[04:33:57] Hey y'all [04:34:09] I need a European who plays video games [06:03:58] Sven_Manguard: ? [06:04:27] Reedy: I resolved the issue. It was with how people refer to PEGI ratings [06:26:32] Reedy: still there? [09:37:43] what's good for disabling action=raw on items? [09:38:24] if my application is actually able to process internal format [10:26:28] (03CR) 10Addshore: [C: 04-1] "From a car point of view this looks fine (as far as I can tell from my tablet)." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/84187 (owner: 10Legoktm) [10:26:54] addshore: car? [10:27:23] Heh [10:27:25] Cr [10:27:30] :P [10:27:40] Silly autocorrect [10:27:41] addshore: also, what do you think of my comments about that on https://bugzilla.wikimedia.org/show_bug.cgi?id=38971#c3 [10:27:57] *goes to look* [10:28:18] Does the code work BTW? :) it looks like a nice and simple change :p [10:28:47] I didn't test it, I don't have a Wikibase dev environment set up yet [10:28:54] I've been meaning to do it for a while now [10:29:19] Kk :) [10:30:12] One thing I have just thought of, it would EB nice to be able to fetch multiple revisions at a time, as with get entities you can get multiple entities at a time for example :p we should try and do this for revisions too! [10:30:24] yeah, thats what my question on the bug is about [10:30:41] also, i have a feeling that my current code would throw a fatal error if the revision you ask for is revdeleted [10:31:03] In that case I think it might want to use the revids as the keys in the array [10:31:16] Hah! We will see :p [10:31:42] you mean have {revisions: {###: {}, ###: {} }? [10:31:42] Caracteres inválidos en el enlace «: {}, »; no están permitidos: <>[]{} [10:31:53] where revisions is a top level key? [10:32:14] Yes [10:32:18] I think [10:32:22] hmmm [10:32:28] idk [10:33:08] also whats missing [10:33:15] is that i cant say "get the last 3 revisions" [10:33:23] you need to already know the revids [10:33:26] I mean, in theory what's to stop &ids=q1/q2&revs=1234/5678 [10:33:58] Hello! Addshore [10:34:16] Hello kolega2357,! [10:35:41] addshore: ideally that would be possible, though it really isnt easy to do in the current structure [10:36:21] https://www.wikidata.org/w/api.php?action=query&prop=revisions&titles=Q76&rvprop=content&format=jsonfm [10:36:30] JSON in JSON! [10:36:33] He he :) if not, for this change being able to get 1 rev is fine, and I think it would be best to be served in its own revisions array [10:37:08] wbgetrevisions xP [10:37:11] Haha, that's looks like an escaped mess xD [10:37:24] Ewww, I want less modules, not more :p [11:48:03] addshore: what's the format which action=wbeditentity accepts? [11:48:15] it's not even the one from EntityContent::serialize() [11:48:58] Take a look at the examples [11:49:08] Its the external serialization [11:50:03] addshore: the one from $serializer->getSerialized( $content->getEntity() ) ? [11:51:07] Mhhm, not sure where it actually comes from, I think its built up, there isn't a single method to get it (other than something like add entity to result maybe) [11:51:21] Take a look at the add*to result methods [11:57:10] addshore, liangent: EntityContent::serialize() gives you the *internal* serialization. that should be used to store entities in the database, and for nothing else. [11:57:46] addshore, liangent: the structure accepted by wbeditentity should be consistent with what wbgetentities returns (i.e. what EntitySerializer generates) [11:58:04] but remember that wbenditentity accepts a *patch*, so the format is not 100% the same [11:58:37] liangent, addshore: and sorry in case i missed some context here :) [11:59:08] DanielK_WMDE: there's a &clear= to override a full entity [11:59:32] liangent: yes, i know. [12:00:00] but it'S still processed as a batch (against an empty entity), not validated and stored. [12:00:06] err, as a patch [12:00:58] DanielK_WMDE: so what's the difference from storing a new one? [12:02:03] [info] => unknown key: id [12:02:04] liangent: ideally, none. i'm just saying that the format is designed to act as a patch, so it's not 200% the same as the external serialization. though it should be very similar. [12:02:18] still have to clean up the output of EntitySerializer... [12:02:37] hm, "id" should be accepted as a key. but talk to addshore and tobi about that, they refactored the code lately :) [12:03:38] liangent: it would be interesting to have a list of things that don't work when trying to use the output of EntitySerializer directly with wbeditentity. Ideally, it would just work, but there may be some things that don't and can't. not sure [12:04:44] liangent: i'd suggest to file a bug about this not working in a streight forward way, so we have this documented and can discuss better solutions [12:04:49] or put it on the mailing list [12:26:24] (03PS13) 10Jeroen De Dauw: Introducing SnakFormatter and SnakFormatterFactory [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/82400 (owner: 10Daniel Kinzler) [12:26:52] (03CR) 10Jeroen De Dauw: [C: 032] Introducing SnakFormatter and SnakFormatterFactory [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/82400 (owner: 10Daniel Kinzler) [12:29:28] addshore: https://www.mediawiki.org/wiki/Extension:Wikibase/API#wbeditentity - all mentioned 'item' keys are shown as 'entity' in wmf live output [12:29:39] is doc outdated or code outdated? [12:30:09] (03Merged) 10jenkins-bot: Introducing SnakFormatter and SnakFormatterFactory [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/82400 (owner: 10Daniel Kinzler) [12:30:16] (03CR) 10Jeroen De Dauw: "lolwat?" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/84207 (owner: 10Liangent) [12:39:04] liangent: the doc is outdated [13:16:47] DanielK_WMDE: thanks [13:16:55] Danielk_wmde, any chance you could check the comments on the change set I had for deprecating entity->new claim ? :) [13:17:04] and what's the correct way to extract integer 42 from string 'Q42'? [13:18:32] Liangent, don't think there is a simple way now, get numeric id is deprecated [13:19:36] liangent: DanielK_WMDE: Kill the doc! [13:19:49] addshore: IdGenerator::getNewId() is expecting an integer [13:20:09] Oh, jeroendedauw, could you check my comments on the patchset deprecating entity new claim also? :) [13:20:26] addshore: you are deprecating entity? [13:20:53] Deprecating entity>newclaim :p [13:21:08] Liangent, it expects a string and returns an int [13:22:01] addshore: eh it's expecting any class implementing it to return an int [13:22:30] Yes.. But you don't need to pass it an int.. [13:23:23] addshore: http://pastebin.com/ACTLEE7j [13:25:10] Heh, labsidgenerator? If you need the INT that's probably the best way. [13:27:08] liangent: if you really need the int for some reason, you can use getNumericId. We quite possibly will need to change it at some future point then though [13:27:56] addshore: link? [13:28:50] https://gerrit.wikimedia.org/r/#/c/82237/ [13:36:11] JeroenDeDauw, addshore: we need getNumericId for the forseeable future - als long as the database schema relies on numeric ids [13:36:19] i see no point in deprecating it without a good replacement [13:36:36] but we should restrict usage to where it's actually needed for the DB layer [13:38:44] addshore: http://pastebin.com/av7KfLt9 [13:38:52] and https://www.wikidata.org/w/index.php?title=Q14916139&action=history [13:39:00] going to use this as my new bot framework [13:40:47] Heh, similar to what I am aiming for ;p [13:43:16] Liangent is your framework code anywhere? ;p [13:43:52] addshore: on labs :p [13:44:36] Not on github anywhere? :p [13:45:35] addshore: in a local git repo [13:45:44] http://pastebin.com/64qCZFaz all work done today for wikibase [13:52:26] addshore: I'm not quite familiar with github operation ... can you give me some instruction? [13:52:57] :) [13:53:14] Make an account, make a repo, push your stuff to the repo :D [13:53:49] addshore: I already have a github account [13:54:20] and the repo -- difficult to do [13:54:42] it's a little more than one repo [13:56:33] the first is a slightly patched mediawiki core. most patches are to allow operations to succeed gracefully even when hooks interrupt them (where the hooks have actually done that) [13:57:16] the second is an Extension:Labs, this one is easy [13:58:08] the third is configuration info, currently not in version control, and lots of it is actually from wmf's config files [13:58:17] addshore: ^ any suggestion about pushing it? [14:15:49] DanielK_WMDE: please to reply to "The next problem with EntityId" (after you think about it for more then a second and have something usefull to add that is ;p) [14:43:24] liangent: ping: for fallback, the HTML title isn't escaping the page name for apostrophes; see https://www.wikidata.org/wiki/Q1991852 [14:51:13] !team [15:15:44] * Sky2042 hellos GerardM [15:42:39] Hi I try to deploy a repo wikidata but I the error "The specified article could not be found on the corresponding site." when "Set a site link". Somebody can help me ? [15:44:33] karima: Can you paste the link of the item on Wikidata and the link to the article? [15:45:25] hoi Sky2042 [15:45:37] http://www.sparql.pro/wiki/Item:Q2 http://en.sparql.pro/wiki/Test [15:46:45] Ah, on your test site, that's going to be fun to debug [15:47:13] may be the problem is with my line in the table Sites INSERT INTO `sites`( `site_global_key`, `site_type`, `site_group`, `site_source`, `site_language`, `site_protocol`, `site_domain`, `site_data`, `site_forward`, `site_config`) VALUES ("ensparqlpro","mediawiki","sparqlpro","local","en","http","en.sparql.pro.","",0,""); [15:47:46] yes is fun... I'm crazy lol [15:48:57] usually is the full url of wiki in the field "site link" ? [15:52:25] Sky2042: hmm only in history page for me? [15:52:44] liangent: I see it on the main item page. [15:52:51] and the problematic character is actually the ampersand [15:53:02] o_o [15:53:08] There's no ampersand in that nlwiki link [15:53:31] Sky2042: labels are different from sitelinks [15:53:51] the label in nl language is actually Eric's Corvette Caper [15:53:59] LOL [15:54:01] okay [15:54:08] blame the importing bot then I suppose :^) [15:54:20] but in history page, it's not escaped into Eric&#039;s Corvette Caper [15:54:29] in html [15:54:41] so I did find a bug? lol [15:54:59] karima: I setup my test instance some time ago and used one of the scripts to import some sample items [15:55:04] just not the one i thought :D [15:55:17] Sky2042: yeah. though it's not really related to fallback [15:55:46] and btw initially I was thinking it might be security-related so I didn't response at first [15:55:47] liangent: Well, yeah, now I can see that. :P [15:55:55] And no worries. [15:56:01] Figured I'd bring it to attention. [15:57:08] Sky2042: but it's a bit tough to find out who's responsible for escaping it [15:57:28] multichill: the doc to config the repo is too light... for the moment T_T [15:57:42] liangent: Just ping the default go to, who I should probably know and don't. :3 [16:00:54] multichill: How populate the table sites with a new site without matrix site ? [16:01:13] It just uses Wikipedia [16:18:36] Sky2042: want to cc https://bugzilla.wikimedia.org/show_bug.cgi?id=54153 ? [16:19:31] sendin' emails that i dont know why [16:19:33] >_> [16:20:31] Sky2042: not your fault :p [16:20:36] I subscribed cc changes on my side [16:20:53] * Sky2042 nods. [20:30:22] !team [20:30:30] Lokopereta: Yes? [20:30:34] jeje [20:30:40] huh [20:30:43] >.> [20:30:45] dat troll [20:30:50] . [20:30:55] JohnLewis: you just got dun trolled [20:31:09] Sky2042: :( [20:31:13] :D [21:19:34] What exactly is badges, btw? [21:20:00] Oh [21:20:00] badges are on? O_o [21:20:11] https://bugzilla.wikimedia.org/show_bug.cgi?id=40810 [21:20:16] That's badges. [21:20:17] Apparently. [21:21:29] yeah [21:21:31] keep up the good work