[08:10:46] Lydia_WMDE: http://tools.wmflabs.org/audetools/wikidata-stats/usage-tracking/2015-11-09:) [08:10:54] http://tools.wmflabs.org/audetools/wikidata-stats/usage-tracking/2015-11-09 :) [08:11:13] or https works https://tools.wmflabs.org/audetools/wikidata-stats/usage-tracking/2015-11-09 [09:51:00] aude: yay! [09:52:17] aude: thanks! :) [11:11:22] Lydia_WMDE: https://phabricator.wikimedia.org/T54385 [11:17:08] Is it possible to view the wikidata code? [11:19:26] Thiemo_WMDE: https://scrutinizer-ci.com/g/wikimedia/mediawiki-extensions-Wikibase/ [11:20:05] Funny how the worst rated class is a test class [11:20:24] (not counting WikibaseRepo, as such tools tend to falsly identify top level factories as bad) [11:20:31] It's a benestar ! [11:20:46] hey :D [11:21:04] benestar: check my awesome new library! http://www.bn2vs.com/blog/2015/11/08/rewindable-php-generators/ [11:21:13] It has all of 1 adapters [11:21:30] saw that already, quite cool (y) [11:21:43] ugh, ok then [11:21:51] benestar: check my other awesome new library http://www.bn2vs.com/blog/2015/11/08/wikidata-wikibase-json-dump-reader/ [11:21:52] :D [11:22:03] though a bit weird that PHP has that strange behaviour [11:22:20] benestar: want weird? try using a generator on HHVM [11:22:32] for the dump reader, I wonder if you'd encounter memory issues in PHP? [11:22:34] It "works", though it behaves different than on PHP >_> [11:22:50] benestar: what kind of memory issues would you encounter> [11:23:16] I think I read somewhere that PHP has problems with big data [11:23:30] but perhaps that also depends on your machine and php configuration [11:23:49] benestar: the iterators are lazy [11:23:54] you only ever load a single entity [11:24:01] of course, what you do with it then is up to you [11:24:20] hah, I see. So you don't load the whole dump at once? [11:24:25] just iterating through it should have very low meory usage [11:24:32] benestar: wtf ofc not [11:24:38] it's 55GB [11:24:38] does it support lookup by id? [11:24:54] I think wikidatatoolkit does load everything at once [11:24:56] benestar: it reads the dump, it does not build any sort of index [11:25:03] ok [11:25:03] sure [11:25:26] I think it makes more sense to not put this in the same lib, so that if you only need the reading part, you dont also have to pull in the other stuff [11:25:38] sure, but it would be a nice extension [11:26:04] benestar: seen https://github.com/JeroenDeDauw/EntityStore and https://github.com/JeroenDeDauw/TermStore ? ;p [11:26:18] only at a glance [11:26:22] Those are very basic [11:40:29] oh this is an amusing read https://www.reddit.com/r/lolphp/comments/3403lu/author_of_php7s_ironic_declarestrict_types1_i/ [11:40:31] benestar: ^ [11:45:50] One of the difficult items: https://www.wikidata.org/wiki/Q690164 <- Anyone has ever looked into modelling administrative law? [11:47:58] JeroenDeDauw, I wish I would get the PHP jokes, but I have never used it :) [11:50:04] tobias47n9e: what I linked is a bunch of non-php devs shitting on php :) [11:50:25] sad thing is that a lot of it is kinda true [11:51:55] JeroenDeDauw, It is usally fun to read, but bad to get involved :) [11:52:37] tobias47n9e: what's the deal with those numbers in your irc handle? Looks even more random than my domain name [11:53:30] 47°N 9°E, the first digits of my home location :) Inside joke for earth scientists :) [11:54:06] At some point I should change my wiki-username to those digits. More creative than tobias1984 [12:15:44] <_joe_> anyone working on wikibase: there is an UBN! ticket that you might want to keep in consideration here https://phabricator.wikimedia.org/T118162 [12:44:30] _joe_: I just added some links to some other tickets to it! [12:56:25] hello - anyone know how i can change the twitter handle on a script and pull the following data: https://www.wikidata.org/wiki/Q34086 [13:00:14] lseactuary, Sentence does not make sense grammatically. Can you try to ask your question again? [13:00:19] sure [13:00:29] so the link above provides all data wiki has on justin beiber [13:00:39] i want to now see what data wiki has on david beckham [13:00:45] britney spears [13:00:46] etc [13:00:48] how do i do this [13:01:34] Are you just interested in all people? I don't see how those are connected? [13:03:50] lseactuary, This is for example a query for all people: http://tools.wmflabs.org/autolist/autolist1.html?q=CLAIM[31%3A5] [13:04:04] Almost 3 million [13:04:26] i have a bunch of twitter handles [13:04:37] i want to pull the data for each of those handles to see what is available [13:05:43] lseactuary, "pull"? with what? Pywikibot? Are you working with the web-API? [13:06:02] so for example for justin beiber [13:06:17] i know one statement is genre and there are 2 types [13:06:23] i want to see similar data for other handles [13:09:32] lseactuary, Did you read this page? [13:09:47] which? http://tools.wmflabs.org/autolist/autolist1.html?q=CLAIM[31%3A5] [13:10:01] https://www.mediawiki.org/wiki/API:Web_APIs_hub [13:10:46] And then I would read this: https://www.mediawiki.org/wiki/API:Presenting_Wikidata_knowledge [13:14:41] lseactuary, And play around with this: https://www.wikidata.org/wiki/Special:ApiSandbox [13:15:03] thanks! [13:29:41] Now I am also intereted in the API. How can I get a claim without knowing the ID: https://www.wikidata.org/wiki/Special:ApiSandbox#action=wbgetclaims&format=json&property=2002&claim=victoriabeckham&rank=normal [13:30:39] I mean: How can I get a valid return containing the item if I just know what claim to look for. [13:31:04] twitter user name = "victoriabeckham", ID = ? [14:07:55] tobias47n9e: use the query engine! [14:08:01] http://query.wikidata.org [14:08:21] addshore, Not possible using the API? [14:08:25] nope [14:08:32] it is possible using the query api though.. [14:09:37] addshore, The question from farther above (14:04) was interested in iterating over known twitter names and getting the data. [14:10:06] yeh, you should be able to do a query to get all known twitter names, and then do stuff with them [14:10:36] addshore, So that will probably not be added to the API? And if not, is there a reason for that? [14:10:55] Why should it be in 'the' api when it is already in an api? ;) [14:11:22] its just at a different URL / endpoint! [14:11:50] addshore, Oh. So it is just not in the API:Sandbox? [14:12:23] https://lists.wikimedia.org/pipermail/wikidata/2015-November/007558.html might be useful if you already have a specific list of ids to look up [14:13:24] well, twitter names, not ids [14:13:28] nikki, Yeah. I too, have to do more reading about this. I have almost no web-programming experience. [14:14:58] tobias47n9e: indeed, that sandbox is only for the Mediawiki API [14:15:34] tobias47n9e: you can just use http://query.wikimedia.org as a sandbox directly as there is no writes, only reads! :) [14:16:54] addshore, Thanks! Now if only lseactuary would stay on a little longer he would so much about Justin Bieber :/ ;) [14:17:06] *he would learn* [14:32:31] can i be banned if i request data for ~150k profiles? [14:33:35] lseactuary: profiles? [14:33:52] so, the query api can spit out fairly big lists [14:34:01] but is there a ban if you ask for too much data [14:34:20] your best bet for getting the entities then would be to use one of the JSON dumps, rather than tens of thousands of api requetss [14:34:53] lseactuary: you should read https://www.mediawiki.org/wiki/API:Etiquette ;) [14:35:12] :) [14:37:22] lseactuary, And if you disconnect after each answer you will miss some interesting comments ! [14:38:12] :D [14:39:23] lseactuary, addshore for example recommended using http://query.wikidata.org for your query. [14:39:57] thx [15:36:35] https://tools.wmflabs.org/mix-n-match/ down? [16:00:58] https://www.wikidata.org/wiki/Wikidata:Wikimania_2016 <-- suggestions urgently needed ~14 days left for comments [16:16:50] lseactuary seems to have found the mailing list. The search for twitter user name data continues :) [16:18:50] I guess he is not going to stop until someone write the query for him and probably still connects it to his website. [16:55:36] :P [16:58:23] Someone has written ther query. Now he needs an entry-level Python tutor :) [16:58:57] I am taking bets if he will get an answer :D [17:00:42] https://lists.wikimedia.org/pipermail/wikidata/2015-November/007571.html [17:10:06] tobias47n9e the query is failry easy [17:12:36] addshore, Yes. But now is asking to get "_" between the words. He is in too deep. [17:13:05] between the words? :P [17:13:35] *goes to read the mailing list* [17:50:16] addshore: never mind the mailing list :) care to review Cenarium's auto-tag patch? https://gerrit.wikimedia.org/r/#/c/194458/ [17:50:18] It would be really good to have something like this, but I can't really tell if the code does what it should... [17:50:31] You have worked with the RC stuff recently, so maybe you have an oppinion [18:10:53] DanielK_WMDE: but I can't really tell if the code does what it should... [18:11:04] hah, that is encouraging :D I'll try and take a look in a bit! [18:11:10] :P [18:11:19] is there just 1 path now? [18:11:22] *patch... [18:11:42] oh, no, a nice chain :P [18:12:34] I'm going to be working later on [18:12:45] so if there's anything that you want me to CR, add me [18:19:16] hoo: added you an hour ago or so. the rdf thing. [18:19:39] ah ok [18:21:46] hoo: touches a lot of files, but it's fairly streight forward [18:22:02] except maybe the tests... [18:44:43] DanielK_WMDE: what about https://gerrit.wikimedia.org/r/#/c/246130/? Markus answered it [18:45:45] hi SMalyshev ;) Give me a poke when you have time to step through and reproduce that ticket UI have being posting on! [18:46:23] addshore: sure! I have a meeting now but I'll have some non-meeting time in about 20-30 mins. thanks! [18:46:29] awesome! [18:50:33] DanielK_WMDE: reviewed that one you linked to, going to have a look at some of the other stuff in the chain now [19:15:11] addshore: ok, I'm ready now [19:15:34] Awesome, *goes to do some stuff [19:15:45] So I will use Q4115189 again [19:15:59] ok, let me open it [19:16:12] specifically Q4115189$40e7aff9-40ba-c125-0617-5c9f2483afd3 :) [19:16:21] reference has is currently 386855b9f5c283dcb93e19d2807bbbe42cae159c [19:16:23] *hash [19:17:03] ok, let me see what the sparql says [19:18:04] So for me sparql find it currently >>> http://tinyurl.com/qym526r [19:18:08] *finds [19:19:21] ok [19:19:30] let me set up the logging and then edit it [19:19:36] okay :) [19:20:55] ok, logging set up [19:20:59] now I'm editing it [19:21:05] *watches* [19:22:24] ok, changed [19:22:29] So the new hash is [19:22:30] 759c2651a64c4e6387469cc3532f8b408e873b3c [19:22:53] http://tinyurl.com/oljf8fj [19:23:18] so now for both queries the statement returns, when using both the current and previous hash! [19:23:29] hm indeed now I see both in the hash [19:23:32] weird [19:23:36] yup :p [19:23:40] in the results I mean [19:23:44] but I think removing the reference then removed them both [19:23:50] let me see what the log shows [19:23:58] =] [19:24:22] don't change anything yet, I want to see what happens there [19:30:45] hmmm, when I manually update it, it's gone. Very weird... [19:33:52] DanielK_WMDE: https://github.com/unclebob/WTFisaMonad/blob/master/Monads.pdf [19:41:05] addshore: some weird stuff is going on there... all the queries look ok and it should be deleted, but somehow it isn't... I'll dig more into it [19:41:45] Awesome! :) [20:18:10] sjoerddebruin: https://www.wikidata.org/wiki/Q876040 [20:18:36] Yeah, I know. But I can't find a news article that is saying that anymore. [20:18:44] Erasmusprijs just says Wikipedia now... [20:19:22] I have the award invitation here. It says Wikipedia community [20:19:28] And it's also in their press release [20:19:50] Hm [20:20:07] https://nl.wikipedia.org/wiki/Gebruiker:Multichill/Erasmusprijs [20:20:12] addshore: thanks for the RC review [20:20:21] Moeten nog een paar jaartallen bij en dan is ie compleet [20:20:26] hey multichill! [20:20:36] hi DanielK_WMDE [20:20:51] DanielK_WMDE: https://de.wikipedia.org/wiki/Benutzer:Multichill/Erasmuspreis for you :-) [20:21:13] Is there a good Sparql tutorial? [20:21:25] \o/ [20:22:11] sec^nd: there are tons of sparql tutorials. no idea how good they are... [20:22:22] * DanielK_WMDE tends to just fiddle with examples and then read up on details [20:23:07] sec^nd: if you know a bit of sql and first order logic, it's pretty untuitive (except for the unintuitive and confusing bits) [20:23:23] but i'm new to sparql myself, just dabbling [20:23:55] Well I guess a question I have is how would you query for things which are subclasses of something, like instanceof [20:24:30] animated film is instanceof film, how do I find all of the objects with properties which are instanceof file? [20:24:39] sec^nd: several of the example queries do that... [20:24:50] https://www.mediawiki.org/wiki/Wikibase/Indexing/SPARQL_Query_Examples [20:26:01] sec^nd: ?x wdt:P31/wdt:P279* wd:Q515. <--- this means "where ?x is an instance of a subclass of Q515" [20:27:46] DanielK_WMDE: what is the "/" in wdt:P31/wdt:P279? [20:28:09] Also what are the periods in queries for? When do I use periods and semicolons? [20:28:18] sec^nd: x/y/z is a path. x/y*/z is a path with y repeating any number of times [20:29:35] sec^nd: for periods and colons, look at how the Turtle notation for RDF works. That should make it clear. [20:30:18] anyway, i'm off for now [20:42:34] Thanks you [20:44:26] aude: Lydia_WMDE: https://tools.wmflabs.org/audetools/wikidata-stats/ Do you know whether that is up to date? [20:44:56] hoo: it is not - katie is creating subpages for dates and this is still the oldest... [20:45:09] ah ok [20:45:24] because I'm seeing different numbers [21:15:51] Lydia_WMDE: Stats like at https://www.wikidata.org/wiki/User:Multichill/Erasmus_Prize would maybe be nice to have as a standard service [21:16:10] *nod* [21:16:11] (and with less crappy layout) [21:16:15] :D [21:16:52] Charlie Chaplin clearly wins! :P [21:49:45] > 930k changes have been propagated via UpdateRepo to date! [22:00:10] <3 [23:49:24] hi [23:50:30] oh [23:57:09] mork: Hello.