily be managed with two properties.
That suggestions don't work in an ideal way is a known issue. It's definitely not a reason to use another property, just because the most appropriate one doesn't get suggested.
--- Jura 07:15, 5 June 2016 (UTC)[reply]

Jura1: that argument doesn't obviously make sense. You say it is easier under one property, and the users have clearly demonstrated that with the usage, then you argue that the status quo should be maintained without reasoning. The information would seem to be able to gathered under one property, and yet when it is brought for discussion it is pretty well dismissed by you. If you are declaring that there is the need for two, and that users need to be struggling through the system to achieve an unobvious goal, then please demonstrate that to the users. — billinghurst sDrewth 09:05, 5 June 2016 (UTC)[reply]

P1412 had been populated mostly by a French library catalog. It includes languages in which people published.

If you know the native language of one of these authors, you can add it to P103. If you want to add the native language of a person, just enter it with P103. No need to bother with qualifiers. (go to the item, type "native lang" then "en" etc.
--- Jura 09:16, 5 June 2016 (UTC)[reply]

I can see the appeal of merging native language (P103) with languages spoken, written or signed (P1412) but I'm concerned that languages spoken, written or signed (P1412) is already too broad and vaguely defined. I couldn't tell you how many languages I speak or write (it could be anywhere from 1 to 50+ depending on what you want to count as speaking or writing a language). native language (P103) on the other hand has a much clearer meaning (I know what my native language is and, as much as I'd like to be natively multilingual, there is nothing I can do to change it :P). If we want native language (P103) to show up when someone searches for "language", that is easily done by adding an alias starting with "language". - Nikki (talk) 09:57, 5 June 2016 (UTC)[reply]

I do not understand the problem. My native language is Dutch, I am fluent in English, I can manage in German a bit in French. It is relevant to ascertain if a given publication COULD be written by me. Boxing me only in Dutch is not helpful. Often a mother tongue is of political value. Thanks, GerardM (talk) 04:05, 6 June 2016 (UTC)[reply]

Wikidata and maps

Currently https://query.wikidata.org/ offers a way to display maps. Which options currently exists to use coordinates from Wikidata to generate maps for Wikipedia?
--- Jura 09:26, 5 June 2016 (UTC)[reply]

@Lydia Pintscher (WMDE): would you give us a status? Llywelyn2000 is interested in this for various lists.
--- Jura 10:20, 7 June 2016 (UTC)[reply]

- I am actually not sure what people are already doing. You can't yet do a sparql query and put the results into an article. You can get the coordinate of an item and use that via Lua. Maybe one of the people more invovled with maps on Wikipedia know more? --Lydia Pintscher (WMDE) (talk) 13:55, 7 June 2016 (UTC)[reply]

- - Is a direct method planned for anytime soon? In the meantime, a Listeria list with coordinates might work.
    --- Jura 14:04, 7 June 2016 (UTC)[reply]

Members of parliament per constitutency

Hi!

I bring here an issue that affects Wikidata but the debate started in el:Project:Αγορά.

User:FocalPoint has created items like member of the Hellenic Parliament (Q24322738) as subclasses of member of the Hellenic Parliament (Q18915989) in order to use them in statements for position held (P39). For example:

⟨ Evangelos Diamantopoulos (Q12877312)  

 ⟩ position held (P39) ⟨ member of the Hellenic Parliament (Q24322738)  

 ⟩

It must be noted that the original position held is just "Member of the Hellenic Parliament" not "Member of Parliament from Kastoria". Per the Constitution of Greece, Article 51 "The Members of Parliament represent the Nation" not their electoral districts.

I have tried to explain that the most acceptable and useful approach is to use member of the Hellenic Parliament (Q18915989) with electoral district (P768) as a qualifier.

⟨ Evangelos Diamantopoulos (Q12877312)  

 ⟩ position held (P39) ⟨ member of the Hellenic Parliament (Q18915989)  

 ⟩
electoral district (P768) ⟨ electoral district of Kastoria (Q14623346)  

 ⟩

FocalPoint has explained that his approach is useful in order to display meaningful information in the infoboxes in Wikipedia, that "Member of Parliament from Kastoria" is something used in everyday speech.

I have tried to explain that what should be done is modify the infobox and not manipulate the data and create entities that do not represent an existing and distinct position. Speaking about a "Member of Parliament from Kastoria" any human understands that it is actually a "Member of the Hellenic Parliament from the electoral district of Kastoria" and not a distinct position. Machines do not understand that and we should not use an approach different than we use for any other parliament in the world.

What do you think? -Geraki (talk) 09:27, 5 June 2016 (UTC)[reply]

Well, we sometimes have elections in only one (or a few) electoral districts at the time, or for only some of the seats in the parliament. The second chamber of the Swedish parliament was such an example. It used to be elections every year, one electoral district at the time. And the EU-elections in every nation can have very unique rules in every nation. In UK, there is not the same rules in the Irish electoral district as in the rest of the nation. We maybe have to accept that we have subclasses like "Member of the X parliament 20AB-20CD, representing political party EFG and electoral district HIJ". But my opinion si that we still have to add such things as "political party", "timeframe", "electoral district" as qualifiers to such statements as position held (P39):"Members of Hellenic Parliament from Kastoria, 2015-2019 representing Liberal party". I am not proposing that we always should use items like "Members of Hellenic Parliament from Kastoria, 2015-2019 representing Liberal party", but that we should allow them, since they already are essential in some cases. -- Innocent bystander (talk) 09:53, 5 June 2016 (UTC)[reply]

I would like to point out that all members of Parliament of Greece (as well as the members of Parliament of Cyprus, following the contributions of User:Xaris333 yesterday), have as position held (position held (P39)) their respective seats characterized by the electoral district as described in several sources in Greek bibliography. This is clearly reflected in many articles of elwiki where the first thing noted is that the person is the (I use Kastoria for example) "Kastoria member of Parliament". Of course each of these parliament seats have the respective category, which has been duly linked in all such items, see (Olympia Teligioridou (Q9850817)). As elwiki uses infoboxes using info from Wikidata, all such infoboxes now indicate not just the general "member of Parliament", but the specific "Kastoria (or other district) member of Parliament) (see for example el:Περικλής Ηλιάδης). Nevertheless, this structure can certainly coexist with the structure indicated above with the qualifiers. May I also note that if any "Kastoria member of Parliament" dies, he is not replaced by just anybody, but (if available) by another "Kastoria member of Parliament". So, the position does exist, very clearly and with real consequences. --FocalPoint (talk) 10:35, 5 June 2016 (UTC)[reply]

I would also like to point out that we already have other items like: Q16707840, duly used in Joseph Hiley (Q6283995) and other MP's from the same constituency. I really cannot see any issue here. --FocalPoint (talk) 10:42, 5 June 2016 (UTC)[reply]

Q16707840 was, I believe, an experiment. It's not an approach used for any other UK MPs that I'm aware of, and (to be honest) removing it at some point was on my to-do list. Andrew Gray (talk) 11:55, 5 June 2016 (UTC)[reply]

My point is that since we can add qualifiers for there is no reason to create a complex network of items that no script can read and use if the developer does not know all of these items. "Members of Hellenic Parliament from Kastoria, 2015-2019 representing Liberal party" would be just a mix of qualifiers for the original political position of just being a member of the parliament. That would mean that if the representative resign before elections or change political group without resigning you would need even more subclasses with repeating qualifiers.

Again, speaking about a "Member of Parliament from Kastoria" a human understands that it is actually a "Member of the Hellenic Parliament from the electoral district of Kastoria" and not a distinct position. Ther are no distinct political bodies to be distinct political positions.

If the infoboxes need to display electoral district then the infobox should be scripted to display the qualifier.

The existence of some examples is countered by thousands of counterexamples using the qualifier, and e.g.

⟨ Barack Obama (Q76)  

 ⟩ position held (P39) ⟨ United States senator (Q13217683)  

 ⟩
electoral district (P768) ⟨ Illinois (Q1204)  

 ⟩

not "US senator for Illinois"

I have to note that you should stop for some time adding these statements (Special:Contributions/FocalPoint) ignoring the discussion. -Geraki (talk) 11:05, 5 June 2016 (UTC)[reply]

Excellent! I have no objection to the use of qualifiers as indicated above. As long as people care enough to fill the data in, this is the most important, to enrich Wikidata. Unfortunately, the members of the Greek Parliament do not have any such qualifiers up to now, but I welcome anyone who might decide to work on this, in addition to any other work done up to now.

As I have already indicated, to the best of my knowledge, all members of Parliament of Greece and Cyprus have now the items under discussion. --FocalPoint (talk) 11:10, 5 June 2016 (UTC)[reply]

@Geraki: no script can read and use if the developer does not know all of these items => this is wrong. Usually there is a pattern :

⟨ member of greek parliament for whatever ⟩ subclass of (P279) ⟨ member of greek parliament ⟩

and

⟨ member of greek parliements for whatever ⟩ electoral district (P768) ⟨ whatever ⟩

that any reasonable script should be aware of. Or maybe

⟨ member of greek parliements for whatever ⟩ has quality Search ⟨ https://www.wikidata.org/wiki/Q23766486 ⟩
electoral district (P768) ⟨ whatever ⟩

. A script that automatically adds labels can be coded from that.

Please also note that the of (P642)

qualifier is sometimes used instead of electoral district (P768)

.

author TomT0m / talk page 11:15, 5 June 2016 (UTC)[reply]

My feeling is that "member of parliament for X" is an unusually precise position and we should probably avoid using it in favour of a different approach. (As a note, I compiled Wikidata:EveryPolitician to try and index the main political properties, and created most of them...).

There isn't a Wikidata-wide standard for defining constituencies, yet, but the most common is to use a qualifier on P39. (The alternative is P768, but this is less commonly used). I'd love to see a consistent worldwide approach to this. Andrew Gray (talk) 11:53, 5 June 2016 (UTC)[reply]

I would also like to see the use of something like represents (P1268) instead of "political party" as qualifier to these kinds of statements. You do not (everywhere) have to be member of the political party you represent. -- Innocent bystander (talk) 12:04, 5 June 2016 (UTC)[reply]

As there is a discussion here, I see no point in mass reverts by Special:Contributions/Spiros790. I am therefore asking User:Spiros790 to stop. --FocalPoint (talk) 13:28, 5 June 2016 (UTC)[reply]

"Member of the Hellenic Parliament" is the accurate and correct term. Items created by FocalPoint (such as Q24322738) are not valid for P39. Other than that I strongly recommend that massive changes like these should be discussed before implementation in order to avoid errors & time-consuming reverts. User FocalPoint should note that discussion takes place before mass changes and edits like his, which are now presented as accomplished facts. Reverts to previous acceptable situation should proceed until this is discussed properly and conclusively. Thank you. -- Spiros790 (talk) 13:34, 5 June 2016 (UTC)[reply]

My best practice

Hoi, I have added thousands and thousands of members of whatever. The format is that they are for instance "Member of the Kerala Legislative Assembly" and I add often manually a constituency for them. It would be silly to have an item for "Member of the Kerala Legislative Assembly for Chadayamangalam" because you would add 140 more items that are not fixed in size nor in time. They are merged, changed in size and whatever. That is best left to the constituency itself. Thanks, GerardM (talk) 16:06, 5 June 2016 (UTC)[reply]

This does not really solves the problem of merging or change of size because we have items for the administrative divisions themselves. Plus the search for absolute consistency is vain since we have some items that have articles specific to some chairs or specific to election results for a specific circonscription. What I think would work, and would allows us to use different model for flexibility while not make harder data extraction is WikiProject Reasoning. author TomT0m / talk page 16:46, 5 June 2016 (UTC)[reply]

When a Wikipedia article does not fit in, tough. There are 280+ Wikipedias, we can not help their inconsistency. We do not need to accomodate a single article because it does not happily fit. Thanks GerardM (talk) 18:06, 5 June 2016 (UTC)[reply]

Some projects have an article about the "Nobel price", some have "Nobel Peace Price" and some even have an article about "2015 Nobel Peace Price". The solution in Tunisian National Dialogue Quartet (Q21078159) is Award:"Nobel Peace Price", time:"2015", article about:"2015 Nobel Peace Price". If you do not have an article about the Peace Price, but only about the Nobel Price that could cause some technical problems. And of course, we do not have items about every version of every price. We do not (yet) have an item for every electoral district in every city/municipality everywhere. In many cases, we do not even have articles about the "city/municipal councils".

We maybe in such cases have to add "Member of City counsil", of "City of Wheresburg", electoral districts:"Nothern suburbs", representing:"Revolutionary Conservative party", from:"2012" to:"2016"? -- Innocent bystander (talk) 18:10, 5 June 2016 (UTC)[reply]

In that case it's less work to create an item "Member of City counsil of Wheresburg" once and for all, maybe even less work to add an item for every elections , for example "Member of City counsil of Wheresburg 2012 - 2016". @GerardM: The only thing we need to do to support a mix of the models is to adapt our extraction queries, that should serve as references for our consumer. "If you want to extract our datas, just do a similar things than our query". The query in question, say if you want the elected of a year should be a mix of "you should check either the qualifier of the statement for the date, or if the value is any subclass of the city counceler of this country that have such a date in their statements and this city as mandate". This should make everyone happy and is not very complicated. author TomT0m / talk page 19:20, 5 June 2016 (UTC)[reply]

We should consider the consequences of such a disastrous idea through the eyes of 280+ languages. Having as few as possible items is better. The notion of unlimited items will make Wikidata useless for most of them. Thanks, GerardM (talk) 01:22, 6 June 2016 (UTC)[reply]

I think it is essential to use items like "Member of Hellenic parliament" instead of "Member of national parliament". It is another kind of job of working in the Parliament of Greece than in the Parliament of any other nation. But I am not so sure about the differences between the "City counsil of Wheresburg" and its neighbour "City counsil of Theresburg". It is most likely more or less the same kind of job.

Using years to those who were elected to First Chamber (Q10501497) (first chamber of the Swedish parliament) could cause some confusion. 1/8 of the chamber was elected every year for a period of 8 years. The set of "member of the first chamber of the Swedish parliament 1939-1947" is then for some time coexisting with those elected for 1941-1949 and 1935-1943. -- Innocent bystander (talk) 06:22, 6 June 2016 (UTC)[reply]

@GerardM: The counterpart of this is that fewer items means less specific items, so in fine it's harder to write complex statements. This mean you always have to qualify them when you write a statement, and in complex statements that mixes dates such as "the parliements members that were elected on 2000 have voted that law in 2002" we risk a qualifier conflict. On the other hand using plain statements in a new item makes the information available for autolabels and minimize redundancy and the risks to get this kind of conflicts is zero. author TomT0m / talk page 06:34, 6 June 2016 (UTC)[reply]

When you always use the same qualifiers it becomes easy. When someone was first elected on a date and ends his tenure at another date you will always be able to know who was in office on a given date. One person can have multiple statements when there is a break. It works well it is unambiguous. Having statements that are all over the place makes for queries that cannot be reused for a different parliament. Consequently your notion fails in the most obvious ways. Thanks, GerardM (talk) 19:11, 6 June 2016 (UTC)[reply]

Currency symbols

Hello everyone,

the question is how the relation between United States dollar (Q4917) and $ should be expressed. To me P558 (P558) seems to be the closest property we have at the moment. Unfortunately a lot of currencies appear at Wikidata:Database reports/Constraint violations/P558 because a currency cannot be converted into a SI base unit (Q223662). Should the constraints be changed or do we need a new property? -- T.seppelt (talk) 06:05, 6 June 2016 (UTC)[reply]

To answer the question T.Seepelt asked here. It is probably a better way to write the constraints here, other than add a lot of exceptions. But exceptions is probably doable in the meantime. We have many units that are unconvertible to SI-units. -- Innocent bystander (talk) 13:16, 6 June 2016 (UTC)[reply]

(On svwiki we have preferred to use ISO 4217 in our infoboxes instead of the unit.)`

The currency-units has one problem that most other units do not have. Sometimes they are written before the number and sometimes they are written after the number. Maybe we then have to include not only the symbol itself, but also the "pattern" it is written in. That probably demands a new property. -- Innocent bystander (talk) 07:01, 6 June 2016 (UTC)[reply]

We currently have currency symbol description (P489) which links to the item about the symbol (e.g. Euro (Q4916) → euro sign (Q18100); although P558 (P558) → €, Unicode character (P487) → €, defining formula (P2534) → \euro and 12 values for TeX string (P1993) are also used on Euro (Q4916)).

A "currency symbol position" property may be needed. I know of only three symbol positions leading (¤1.99), medial (1¤99) and trailing (1.99¤) and I know of no instance where the presence or absence of a space between the symbol and value matters, so an item value would seem doable, however some symbols e.g. € are used in multiple positions (generally how the symbol for the currency it replaced was used I suspect). There may of course be others I'm not aware of. Thryduulf (talk: local | en.wp | en.wikt) 11:08, 6 June 2016 (UTC)[reply]

It is possible to put qualifiers also to such a property, so different use of € is not a big problem. The median-version looks like a challenge to accomplish in the client. -- Innocent bystander (talk) 13:16, 6 June 2016 (UTC)[reply]

Simply change, for example, euro sign (Q18100) to "instance of"="currency prefix" and likewise for those which are "currency suffix". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:01, 6 June 2016 (UTC)[reply]

This highly depends on the language. In German € is a suffix (z.B. 3,50 €). Therefore instance of (P31) is not the ideal solution because we would have to use qualifiers on these claims. We need another system. Maybe even a "currency symbol position" property... -- T.seppelt (talk) 15:16, 6 June 2016 (UTC)[reply]

It also depends on currency - e.g. $ when used for United States dollar (Q4917) is a prefix, when used for Canadian dollar (Q1104069) $ is a suffix in Quebec (Q176) and a prefix elsewhere. When cifrão (Q1091860) is used for Brazilian real (Q173117) it is a prefix, when used for Cape Verdean escudo (Q4591) it is a suffix for whole amounts and a decimal separator for fractional amounts. Thryduulf (talk: local | en.wp | en.wikt) 22:14, 6 June 2016 (UTC)[reply]

cent/penny-symbols

Discovered that we also have "cent"-symbols in some currencies. -- Innocent bystander (talk) 08:00, 7 June 2016 (UTC)[reply]

Wikidata weekly summary #212

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

Events/Blogs/Press
- Upcoming: OpenData.ch
- Upcoming: HackHPI
- Upcoming: ELAG
- Documentation of the outcomes of WikiCite
- Wikidata showing chemical properties with references
Other Noteworthy Stuff
- SQID now also shows references for statements and received a whole bunch of other nice features
- wikidata-cli now has a wdsparql command to work with SPARQL from the command line
- It seems we already have almost 0.4% of the current population of Finland on Wikidata.
Did you know?
- Newest properties: exact match, reserve number (Canada), Shakeosphere person ID, mains voltage, HKMDb film ID, relegated, promoted, NIOSHTIC-2 ID
- Query examples: body of water with the most watercourses ending in it (source), French communes with names ending in ac (source), birthplaces of Europeana280 artists (source), popular eye colors (source)
Development
- Fixed the bug where the language selector in Special:NewItem doesn't recognize certain languages (phabricator:T115792)
- Fixed issues that prevented editing in Internet Explorer (phabricator:T136543)
- Added new visualization to the query service to visualize graphs like family trees. It is now deployed yet but will come soon.
- Worked on new visualization for the query service to visualize results with data on different dimensions
- Improved build and deploy process for the query service. The site should load faster now.
- Improved the formatting of dates and more in the query service.
- More fixes to get ArticlePlaceholder ready for the second round of wikis
- Worked on a fix for the issue where sometimes an old label is shown in the main header of an item (phabricator:T135714)
Monthly Tasks
- Hack on one of these.
- Help develop the next summary here!
- Contribute to a Showcase item
- Help translate or proofread pages in your own language!
- Add labels, in your own language(s), for the new properties listed above.

Read the full report · Unsubscribe · Lydia Pintscher (WMDE) 14:28, 6 June 2016 (UTC)[reply]

Qualifier

Hello. 1986–87 Cypriot Fourth Division (Q17436662) was split into three groups. We have three winners winner (P1346), one for each group. And we have teams that were promoted (P2881) from each group and teams that were relegated (P2882) from each group. Which qualifier can I use to show for example that Libanos Kormakiti (Q18228530) was the winner of Nicosia-Keryneia Group? (I don't want to have 3 different Wikidata pages since in Wikipedia we used to have only one article with all the groups and not to separate articles, one for each group). Xaris333 (talk) 17:30, 6 June 2016 (UTC)[reply]

@Xaris333: I guess a solution for more complex situation would be to be neutral in better/worse sport division and to have a unique "next league" property. We can know if the league is better ranked or not by knowing the relative rank of the league themselves. author TomT0m / talk page 17:52, 6 June 2016 (UTC)[reply]

@TomT0m: The groups were parallel. They didn't have any relation between them... "Next league"? Like followed by (P156)? I can't use it.Xaris333 (talk) 18:21, 6 June 2016 (UTC)[reply]

@Xaris333: That's why it's relevant. It works also if the leagues don't really have relations. I actually did not mean followed by (P156) but it might worth an examination, I meant something like "team that change league" instead of a pair of properties "promoted" qualified with the league. A qualifier on https://www.wikidata.org/wiki/Property:P1923 Search could also work and would minimize redundancy :

⟨ the league ⟩ participating team (P1923) ⟨ some team that actually won ⟩
next league Search ⟨ the upper league ⟩

.

I don't understand why you don't want to create items, they absolutely don't have to be linked to an article. it's a recipe for headeachs, really. Plus another wikipedia might do another choice, and we don't want our model to be one wikipedia centric otherwise this will be a mess here sooner or later. author TomT0m / talk page 19:19, 6 June 2016 (UTC)[reply]

Even the Cyprus Football Association referred to 1986–87 Cypriot Fourth Division (Q17436662) as a single championship, not 3 separates. I don't think there will be a Wikipedia that will create 3 different articles, one per group. But, OK, it is the only solution right now. I am not going to use it... Xaris333 (talk) 19:33, 6 June 2016 (UTC)[reply]

Swedish Football Division 6 (Q7654676) has ~80 groups. The groups can be reorganized from one year to another! A change of groups inside this tier is considered a matter of convenience (shorter travel within the group), not a matter relegation/promotion. -- Innocent bystander (talk) 07:54, 7 June 2016 (UTC)[reply]

Good luck to enter the results without the appropriate items :) @Xaris333: If nobody can help you with a good solution, I guess you'll have to go with it. If it's only a link problem, I guess it's related to WD:XLINK, you can add you usecase. In lua, you can make the infobox code generate the link to the appropriate article, for example. author TomT0m / talk page 15:38, 7 June 2016 (UTC)[reply]

@TomT0m: I just wanted to ask. If there is not a solution, its OK. Thanks. Xaris333 (talk) 17:38, 7 June 2016 (UTC)[reply]

Active links for commons categories

I am sure this must have been discussed, and I would appreciate if someone could point me out to a discussion or update on the status. In P373 statements, Commons categories do not appear for me as active links (they appear as text). See Q7921517 as one of the example (the top statement).

Is this a bug or a feature?
Could it be that this has smth to do with my configuration (I am on Windows 7, last version of FF)?

Thanks--Ymblanter (talk) 07:26, 7 June 2016 (UTC)[reply]

Sometimes they appear as links for me, sometimes just as text. In the example you supply, I do have a clickable link. I have no idea how the software (or ours) picks it out. I'd have thought it has something to do with loading times. Jared Preston (talk) 07:57, 7 June 2016 (UTC)[reply]

This problem has been going on for nearly a year now (see phab:T115794) and I haven't seen anything to suggest anyone is working on it. :( - Nikki (talk) 08:51, 7 June 2016 (UTC)[reply]

The JavaScript code is very glitchy and you sometimes need multiple reloads to have the links. (This also affects other JavaScript stuff on the page.) --Srittau (talk) 09:22, 7 June 2016 (UTC)[reply]

Relationship between `cellular response to vitamin D` and `vitamin D`

Currently Q14907166 has no statement that defines a relationship to Q175621. I would like to add one but at the moment I can't think of a Wikidata property that's applicable. ChristianKl (talk) 08:40, 7 June 2016 (UTC)[reply]

I added a relationship one level higher at response to vitamin D (Q14599705), has cause (P828) vitamin D (Q175621). Does that seem right? --99of9 (talk) 06:14, 8 June 2016 (UTC)[reply]

Which countries in 1754?

Notified participants of WikiProject Historical Place @Llywelyn2000:

How should we query countries in the past?

w:List of state leaders in 1754 has lists of entities in this years. It's part of a series for other each year. How would we find the same on Wikidata?
The general idea might be to query for "P31=country" and look for start and end dates within a range. For most years, I think this is unlikely to give anything useful.
Maybe a list of applicable items should be included on 1754 (Q7623) ?

I'm interested in more suggestions.
--- Jura 10:33, 7 June 2016 (UTC)[reply]

I would like to (again) rise the questions if such items as "Farfaraway during the Middle Ages" really should be regarded as countries? Swedish Empire (Q215443) has the statement: "instance of:former country", when I think it rather should have a statement like "instance of:aspect of history/of:Sweden/from:1611/to:1721". -- Innocent bystander (talk) 11:32, 7 June 2016 (UTC)[reply]

w:List of state leaders in 1754 has "Kingdom of Sweden (1721–1809)". Unfortunately, this redirects somewhere else and doesn't map to an item. Interestingly, Swedish Empire (Q215443) has P571/P576 which could be a substitute for the start/end dates mentioned. This actually does give a somewhat useful list [4] (even if the 171 items don't include Sweden). Maybe we just need to cover the gaps in that approach.
--- Jura 11:52, 7 June 2016 (UTC)[reply]

Historical data on Wikidata is quite lacking. There are only three listed heads of state/government froum 1754: query. --Yair rand (talk) 16:39, 7 June 2016 (UTC)[reply]
- You get 5 if you change >/< to >=/<= :)
  Obviously, without the basic structured, it's even more difficult to add data.
  --- Jura 18:20, 7 June 2016 (UTC)[reply]

Tiny datathon on lighthouses

We'll be arranging an event at the National Archives in Finland about lighthouses. For the purpose, I have been experimenting with Wikidata-driven infoboxes. However, it seems that some properties either do not exist, or may have been decided to be used through qualifiers. Here are some of my worries, how should I express them?

Lit date – with significant event (P793) and ___?
Date of automatization
Deactivation date – with significant event (P793) and ___?
Focal height – with height (P2048) applies to part (P518) <what is the item for the light of the lighthouse?>
Focal height from sea level – with elevation above sea level (P2044) similarly?
Light reach – Can vehicle range (P2073) be used to describe the reach of the light, and what would be the correct term?
Light source
Lens
Light intensity
Foundation – With made from material (P186) with applies to part (P518) foundation (Q191360)?
Construction – The same
Area:
- Gross area
- Floor area
- Apartment area
Fogsignal
Racon
Marking

Thanks for your help! /Susannaanas (talk) 12:46, 7 June 2016 (UTC)[reply]

automization should use
⟨ Example lighthouse ⟩ significant event (P793) ⟨ automation (Q184199)    ⟩
point in time (P585) ⟨ 1 September 1985 ⟩
Lit date could use significant event (P793) with an appropriate item (you don't need to discuss the creation of items if they don't exist) and point in time (P585) but maybe start time (P580) could be used?
Deactivation date similary could use significant event (P793) or end time (P582).
For the focal length properties, lamp (Q1138737), light bulb (Q1318740) or new items may work.
For light reach vehicle range (P2073) is not appropriate (that's how far a vehicle can travel without refuelling). I think a new property may be needed here "visible range" perhaps?
light source - source of energy (P618) is probably best.
Lens - probably simplest is applies to part (P518) with a specific item for the type of lens (again create as needed)?
We don't have a property I can find for intensity, but be clear what you mean when proposing one (w:Light intesity is a disambig page with half a dozen potential relevant meanings).
Foundation and construction material, yes I agree with your suggestions.
Area: area (P2046) with applies to part (P518) I think.
Fogsignal and racon - I don't know what these are. Is light characteristic (P1030) relevant?
Marking - inscription (P1684) might apply but I don't think that's quite what you are after, color (P462) perhaps?

I'm not a subject expert though so I may be missing some things. Thryduulf (talk: local | en.wp | en.wikt) 13:31, 7 June 2016 (UTC)[reply]

Thanks a lot, this got it going already!

I think a new property for vehicle range (P2073) could be created. I'll wait until after the event with this. I am also not a subject expert.
Light source / light - is different from source of energy (P618). I will be using that for electricity.
For light intensity I hope there will be suggestions.
I'll be using light characteristic (P1030), I will see if fogsignal and racon are included in that.
For marking, I will also see how it's used in WP to get an idea.

Cheers, Susannaanas (talk) 14:04, 7 June 2016 (UTC)[reply]

It seems that we have quite a lot of items about lighthouses, but not too many with properties beyond location and heritage status. If this is within the scope of your workshop, you might want to do a infobox mapping & import exercice.
--- Jura 14:12, 7 June 2016 (UTC)[reply]

Yes, it seems that's the thing to do, but I will run short of time... I will look into this, and see where I find the time :-) /Susannaanas (talk) 14:20, 7 June 2016 (UTC)[reply]

Should I start by making a version of the IB template? And create a project? /Susannaanas (talk) 14:27, 7 June 2016 (UTC)[reply]
- Sounds good. The advantage we had with films is that most properties had datatype item. It might be harder with quantities - partial support/improved support for that was added to HarvestTemplates (and QuickStatements) just recently. Help:Import Template:Bio from itwiki might be another format. I had used it for a fairly simple (follow-up) import project. Not sure what works best to get people started.
  --- Jura 14:43, 7 June 2016 (UTC)[reply]

New project started, WikiProject Lighthouses!! Worked half-way Finnish and English properties for lighthouses!! Help check the project table and join mapping!! /Susannaanas (talk) 17:22, 7 June 2016 (UTC)[reply]

TED speaker challenge

From 6 June to 6 July we are holding a writing challenge about TED speakers. Everyone can participate by writing about people who have held a TED talk. In January TED made a data donation with metadata about 2000+ TED talks. For the overview list of the TED speakers in the competition, see here: Wikidata:TED/TED speakers. For an overview of all the TED talks that these people have held, see here: Wikidata:TED/Ted Talks. Often the talks have multi-lingual subtitles, but you can also access the subtitle transcriptions directly with a link to quote the talks for use in Wikipedia articles in other languages than English. More information about the challenge, the points, the prizes, and the sign-up is here: TED speakers challenge. Jane023 (talk) 13:47, 7 June 2016 (UTC)[reply]

Just to add: in "writing about", we also include improving Wikidata items. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:33, 7 June 2016 (UTC)[reply]

Please merge

Q42117 and Q23925151. I guess few weeks ago someone created Q23925151 and messed up the English interwiki connection to it. Might be worth following up to see if that user hasn't messed anything else. I don't know enough about wikidata-fu to trace that, however. --Piotrus (talk) 03:08, 8 June 2016 (UTC)[reply]

Merged. --Edgars2007 (talk) 04:31, 8 June 2016 (UTC)[reply]

Looking for a great person to do the community communication around Wikidata

Hey everyone :)

Over the past 4 years - since the beginning of the development - I have been doing the community communications around Wikidata. The main objectives being that there is good communication between the editors, the development team and everyone else around Wikidata and that the project is a healthy and welcoming place. Since then Wikidata has grown massively and I have also taken on the product management for Wikidata. Both tasks are getting too big to be handled by one person while still giving editors and the development team the attention they deserve. Because of this I am now looking to hire someone who takes over the community communications part. I hope we can find an enthusiastic and knowledgeable person who can support you all.

The job add is here: https://wikimedia.de/wiki/Project_Manager_(m/f)_Community_Communication_Wikidata If you have any questions feel free to reach out to me. If you know someone who would be a good fit please do reach out to them and encourage them to apply.

PS: Don't worry. I'll still be around and available to you all but focusing more on the product management. Wikidata and you all are still very precious to me.

Cheers --Lydia Pintscher (WMDE) (talk) 07:50, 8 June 2016 (UTC)[reply]

Sad to hear, how about Innocent bystander?
--- Jura 08:07, 8 June 2016 (UTC)[reply]
Me, communication?! My wife cannot stop laughing! -- Innocent bystander (talk) 08:21, 8 June 2016 (UTC)[reply]
I think you'd to translate community talk into tech speak and the opposite, as you do most of the time .. not doing mindreading on her ;) -- Jura 08:28, 8 June 2016 (UTC)[reply]
@Lydia Pintscher (WMDE): Given the nature of the project, it would be a good idea to always list acceptable location of the potential employee in question... e.g. is it Germany-based only? Anywhere? and etc. --Izno (talk) 11:30, 8 June 2016 (UTC)[reply]
- I would prefer someone who can work in the office in Berlin but I am open to remote work. --Lydia Pintscher (WMDE) (talk) 12:48, 8 June 2016 (UTC)[reply]
Also, what (if any) is the level of German (and/or other language) fluency required for this position? Thryduulf (talk: local | en.wp | en.wikt) 11:52, 8 June 2016 (UTC)[reply]
- Only proficiency in English is needed. Any additional language is a big plus. --Lydia Pintscher (WMDE) (talk) 12:48, 8 June 2016 (UTC)[reply]

@@ Line 802: / Line 802: @@
 *::I think you'd to translate community talk into tech speak and the opposite, as you do most of the time .. not doing mindreading on her ;) -- [[User:Jura1|Jura]] 08:28, 8 June 2016 (UTC)
 *{{ping|Lydia Pintscher (WMDE)}} Given the nature of the project, it would be a good idea to always list acceptable location of the potential employee in question... e.g. is it Germany-based only? Anywhere? and etc. --[[User:Izno|Izno]] ([[User talk:Izno|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 11:30, 8 June 2016 (UTC)
+** I would prefer someone who can work in the office in Berlin but I am open to remote work. --[[User:Lydia Pintscher (WMDE)|Lydia Pintscher (WMDE)]] ([[User talk:Lydia Pintscher (WMDE)|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 12:48, 8 June 2016 (UTC)
 *Also, what (if any) is the level of German (and/or other language) fluency required for this position? [[user:Thryduulf|Thryduulf]] (talk: [[user talk:Thryduulf|local]] | [[w:user talk:Thryduulf|en.wp]] | [[wikt:user talk:Thryduulf|en.wikt]]) 11:52, 8 June 2016 (UTC)
+** Only proficiency in English is needed. Any additional language is a big plus. --[[User:Lydia Pintscher (WMDE)|Lydia Pintscher (WMDE)]] ([[User talk:Lydia Pintscher (WMDE)|<span class="signature-talk">{{int:Talkpagelinktext}}</span>]]) 12:48, 8 June 2016 (UTC)

Wikidata:Project chat: Difference between revisions

Revision as of 12:48, 8 June 2016

Research about user participation in Wikidata - call for participation (update)

Problems with IE

Wikidata Module:Taxobox

Arm

Importing Wiki Loves Monuments lists into Wikidata

Approaches

Issues

Importing directly from the WLM database into Wikidata

Import from the reference sources used to created the WLM databases

Discussion

Viking River Cruises

Sitelink to redirect - need opinions

Recap

Hide some bots from watchlist?

Linking to Wikidata from Meta to show tem names in specific languages

Data Extractaion

Pywikibot question

Verifiability policy

Wiki-related properties

some questions about Properties

Wikidata helpers needed in Edinburgh for Wikidata training event

Adding Language-labels for an Item

Historical maps properties

Medicines

How to add labels in sign languages ?

Link WP article with different name

Single value constraint for P856

Q3998906

P17 and military bases

Emergency phone numbers

Is there a way to search for a specific value for a specific property?

Add more than one geographic coordinate for Q3440914

Native language property (P103) and P1412

Wikidata and maps

Members of parliament per constitutency

My best practice

Currency symbols

cent/penny-symbols

Wikidata weekly summary #212

Qualifier

Active links for commons categories

Relationship between `cellular response to vitamin D` and `vitamin D`

Which countries in 1754?

Tiny datathon on lighthouses

TED speaker challenge

Please merge

Looking for a great person to do the community communication around Wikidata

Navigation menu

Search