[09:26:07] <jubo2>	 Hi. I need to learn about the Wikibase export-import process. Any good links to read would be appreciated
[09:26:42] <jubo2>	 What works, what is not implemented and what the user should not even consider trying?
[09:27:32] <andre__>	 https://www.mediawiki.org/wiki/Wikibase/Indexing/RDF_Dump_Format
[09:28:07] <andre__>	 I guess.
[09:32:32] <jubo2>	 I'm also guessing about things andre__
[09:34:32] <jubo2>	 For my use case it would be useful to maintain property compatibility if possible
[09:37:34] <jubo2>	 A reasonable sustainable ability for the Consumerium use case would be to have one Wikibase Client being able to access 2 Wikibase Repositories when it has access to both of the databases.
[09:44:40] <jubo2>	 So we could use a regularly refreshed set Wikidata's items, claims and qualifiers in such a manner that it is not a "fork and growing entropy between" issue
[09:49:21] <jubo2>	 Consumerium contributors would be entering a lot of data on businesses if it just benefited non-WMF wikis. Repeated export-import to mirroring database accessable to Wikibase Client in addition to accessing a 2nd wikibase
[09:50:20] <jubo2>	 .. that would remove the disincentive to add to Wikidata
[09:51:42] <jubo2>	 I am of the understanding that the question about what kind of rules should be used for the selection of which database to use if they have overlapping hits is a very challenging one but surely not impossible
[09:59:30] <rawknuckles>	 hi, a question regarding notability. I'm thinking of adding item pages where each item represents an unique hot wheels car. According to the notability criteria, it (according to me) satisfies #2. Do you think it's valid?
[10:02:29] <pintoch>	 rawknuckles: if it is about a particular model, I would say yes (but not if it is one of the physical instances of a model)
[10:03:16] <pintoch>	 (do you have sources to describe them?)
[10:39:44] <rawknuckles>	 @pintoch Thanks, yes it would be an unique model (and not a physical instance since those would be tremendously many)
[10:42:34] <rawknuckles>	 This also had me thinking about other toys. For instance, other model car companies like Matchbox, Majorette would have their version (model) of a particular car (and in some cases, a completely unique design). So then the question I have is, does it become okay to make individual pages for each such model.
[10:44:06] <rawknuckles>	 Example: if Matchbox, Majorette and Hot Wheels each make a model of Porsche 911, then is it okay to have 3 such pages each identifying their individual designs?
[10:44:58] <rawknuckles>	 Or would we rather have an item page called Porsche 911 (1:64, Toy) with links to those 3 manufacturers
[10:45:52] <pintoch>	 good question
[10:46:45] <pintoch>	 intuitively I would create three different items in that case, but I am not familiar at all with this domain. you might get more informed answers at the Project Chat
[10:47:13] <reosarevok>	 I would also do three, but also not my field :)
[10:47:57] <rawknuckles>	 oh okay, thanks! I'll try to see what the folks there have to say about this.
[11:00:55] <floatingpurr>	 Hello there. I'm relatively new to Wikidata and I'm wondering what is "the" way to massively load data. I'm interested in loading a dataset of all the italian schools (http://dati.istruzione.it/opendata/opendata/catalogo/elements1/?area=Scuole) and I'm wondering if (i) the Italian Open Data License v2.0 could be fine for my purpose, and (ii) what are the steps that I need to follow to load this dataset (maybe proposing a bot
[11:07:11] <pintoch>	 floatingpurr: in theory the https://www.wikidata.org/wiki/Wikidata:Data_Import_Hub would be the designated place to discuss that I think
[11:07:23] <pintoch>	 it is not extremely functional yet but it is being revamped
[11:07:45] <pintoch>	 and at least you will get informed answers there
[11:10:32] <floatingpurr>	 hey pintoch, thank you! I found the description of the process not so clear (I do not exclude it's a problem of mine). Anyway, I'm going to follow your hint and to post my question there. Thanks : )
[11:11:29] <pintoch>	 floatingpurr: no no, it's clearly not a problem of yours - we are aware of the issue and actively working on it :) by using the space you will contribute to its revival ^^
[11:11:54] <pintoch>	 feel free not to use the existing templates if they do not work for you, you can just post a message there
[11:12:49] <floatingpurr>	 I'll do my best! Happy to contribute to this amazing project! : )
[11:23:19] <reosarevok>	 floatingpurr: "indicare la fonte delle Informazioni e il nome del Licenziante, includendo, se possibile, una copia di questa licenza o un collegamento (link) ad essa." might make it tricky, since you'd be re-licensing to CC0 by importing
[11:23:29] <reosarevok>	 But I'm not an expert, so do ask there and see :)
[11:27:32] <floatingpurr>	 Yes reosarevok, I guess it could be an issue. I'll do ask : )
[12:48:51] <repole>	 Hello I'm not sure where to ask, so I 'm sorry if this is not the right place. I want to use monthly page view count of wikidata on my project. Are there dump files for this? in advance
[14:24:58] <marktraceur>	 Thiemo_WMDE: I'm coming back to your advice after a hectic weekend, I have a quick question and then maybe some longer questions for email
[14:25:31] <marktraceur>	 Thiemo_WMDE: I like the idea of simply letting the EntityView class do the work, but I'm a little confused whether I'm meant to be creating the EntityView myself or using the DispatchingEntityViewFactory to do it for me
[14:26:23] <marktraceur>	 Thiemo_WMDE: If the latter, it looks like I'm going to have to do some changes to Wikibase to allow something outside of the WikibaseRepo class to use the entityTypeDefinitions to create the factory, or maybe just return the factory itself...
[14:27:02] <marktraceur>	 Thiemo_WMDE: But the EntityView class doesn't seem like it's meant to be called directly, so I'm a little stuck there
[14:31:54] <DanielK_WMDE_>	 marktraceur: if you know the concrete type of EntityView (e.g. MediaInfoView), you can create it directly.
[14:32:09] <DanielK_WMDE_>	 EntityView itself is abstract. 
[14:33:12] <DanielK_WMDE_>	 marktraceur: The factory uses entityTypeDefinitions. Returning the factory from entityTypeDefinitions would get us into a tail spin :)
[14:33:57] <marktraceur>	 DanielK_WMDE_: I mean returning the factory from a WikibaseRepo instance
[14:34:11] <marktraceur>	 But your way sounds better
[14:34:39] <DanielK_WMDE_>	 WikibaseRepo returns that factory from getEntityParserOutputGeneratorFactory
[14:34:54] <DanielK_WMDE_>	 ah, it'S wrapped. i see what you mean
[14:35:01] <DanielK_WMDE_>	 yes, you could also do that
[14:35:33] <DanielK_WMDE_>	 but as I said: if you know the exact type already, you can create it directly. unless you don't want to know about all the stuff it needs in the constructor.
[14:35:40] <DanielK_WMDE_>	 in that case, going via the factory is nicer.
[14:37:49] <marktraceur>	 Yeah, that's what I was thinking, but changing Wikibase is harder than changing MediaInfo
[14:37:52] <marktraceur>	 I'll try your way
[14:38:01] <Artix>	 Hello!
[14:38:51] <DanielK_WMDE_>	 marktraceur: should be straight forward eitheer way
[14:40:01] <Artix>	 Any particular reason Quickstatements is slow?
[14:40:43] <marktraceur>	 DanielK_WMDE_: Ugh, yeah, there's a lot of duplication that will go into calling the constructor myself, I'm going in
[14:42:18] <pintoch>	 Lucas_WMDE: follow up on my question from yesterday: I changed my mind and will use wbformatvalue instead (it should not generate so many queries)
[14:42:24] <pintoch>	 any idea where I could find out
[14:42:41] <pintoch>	 what the "options" parameter is expecting for this action? https://www.wikidata.org/w/api.php?action=help&modules=wbformatvalue
[14:42:47] <pintoch>	 (sorry for the line break)
[14:45:15] <Lucas_WMDE>	 hm, I don’t really know
[14:45:45] <pintoch>	 by sniffing my browser's calls it looks like it accepts {"lang":"en"}, I can add that to the wiki
[14:46:03] <pintoch>	 (but if it's a JSON parameter there could potentially be more options…)
[14:46:53] <DanielK_WMDE_>	 marktraceur: just split that one method in WikibaseRepo into two. Should be quite simple. 
[14:47:11] <pintoch>	 Lucas_WMDE: that's actually all I need, sorry for disturbing
[14:47:27] <marktraceur>	 DanielK_WMDE_: Yeah, there are still some things I need to pull together to get it working on the MediaInfo end, so I'll take a while to test it, but the Wikibase patch isn't difficult
[14:47:48] <Lucas_WMDE>	 wait, wait! :D
[14:48:08] <Lucas_WMDE>	 pintoch: afaict QuantityFormatter is the only formatter that uses it
[14:48:36] <pintoch>	 oh yeah? but surely external-ids with formatter urls must also rely on that, no?
[14:48:40] <Lucas_WMDE>	 with three options: whether to show the uncertainty margin, how much rounding to apply, and whether the unit is included or not
[14:48:48] <pintoch>	 ah right, the options
[14:48:57] <pintoch>	 yes ok
[14:49:12] <Lucas_WMDE>	 oh yeah, "lang" is a generic one I think
[14:49:16] <Lucas_WMDE>	 (sorry, just saw you earlier message)
[14:50:05] <Lucas_WMDE>	 oh, and LatLongFormatter also has several options
[14:51:12] <pintoch>	 okay… if you tell me how to find these out I can add them to the wiki
[14:52:06] <Lucas_WMDE>	 I ran
[14:52:09] <Lucas_WMDE>	 find -name '*Formatter.php' -exec grep --color=auto -rF OPT_ {} +
[14:52:11] <Lucas_WMDE>	 in my vendor/
[14:52:13] <Lucas_WMDE>	 directory
[14:52:18] <Lucas_WMDE>	 (sorry, extra line break)
[14:52:28] <Lucas_WMDE>	 unfortunately it looks like these libraries aren’t in codesearch.wmflabs.org
[14:56:06] <Lucas_WMDE>	 there are some here: https://github.com/search?utf8=%E2%9C%93&q=%22OPT%22+user%3Awmde+user%3ADataValues+extension%3Aphp&type=Code
[15:02:27] <Lucas_WMDE>	 pintoch: ^
[15:03:07] <pintoch>	 ah thanks a lot :)
[15:03:38] <Lucas_WMDE>	 yw d)
[15:03:40] <Lucas_WMDE>	 * :)
[15:45:58] <Artix>	 anyone here?
[15:51:22] <pintoch>	 Artix: yes?
[15:51:56] <Artix>	 I was about to say, some gy is editing on quick statements at a ridiculous rate.
[15:52:09] <Artix>	 my batch can't even get through without being very slow
[15:52:21] <Artix>	 I was about to ask if someone tell him to slow down
[15:52:30] <Artix>	 cuz he didn't read my talk page comment
[15:57:51] <pintoch>	 hmmm, yeah, 368/min seems a bit too much
[15:58:09] <Artix>	 I have an Alternate account
[15:58:10] <pintoch>	 let's try to invoke an admin
[15:58:12] <pintoch>	 !admin
[15:58:12] <wm-bot>	 Attention requested  HakanIST sjoerddebruin revi
[15:58:21] <revi>	 Hi
[15:58:48] <revi>	 You need to give me the link to them
[15:59:05] <pintoch>	 https://www.wikidata.org/wiki/Special:Contributions/Renamerr
[15:59:26] <Artix>	 https://wikidata.wikiscan.org/hours/24/users
[15:59:29] <revi>	 I'm pretty sure Ops guy won't like him either lol
[15:59:36] <Artix>	 350 edits per minute
[15:59:38] <revi>	 wow
[15:59:46] <revi>	 yeah, he need to be forcibly slow down
[15:59:54] <Artix>	 i once edited at that rate. got a stern warning from Lydia (wikidata staff)
[16:00:01] <Artix>	 and a block from multichill
[16:03:04] <revi>	 blocking him now
[16:07:42] <Artix>	 +1 for keep in 2019 Revi
[16:07:47] <revi>	 lol
[16:08:02] <revi>	 btw, another admins will have to watch his talkpage since it's 1AM and I'm about to go to bed
[16:08:31] <Artix>	 안녕히 주무세요
[16:08:33] <Artix>	 (I tried)
[16:09:11] <revi>	 100 out of 100
[17:07:32] <sjoerddebruin>	 We should really have some consensus about this. https://www.wikidata.org/w/index.php?title=Q646347&action=history
[17:24:38] <floatingpurr>	 Hey guys, it seems like blazergraph development has been discontinued. Do you know if wikidata is looking for an alternative triple store?
[17:45:30] <DanielK_WMDE_>	 floatingpurr: i have heard some FUD, but nothing solid. what's your source?
[18:27:41] <sjoerddebruin>	 You may notice that I fixed something around scientific articles. https://grafana.wikimedia.org/dashboard/db/wikidata-datamodel-statements?refresh=30m&panelId=3&fullscreen&orgId=1
[18:28:38] <Lucas_WMDE>	 sjoerddebruin: I don’t follow, was “scientific article” a subclass of “event” somehow?
[18:28:44] <Lucas_WMDE>	 does this graph include subclasses?
[18:28:47] <sjoerddebruin>	 Yes
[18:28:56] <Lucas_WMDE>	 ok, cool
[18:29:10] <sjoerddebruin>	 the label of https://www.wikidata.org/wiki/Q4119870 might be confusing
[18:29:21] <Lucas_WMDE>	 ah
[18:29:23] <Lucas_WMDE>	 I see
[18:29:38] <sjoerddebruin>	 The French one seems wrong as well.
[18:30:41] <sjoerddebruin>	 And not sure if "wissenschaftliches Werk" should be a German alias.
[18:30:55] <Lucas_WMDE>	 according to the German description it’s a class of text
[18:31:06] <Lucas_WMDE>	 but the statements look like it should be the action? the act of writing academically?
[18:31:24] <sjoerddebruin>	 That's what the English article is about.
[18:31:38] <Lucas_WMDE>	 the German article is about the kind of text
[18:31:50] <Lucas_WMDE>	 with a subsection “Wissenschaftliches Arbeiten” which looks like it’s more about the act
[18:32:08] <sjoerddebruin>	 Seems like there was confusion since the start: https://www.wikidata.org/w/index.php?title=Q4119870&limit=250&action=history
[18:32:40] <sjoerddebruin>	 The DNB entry seems to be about the act as well
[18:37:07] <sjoerddebruin>	 So... confusing...
[19:03:28] <reosarevok>	 Heh
[19:03:55] <reosarevok>	 So, our national library wants to add data for like 50k people to WD
[19:04:14] <reosarevok>	 But they didn't think about asking us at WMEE first, and they got blocked by Harmonia_Amanda :D
[19:04:30] <reosarevok>	 (perfectly legitimately blocked, mind)
[19:05:29] <reosarevok>	 Is there any better (read: faster) way than pywikibot for adding stuff? They complain that the 10 second waits would make the project take forever (which I think would be fine, but I thought I'd ask just in case)
[19:07:33] <sjoerddebruin>	 10 second wait sounds extreme.
[19:08:01] <reosarevok>	 Heh
[19:08:08] <reosarevok>	 One of my bots had a 30 second wait
[19:08:16] <reosarevok>	 Because even at 10 I got told I was going to fast
[19:08:42] <reosarevok>	 What is a good wait then? :D (maybe my code was fucked up in different ways back then)
[19:09:04] <reosarevok>	 They wanted to add info with QS, but that sounds like a bigger pain in the ass
[19:10:30] <sjoerddebruin>	 5 seconds seems like the default of pywikibot
[19:10:43] <sjoerddebruin>	 https://www.mediawiki.org/wiki/Manual:Maxlag_parameter
[19:10:49] <reosarevok>	 Ok :) 
[19:35:16] <reosarevok>	 What's the expected use of P101 vs P106 for people? 
[19:35:54] <reosarevok>	 Say, should someone have P106 Q3640160 or P101 Q7173 (or both)
[19:36:03] <reosarevok>	 It doesn't seem to be properly explained anywhere
[19:36:27] <Lucas_WMDE>	 I think both, but I’m not sure
[19:37:54] <reosarevok>	 Sigh 
[19:37:58] <nikki>	 I would imagine that p101 is unnecessary there
[19:38:09] <nikki>	 because marine biologist already says marine biology is the field
[19:38:18] <reosarevok>	 So, P106 linguist, P101 German or whatnot?
[19:38:21] <sjoerddebruin>	 It's more for specific branches yeah
[19:38:27] <reosarevok>	 And if so, should P101 be used as a qualifier for P106?
[19:38:36] <sjoerddebruin>	 Or https://www.wikidata.org/wiki/Q166153 :)
[19:38:41] <sjoerddebruin>	 And no
[19:38:42] <nikki>	 but if there's a more specific field than the occupation, then both would make sense
[19:38:54] <reosarevok>	 Hmm
[19:39:08] <sjoerddebruin>	 I think Siri shows it. :)
[19:39:15] <reosarevok>	 But what if a person is say a linguist and a painter - German as a field applies only to linguist
[19:39:22] <reosarevok>	 Not to their work in general
[19:39:37] <sjoerddebruin>	 "applies to part"? :)
[19:39:50] * reosarevok shrugs
[19:39:59] <reosarevok>	 Still feels having it as a qualifier would make more sense, but eh :p
[19:40:39] <nikki>	 a qualifier sounds sensible to me, but then it's probably going to be inconsistent >_<
[19:58:15] <reosarevok>	 nikki: isn't that true of most of WD? :p
[19:58:45] <nikki>	 yeah :/
[20:11:42] <reosarevok>	 Of course, I would probably just create "German philologist" and use P106 :p
[20:15:31] <sjoerddebruin>	 :O