User talk:Magnus Manske/Archive 9

all talk archives and subpages ^{user pages · m:special:CentralAuth · JIRA}

Welcome to Wikidata, Magnus Manske!

Wikidata is a free knowledge base that you can edit! It can be read and edited by humans and machines alike and you can go to any item page now and add to this ever-growing database!

Need some help getting started? Here are some pages you can familiarize yourself with:

• Introduction – An introduction to the project.
• Wikidata tours – Interactive tutorials to show you how Wikidata works.
• Community portal – The portal for community members.
• User options – including the 'Babel' extension, to set your language preferences.
• Contents – The main help page for editing and using the site.
• Project chat – Discussions about the project.
• Tools – A collection of user-developed tools to allow for easier completion of some tasks.

Please remember to sign your messages on talk pages by typing four tildes (~~~~); this will automatically insert your username and the date.

If you have any questions, don't hesitate to ask on Project chat. If you want to try out editing, you can use the sandbox to try. Once again, welcome, and I hope you quickly feel comfortable here, and become an active editor for Wikidata.

Best regards! --Guerillero | Talk 21:59, 2 January 2013 (UTC)

Adding Orbis Latinus to Mix'n'match

Hello,

thanks for your great tools for Wikidata! Would it be possible to add the online version of Orbis Latinus [1] to the catalogs in Mix'n'match? It would go really nicely with GeoNames and Getty for research purposes. Adam Harangozó (talk) 16:15, 6 August 2015 (UTC)

Listeriabot

I agree with Multichill, above. Love this bot. I also have two extra things I would like to be able to do with the lists:

• Possibility to number the rows of the list.
• Possibility tot sort names in a column to last name instead of first name

See for example this list, where I would like to have a list sorted by last name and numbered, so that I can see the number of Filipino senators in the blink of an eye, instead of having tot count the numbers of rows myself. Hope you can help out. Magalhães (talk) 06:00, 28 June 2015 (UTC)

 Support 1000 % :) --Hsarrazin (talk) 16:35, 29 July 2015 (UTC)

Both done, plus some qualifier options. Check the updated manual. Major refactoring took place under the hood, so watch for malfunctions in the next few days! --Magnus Manske (talk) 22:02, 1 August 2015 (UTC)

Manual update seems to have stopped working, see en:User:Spinster/List of works by Vincent van Gogh / here: Fatal error: Call to undefined method Listeria::getIremPropFirst() in /data/project/listeria/shared.inc on line 431

Something else. For statements linking to other statements like collection (P195) you put the whole list in the table, but you don't seem to do that for string based statements like catalog code (P528). Can you also output the full list for string based statements? Or maybe make it configurable in the columns? Multichill (talk) 13:54, 2 August 2015 (UTC)

Should both be fixed now. --Magnus Manske (talk) 14:06, 2 August 2015 (UTC)

Excellent. Thank you, Multichill (talk) 14:44, 2 August 2015 (UTC)

Adding the big ones like IMDB to Mix'n'match?

Hi Magnus, I was looking at IMDb ID (P345) and I wondered if you ever considered adding it to Mix'n'match? Other big identifiers would be VIAF ID (P214) or GeoNames ID (P1566). Multichill (talk) 17:28, 5 July 2015 (UTC)

I'd love to see Geonames! I've not used Mix'n'match so far because there's nothing on the list that I'm really interested in or know anything about, but linking to Geonames sounds like fun. :) Their data is already available in a tab-separated format (see allCountries.zip on http://download.geonames.org/export/dump/, the columns are described at the bottom of the page). I imagine we could generate a lot of suggested matches by looking for items with similar coordinates and a similar name, and preferably also a similar type (feature class, feature code). I've been considering mapping http://www.geonames.org/export/codes.html to Wikidata items but I haven't really had a reason to start yet. If it would be useful for that, do let me know. - Nikki (talk) 23:32, 5 July 2015 (UTC)

@Multichill, Nikki: See Wikidata:Bot requests/Archive/2014/10#Done. These databases are too large.--GZWDer (talk) 18:42, 16 July 2015 (UTC)

So how big is acceptable? TGN has 1.2 million entries and several others have over 100,000. If the entire dataset is too big on its own, it wouldn't be difficult to split it into smaller datasets and I'm quite happy to do that myself if someone tells me how small they need to be. - Nikki (talk) 21:05, 16 July 2015 (UTC)

Reasonator ISNI

Hey, ISNI link is fine from Wikidata, but it is wrong from Reasonator. --JulesWinnfield-hu (talk) 11:10, 11 July 2015 (UTC)

Some values are preprocessed [2]. --JulesWinnfield-hu (talk) 22:32, 13 July 2015 (UTC)

Duplicity

Hi :) Duplicity is really cool! I've been using Articles without item but always found it quite slow to use, because it wasn't very easy to compare things and Duplicity makes that so much better. I do have a few requests and questions though:

• Could you make the buttons always appear in the same place? Since they're displayed after the Wikipedia page title, they move about depending on how long the page title is, which is really bad for muscle memory. Fixed before I could even finish my comment. :)
• Could you exclude English pages with en:Template:Wiktionary redirect? Those are soft redirects to Wiktionary and aren't notable in Wikidata, so shouldn't be linked to items. Similarly, the German templates de:Vorlage:Obsolete Schreibung and de:Vorlage:Falschschreibung are soft redirects to other items that shouldn't be linked to Wikidata items.

I am already excluding several of these, at update time (once per hour), though through categories rather than templates. Added "Redirects to Wiktionary".

• An option to automatically skip pages with no suggestions would be useful. Adding new pages is more work than confirming that something matches, so if you're feeling lazy and are in the mood for something easy, only looking at things with suggestions would be nice.

No suggestions can come easily for non-English words when it's not a biography (title=name). I did put in the search box for a reason, so one could try more likely terms. Also, getting suggestions is "expensive", as it is using Wikidata search, so skipping could be slow.

• On a similar note, perhaps there should be a way to say that the suggestion is not a match, without having to find the right match or create a new item? While trying it out I skipped several where I could tell that the suggestion was wrong but didn't feel like I knew enough to say what would be correct.

Just skip it; action is not mandatory ;-)

• An easier way to say that something is a disambiguation page would be nice, e.g. a button which would create an item with "Wikimedia disambiguation page" as the description and a instance of (P31) Wikimedia disambiguation page (Q4167410) statement. Disambiguation pages seem to show up fairly often in Articles without item and I already got one in Duplicity too. Also fixed before I could finish my comment. Works nicely!
• Sometimes the interface goes weird and really narrow, and then when I press skip, it goes back to https://tools.wmflabs.org/wikidata-todo/duplicity.php? Looking at the console, it shows the error duplicity.php?wiki=enwiki:53 Uncaught SyntaxError: Unexpected identifier. The line at that point is `var title = '2005 Greenlandic Men's Football Championship' ;` - it looks like the code needs to escape apostrophes.

Should be fixed now.

• Are you planning to add support for more languages? How about namespaces? (templates and modules in particular are often copied between projects)

- Nikki (talk) 11:28, 12 July 2015 (UTC)

Will add more languages on demand. Will look into templates/modules/categories, but I feel they'd "spam" the list I have now. --Magnus Manske (talk) 11:41, 12 July 2015 (UTC)

Oh, Magnus, please add ruwiki. --Infovarius (talk) 21:18, 13 July 2015 (UTC)

Quickstate / unk

Is there a way to set "point in time" to "unknown" when adding numbers with QuickState? --- Jura 15:51, 13 July 2015 (UTC)

No. --Magnus Manske (talk) 15:53, 13 July 2015 (UTC)

Ok. Just realized that I can't add actual dates either (#QuickStatements_:_quantity). --- Jura 16:11, 13 July 2015 (UTC)

Multi-beacon

Hi Magnus,

I tried to do a look-up on P646 for "/m/015f0p" (and a series of others). Any idea why it doesn't map to Q623333 ? --- Jura 21:39, 13 July 2015 (UTC)

/m/021c5h /m/02l8p9 /m/0gd1z /m/01djxm /m/0cd4d

A few more above. --- Jura 06:55, 14 July 2015 (UTC)

I used the other Beacon module instead. --- Jura 19:01, 14 July 2015 (UTC)

Location in Reasonator

For information, it seems that there is an issue with locations displayed in Reasonator: a public artwork in Wahsington D.C. could be located in Australia https://tools.wmflabs.org/reasonator/?q=Q16828178 , even if data don't seem to be wrong. The issue on Reasonator is certainly known and more difficult to be resolved than I can think; there are so many different cases. I hope a solution could be found. Best regards --Shonagon (talk) 17:49, 15 July 2015 (UTC)

Arabic terminator

Hi Magnus, I was playing around with Terminator to show Emnamizouni around and we noticed Arabic is missing. Could you please add it? Multichill (talk) 19:53, 18 July 2015 (UTC)

Added language code "ar", should show after next update. --Magnus Manske (talk) 23:18, 19 July 2015 (UTC)

Catscan and autolist

Hi, sorry to pester you with low-content remarks, but I must say that to me the ultimate maintenance tool would be an autlolist option in catscan (I mean: the corresponding item matches some query). --Zolo (talk) 17:08, 19 July 2015 (UTC)

As "Get the corresponding Wikidata item for the individual page" on http://tools.wmflabs.org/quick-intersection/index.php ? --- Jura 17:39, 19 July 2015 (UTC)

Second. And it is better if this can also create missing items. However Catscan2 in not finished ([3])--GZWDer (talk) 05:30, 20 July 2015 (UTC)

FWIW, I am working on something that would solve these issues and more. Will take some time though. --Magnus Manske (talk) 11:43, 20 July 2015 (UTC)

SourcererBot

I think your bot is not working correctly here. It says "International Standard Name Identifier", but is using GND ID (P227). Sjoerd de Bruin (talk) 12:52, 21 July 2015 (UTC)

Thanks, fixed now. Running again to add proper values, but not sure how to remove the bad ones. Looks complicated... --Magnus Manske (talk) 13:44, 21 July 2015 (UTC)

I see the bot still adding strange qualifiers (already left a message on the talk page of the bot). Mbch331 (talk) 14:28, 21 July 2015 (UTC)

I think you still got the broken one (edit was before my reply here). Seems to work (here too). --Magnus Manske (talk) 15:18, 21 July 2015 (UTC)

problem with Sourcerer

Hello Magnus,

I can't have Sourcerer work properly on items. When I click on the link, I have "Checking external links... to go", then nothing more...

Can you fix it, or is this tool dead ?

Thanks for your help --Hsarrazin (talk) 18:26, 21 July 2015 (UTC)

Try it now. --Magnus Manske (talk) 19:55, 21 July 2015 (UTC)

sorry, no change :( --Hsarrazin (talk) 17:37, 22 July 2015 (UTC)

Listeria question

Concerning Listeria, that I begin to test more seriously, I'd like to know how to sort out items where the value of a claim has been deprecated, so that they don't appear in the list afterwards.

Is there a way to do it ? --Hsarrazin (talk) 17:39, 22 July 2015 (UTC)

Not yet. There are not many deprecated statements. I'll refactor listeria sometime. --Magnus Manske (talk) 19:23, 22 July 2015 (UTC)

there will be more and more, as I use it to add precise dates for French writers see Abad, top of list here ;)

Alphabetic sorting is my second problem, but can't be solved from wikidata, afaik :)

thanks for this very useful tool. --Hsarrazin (talk) 20:04, 22 July 2015 (UTC)

 Support 1000 % :) --Hsarrazin (talk) 16:35, 29 July 2015 (UTC)

Duplicity tool

According to the intro of the tool article in bad categories are skipped. However I regularly come across articles that have been nominated for deletion on nlwiki. For example nl:Léon van den Haute it was nominated for deletion on July 8th 2015 and it was suggested to me just now. What is the reason the article was suggested? Mbch331 (talk) 09:38, 25 July 2015 (UTC)

Because I currently exclude "bad categories" from de and en? --Magnus Manske (talk) 12:47, 25 July 2015 (UTC)

That explains a lot. Didn't know that part, because the tool doesn't say you only exclude bad categories from de and en. They way it's mentioned at the tool it looks like you exclude bad categories from all wikis. Mbch331 (talk) 12:52, 25 July 2015 (UTC)

Catalogus Professorum Halensis

Hallo Magnus! Catalogus Professorum Halensis ID (P2005) ist jetzt vorhanden. Kannst du die Matches nach Wikidata übertragen? Jonathan Groß (talk) 06:43, 27 July 2015 (UTC)

Läuft! --Magnus Manske (talk) 10:22, 27 July 2015 (UTC)

Erzeuge auch mal eben "sub items" für alle, die noch kein Item haben. --Magnus Manske (talk) 10:25, 27 July 2015 (UTC)

Heißt das, items ohne Wikipedia-Artikel? Kannst du dann auch gleich eine Rotlinkliste für dewiki anlegen? Jonathan Groß (talk) 10:27, 27 July 2015 (UTC)

Keine Rotlinkliste (muss noch 'nen Tool schreiben!), aber ich kann die komplette Liste anbieten, mit Rotlinks kursiv. --Magnus Manske (talk) 11:32, 27 July 2015 (UTC)

@Jonathan Groß: Jetzt auch mit redlinks! --Magnus Manske (talk) 13:00, 4 August 2015 (UTC)

Danke dir (: Jonathan Groß (talk) 15:19, 4 August 2015 (UTC)

Mitglieder der Heidelberger Akademie der Wissenschaften

Hallo Magnus, ich habe schon wieder was auf der Pfanne: Eine Liste der Mitglieder der Heidelberg Academy for Sciences and Humanities (Q833738) inklusive Angaben zu Art und Startzeit der Mitgliedschaft, Klassenzugehörigkeit und Sterbedaten. Hier ist die Originaldatenbank, ich hab aber schon alles rausgezogen, siehe hier. Wär das auch was für Mix'n'Match? Jonathan Groß (talk) 10:43, 27 July 2015 (UTC)

Klar! Warte auf Deine Genehmigung bei Google Docs. Mann, sind wir Deutsch! ;-) --Magnus Manske (talk) 11:54, 27 July 2015 (UTC)

Ich dachte, der Link genügt. Genehmigung ist jetzt raus :) Jonathan Groß (talk) 14:15, 27 July 2015 (UTC)

Ist importiert. Habe Deine Kalenderdaten nicht verwendet, die scheinen mit den Personen nix zu tun zu haben... --Magnus Manske (talk) 14:38, 27 July 2015 (UTC)

Die Kalenderdaten waren Sterbetage. Wohin hast du die Daten importiert? Direkt auf die Wikidata-Items? Dann fehlen aber noch weitere Qualifikatoren. Jonathan Groß (talk) 23:13, 27 July 2015 (UTC)

Importiert in Mix'n'match. Das Tool ist gut, um matches zu finden, nicht so sehr als Metadaten-Transport geeignet. --Magnus Manske (talk) 14:59, 28 July 2015 (UTC)

Die Statements müssten sich ja auch mit QuickStatements hinzufügen lassen, sobald das Matching fertig ist. Jonathan Groß (talk) 18:39, 28 July 2015 (UTC)

Übrigens, der Eintrag zu Géza Alföldy (Q80258) ist irgendwie komisch. Scheint zweimal vorhanden zu sein. Ich habe einen als N/A markiert. Jonathan Groß (talk) 20:09, 28 July 2015 (UTC)

Ich stelle mir die Einträge am Ende so vor wie bei Ekkehard Ramm (Q20742624): Detaillierte Angaben zur Mitgliedschaft, die sich aus der Tabelle und der Match-Liste schöpfen lassen. Jonathan Groß (talk) 11:43, 31 July 2015 (UTC)

Qualifiers in Listeria

Hi Magnus :) Is there a way to list a qualifier's value regardless of the property's value? On User:Nikki/P553 I was trying to make a list of the remaining uses of website account on (P553) and wanted to display the qualifier website username or ID (P554) so that I could see what sort of values are being used for the qualifier. I tried "p553//p554" and "p553/p554" but neither of those work. - Nikki (talk) 16:50, 31 July 2015 (UTC)

You can not just do property/qualifier_prop; you need to specify which item you want the qualifier for. So P553/Q866/P554 would give you the YouTube membership etc. Check the docs. Not sure this is ideal, but that's the way it is now. --Magnus Manske (talk) 18:27, 31 July 2015 (UTC)

@Nikki: Update: New option. See Template:Wikidata list, and User:Magnus Manske/listeria test2 for an example. --Magnus Manske (talk) 21:41, 1 August 2015 (UTC)

Thanks for adding it (and so quickly)! The "item" and "label" columns seem to have broken for lists here on Wikidata though, it's producing empty table cells (see User:Nikki/P553 again). - Nikki (talk) 00:03, 2 August 2015 (UTC)

Sounds good. I tried to apply it at User:Jura1/name days, but that didn't quite work out. That list is a bit tricky anyways. --- Jura 07:40, 2 August 2015 (UTC)

label issue fixed. --Magnus Manske (talk) 14:19, 2 August 2015 (UTC)

Arbitrary access on enwiki

I saw your question on Twitter and it doesn't seem to have any replies, so... according to phab:T100788, arbitrary access for enwiki is tentatively scheduled for the 18th. - Nikki (talk) 10:29, 2 August 2015 (UTC)

Uncorrect list on "people without gender"

Hello Magnus,

Tonight, for a change, I tried http://tools.wmflabs.org/wikidata-todo/?action=nogender

And the list I got was composed only of :

• Category:People from Delaware
• Category:People from São Paulo (state)
• Category:People from West Bengal
• Category:People from Tennessee
• Category:People from South Australia
• Category:People from Kerala

etc.

I don't think it is possible to assign a gender to these items, that are certainly not Q5 :)

Wanted to let you know, so that you can correct the request for the tool :) --Hsarrazin (talk) 21:08, 2 August 2015 (UTC)

Wow, that is really from the olden days of Wikidata :-) Fixed it (kinda), but a better way is AutoList. Should probably take this tool offline, too ancient. --Magnus Manske (talk) 10:19, 3 August 2015 (UTC)

Yup, but the random list was nice, while Autolist always gives the list sorted on Qid, which leads to always getting the same items, impossible to determine, because of language ;)

also, I would not consider those as Q5 either ;)

   olfactory receptor, family 7, subfamily E, member 46 pseudogene
   adenosylhomocysteinase pseudogene 3
   ribosomal protein L22 pseudogene 11
   heterogeneous nuclear ribonucleoprotein A1 pseudogene 39
   plakophilin 4 pseudogene 1
   von Willebrand factor C domain containing protein 2-like
   ribosomal protein L23 pseudogene 5

D --Hsarrazin (talk) 17:52, 5 August 2015 (UTC)

Well, when I say Autolist is better, I have my reasons. You choose to drive a Tin Lizzy, you get fail. Consider using Autolist to generate a PagePile, you can then filter it to make a random subset. Or several. --Magnus Manske (talk) 18:30, 5 August 2015 (UTC)

--Hsarrazin (talk) 17:52, 5 August 2015 (UTC)

Magnus Manske, I can assure you, I use Autolist, a lot… :)

that was just changing tool, for fun ;) If you tell me it's obsolete, I won't use it.

I'll have a try at PagePile, once I understand what it does :) --Hsarrazin (talk) 20:30, 5 August 2015 (UTC)

Not in the other language

Only Wikipedia is available. Please improve.--GZWDer (talk) 06:23, 3 August 2015 (UTC)

Any project in particular? --Magnus Manske (talk) 10:19, 3 August 2015 (UTC)

e.g. Wikivoyage. Others may also be supported.--GZWDer (talk) 01:50, 4 August 2015 (UTC)

Added a few. --Magnus Manske (talk) 12:26, 4 August 2015 (UTC)

authority_control.js, a missing dependency

Hi,

Thank you for the authority_control.js script. I tried to use it, but my browser failed to load the script, while sending this error :

Error: TypeError: $(...).dialog is not a function

Line: 66

The message hinted a missing library. I found that if I use the script wikidata useful.js, it works. Strange.

Ltrlg thinks [4] that the line mw.loader.using( 'jquery.ui.dialog', ... ) is missing in authority_control.js.

Regards,

Cantons-de-l'Est (talk) 18:01, 3 August 2015 (UTC)

There may be some race condition. Try reloading, with of without force-reload. Once it "sticks", it should work. --Magnus Manske (talk) 18:07, 3 August 2015 (UTC)

Sync Wikidata with dump

Could you add the new dump to WDQ when it comes available? WDQ still doesn't likes merged items... Sjoerd de Bruin (talk) 18:18, 3 August 2015 (UTC)

Latest JSON dump has bugs, they're working on it. Next one. --Magnus Manske (talk) 12:16, 4 August 2015 (UTC)

GeneaWiki

The "Discuss" link on GeneaWiki doesn't appear to lead anywhere, so I'll leave some suggestions here. Hope that's okay.

• The tool is great at generating genealogical trees, but it would also be useful to be able to generate other kinds of trees using different properties (other than P40/P25/P22).
• The canvas width seems to be locked to 1024. Any reason it couldn't be 100% width?
• Having languages other than English as an option would also be useful.

Thanks for the excellent tool, by the way. --Yair rand (talk) 22:16, 6 August 2015 (UTC)

signature (P109) in WDFIST

To add these, it would be helpful if SVG files were displayed. I had forgotten how cumbersome it is to add images by hand ;) --- Jura 17:14, 9 August 2015 (UTC)

People Australia identifier (P1315) in mix'n'match

Hi Magnus, could you add NLA Trove people ID (P1315) and maybe Libraries Australia ID (P409) to mix'n'match? See for example the record for Suharto (Q44819): http://trove.nla.gov.au/people/449040 . It contains both values. It's hosted by trove and they have an api. Never used it, but you should probably be able to pull the info out of it. @Wittylama: used to work there so he might have more background info.

My first usage will be for painters. We're trying to add an authority control link to every painter (who has a painting on Wikidata). Most general well know painters are in Union List of Artist Names ID (P245), RKDartists ID (P650) or both. For the bit more obscure American artists Smithsonian American Art Museum person/institution ID (P1795) seems to cover it. I recently uploaded several Australian collections and I hope NLA Trove people ID (P1315) will cover the locals there. Multichill (talk) 20:00, 9 August 2015 (UTC)

Trying to import >1M people for NLA Trove people ID (P1315)... --Magnus Manske (talk) 13:03, 10 August 2015 (UTC)