[05:16:46] (03CR) 10Legoktm: "(2 comments)" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81671 (owner: 10Addshore) [08:49:08] (03PS7) 10Michał Łazowik: (bug 40810) Add badges support to Item [DO NOT MERGE] [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 [08:49:16] Tobi_WMDE: http://www.scivee.tv/node/55318 [08:49:27] der player ist aber mist, man kann nicht springen [08:49:47] (03CR) 10jenkins-bot: [V: 04-1] (bug 40810) Add badges support to Item [DO NOT MERGE] [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 (owner: 10Michał Łazowik) [08:50:30] Tobi_WMDE: dafür haben sie ein vernünftiges url-schema: die flv-datei gibt's unter http://www.scivee.tv/assets/video/55318 [08:53:04] Tobi_WMDE: slides brightbyte.de/repos/papers/2012/Wikidata-SWIB-2012.pdf [08:54:15] DanielK_WMDE: so, the diff viewing is not broken, but the header is "linki / eowiki / name" instead "linki / eowiki" [08:54:21] with name not localized [08:54:57] I think I'll have to do something similar to claims (in Wikibase lib/repo) [08:55:05] lazowik: yea, these headers are not localized at all. [08:55:09] i think this is fine for now [08:57:01] DanielK_WMDE: I'll add one test and you can recheck [09:00:31] lazowik: ok [09:00:36] lazowik: i am cleaning up some code for diff viewing [09:00:40] Tobi_WMDE: what's your take on https://bugzilla.wikimedia.org/show_bug.cgi?id=53554 ? [09:00:46] i can take a look at how to accommodate the badges there [09:00:50] i think this is either very simple to fix, or very hard... which one is it? [09:01:15] oh that bug :) [09:01:27] aude: I only need to replace the third step on path (name/badges) to localized text [09:02:58] links / enwiki / name [09:03:08] or links / enwiki /badges [09:03:18] and name/badges need to be localized [09:03:46] now it's a recursive function going down to atomic diff while creating path that is viewed in header [09:04:15] and claims use own, quite complicated formatting [09:04:19] aude: ^ [09:08:05] lazowik: nothing in that path is currently loclaized [09:08:13] I know [09:08:24] i wouldn't try to add that now [09:08:30] it's a separate issue [09:08:53] but I don't know whether this should be changed, or another formatter added [09:08:58] like claims have [09:09:30] lazowik: don't worry about it yet [09:09:35] ok [09:09:38] shoudl be done in a separate patch [09:10:27] (03PS8) 10Michał Łazowik: (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 [09:10:35] DanielK_WMDE: ^ should be ok now [09:10:43] hope so :) [09:11:50] wait [09:11:57] something strange I notice [09:13:15] DanielK_WMDE: that's mostly because we prevent default behaviour, I'm not exactly sure why. I guess for input fields like label and descriptions it should not be too hard to enable it again, whereas it would not make a lot of sense for where the entity selector is used [09:13:21] (03PS9) 10Michał Łazowik: (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 [09:17:33] (03PS10) 10Michał Łazowik: (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 [09:18:10] DanielK_WMDE: ^ now, there was a totally unrelated variable added... no idea how it ended up there [09:18:20] and a missing asterisk in comment... [09:19:30] Tobi_WMDE: yes, i agree. could you puit a short note on bugzilla, please? [09:22:08] DanielK_WMDE: done [09:22:21] lazowik: for patch set 7, did you rebase *and* make changes? [09:22:25] that makes review a pain... [09:22:29] Tobi_WMDE: thanks [09:22:58] uh, I'm not sure [09:23:07] I didn't know that... [09:23:36] the problem is that i can't see a clean diff showing what you changed [09:23:42] i have to review everything again [09:23:54] sorry :( [09:23:55] or go through all the changes caused by the rebase, too [09:24:14] lazowik: not a big deal for a smallish thing like this, but please consider it in the future [09:24:20] sure! [09:24:43] * aude sure i made that "mistake" before [09:26:47] yea, me too. it happens. [09:26:54] but it's good to try and keep in mind [09:28:14] DanielK_WMDE: hmm, I can't see any changes that are not mine (from 6 to 10) [09:30:05] lazowik: select 6 as the "reference version", than look at PS 7. that should give you the diff between 6 and 7. but it includes stuff from the EntityId refactoring [09:30:53] lazowik: if you doin't select a reference version, the diff is always against that commit's parent, which is alsways "clean", but doesn't let me see what you changed between one PS and the next. [09:31:37] I had 6 as reference comparing with 10, now 6 with 7 an nothing... [09:31:45] leave it [09:32:08] if you mean q => Q it's mine [09:32:11] lazowik: hm? it's fine, i sorted it out by now [09:32:27] ok [09:33:17] lazowik: PS 7 against base has no change to the release notes. PS 7 against PS 6 does: https://gerrit.wikimedia.org/r/#/c/74676/6..7/RELEASE-NOTES.md [09:33:44] lazowik: anyway, the code looks good. i'll do some testing now, and merge if all goes well. [09:33:52] ok [09:34:13] DanielK_WMDE: might want to do that on different db [09:35:15] bah :) [09:35:28] if it breaks the page, i'll delete the page [09:35:30] (03PS7) 10Addshore: Add editentity claims support [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81671 [09:35:58] hehe [09:47:56] lazowik: this patch is not intended to cover serialization in the external form (used by the API), right? [09:48:04] at least, it doesn't :) [09:48:08] nope [09:48:12] ok [09:48:38] it isn't in datamodel, right? [09:48:55] I'll do the api next [09:50:54] getting & editing at once [09:52:55] lazowik: cool. please coordinate with addshore, he has been doing a ton of refactoring in the api code lately. [09:53:03] ok [09:53:05] *waves* [09:53:40] :D [09:53:44] <: [09:53:54] addshore: I think I'll just copy aliases [09:53:59] it's quite similar [09:54:21] (if aliases are a good code...) [09:54:31] as int he api modules for aliases? :> [09:55:10] indeed looks good :) [09:55:30] not entirely sure what your planning! but ping me if you need me :) [09:55:39] sure [09:56:49] lazowik: best start with SiteLinkSerializer. It's in Wikibase/lib [09:57:01] ok [09:57:41] (03CR) 10Daniel Kinzler: [C: 032 V: 031] (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 (owner: 10Michał Łazowik) [09:57:49] \o/ [09:57:50] (03Merged) 10jenkins-bot: (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - 10https://gerrit.wikimedia.org/r/74676 (owner: 10Michał Łazowik) [09:57:54] yay! [09:57:56] \o\ [09:57:58] yay :D [09:58:02] /o/ [09:58:07] \o\ |o| /o/ [09:58:18] ~o~ [09:58:24] :D [10:00:06] DanielK_WMDE: http://www.youtube.com/watch?v=ITWnyJoX0c0&t=0m13s [10:00:12] and I know this guy... [10:02:13] addshore: https://www.wikidata.org/w/index.php?diff=67028525 :< [10:02:35] >.< [10:02:48] legoktm: its not running at the moment, I think I have already been linked to that one [10:02:52] ok [10:02:53] https://www.wikidata.org/wiki/Wikidata:Requests_for_permissions/Bot/SamoaBot_37 [10:02:54] fyi [10:02:55] it should never ever ever do that :P [10:03:12] it shouldnt be possible for it to overwrite a link >.< [10:07:20] (03PS8) 10Addshore: Add editentity claims support [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81671 [10:07:29] legoktm: fyi ^^ ;p (fixed your comments) [10:07:55] ty :D [10:07:58] i was wondering [10:08:06] if i try to add 6 claims [10:08:09] the first 5 are fine [10:08:15] then the 6th one is malformed or something [10:08:37] i'm just going to get a standard error which doesn't tell me the first 5 went through right? [10:08:55] (03CR) 10jenkins-bot: [V: 04-1] Add editentity claims support [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81671 (owner: 10Addshore) [10:09:04] :< [10:09:29] hmm, in that case it depends whats wrong with the 6th claim? [10:09:45] but yes, it would fail :P [10:09:52] idk, something terrible :P [10:09:58] xD [10:11:16] (03CR) 10Daniel Werner: [C: 032] Re-evaluated listview and snaklistview module dependencies [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/80347 (owner: 10Henning Snater) [10:11:48] addshore: legoktm it should fail the entire edit [10:11:54] aude: yus [10:12:18] https://integration.wikimedia.org/ci/job/mwext-Wikibase-testextensions-master/4057/console O_po [10:12:37] (03PS4) 10Tobias Gritschacher: toolbarcontroller: Detach event handlers before re-attaching. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/79776 (owner: 10Henning Snater) [10:13:26] aude: ok, sounds good :D [10:14:02] aude, any idea whats up with those errors? O_O they are in client... nothing to do with anything I have touched as far as I can tell [10:14:17] (03CR) 10jenkins-bot: [V: 04-1] toolbarcontroller: Detach event handlers before re-attaching. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/79776 (owner: 10Henning Snater) [10:16:07] looking [10:16:29] ok, investigating [10:16:34] :< [10:17:15] (03CR) 10Tobias Gritschacher: [C: 04-1] "1. the move functionality should be made experimental before merging it. we want to be able to enable it on a later point. that could be m" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/80543 (owner: 10Henning Snater) [10:20:25] addshore: for starters: should I write a separate module? [10:20:43] or include that in wbsetsitelink ? [10:22:15] lazowik: i may have to revert the badge change - it broke some tests for the client extension. [10:22:23] :( [10:22:27] the change propagation code is making assumptions about what a diff for site links looks like [10:22:29] let's see how quickly it can be fixed [10:22:38] or we can exclude change handler test [10:22:43] until it's fixed [10:22:59] lazowik: katie is owrking on it, but at the moment, it keeps jenkins from approving any changes [10:23:06] ok [10:23:54] if it doesn't look like i can fix before lunch, then we either revert or have the test skipped [10:23:56] (03CR) 10Daniel Kinzler: [C: 032] Improvements to ByPropertyListUnserializerTest [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81554 (owner: 10Jeroen De Dauw) [10:24:24] aude: ok [10:25:22] (03Merged) 10jenkins-bot: Improvements to ByPropertyListUnserializerTest [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81554 (owner: 10Jeroen De Dauw) [10:37:01] (03PS1) 10Aude: skip change handler test, until it is fixed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81904 [10:38:37] (03CR) 10jenkins-bot: [V: 04-1] skip change handler test, until it is fixed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81904 (owner: 10Aude) [10:40:06] (03PS2) 10Aude: skip change handler test, until it is fixed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81904 [10:44:14] i can't edit on master [10:44:16] Exception Caught: $data needs to have a "badges" key [10:44:43] lazowik: ^ [10:44:57] DanielK_WMDE: ^ [10:45:02] mhm [10:45:20] we might need api support [10:45:21] now It's better to revert [10:45:28] together with the data model change [10:46:08] aude: you can't edit site links? but i tried that... [10:46:08] aude: not related [10:46:24] DanielK_WMDE: no i can't [10:46:31] hm :/ [10:46:53] that's odd. can you do the revert and supply lazowik with a stack trace? [10:46:58] i can join the bug hunt, if you like [10:46:58] I'd go with revert, too much fuzz [10:47:53] investigating [10:48:13] I think that it shouldn't be treated as exception [10:48:28] lazowik: what shouldn't? [10:48:38] requiring badges key [10:48:45] Tobi_WMDE: which api module does the ui use for site links? [10:48:56] we could just pretend it was an empoty list of badges. [10:49:01] but why and how does it happen at all? [10:49:08] I don't remember [10:49:09] i9 was able to edit without a hitch [10:49:13] exactly [10:49:17] * aude needs to know which module to look at [10:49:27] but I think this was related to patching [10:50:10] hm.... [10:50:23] i didn't try to generate an edit conflict to try patching... [10:50:36] it somehow deleted the key during patchin [10:50:46] but I'm not sure if it's that [10:51:17] DanielK_WMDE: I think it's easier to just see print_r of patching effect in tests [10:52:29] hmm, the badges stay there even if they're empty [10:52:34] lazowik: but you have to trigger the patching somehow [10:52:43] from tests :) [10:52:50] our unit tests actually should do that... i wonder why they don't [10:52:58] patching is there [10:53:00] i didn't see any errors [10:54:54] ok, it uses set site link api module [10:55:48] aude: you can't edit _any_ sitelinks? [10:56:12] only tried one item [10:56:43] can you get current serialization of it? [10:58:19] aude: setSiteLink [10:58:30] Tobi_WMDE: thanks [10:59:29] lazowik: http://pastie.org/8283064 [10:59:38] when i do $entity->serialize() [11:00:04] still debugging [11:01:23] DanielK_WMDE: what should the unit tests do that they dont? :P [11:01:52] it looks like the change op propertly adds "frwiki":{"name":"Canada","badges":[]} to the entity [11:02:20] addshore: apparently, test patching of sitelinks [11:02:34] (as used when resolving edit conflicts) [11:04:55] DanielK_WMDE: patching is tested in ItemTest [11:05:44] ...and that passes... [11:06:26] it just doesn't make sense [11:08:30] but if aude can't edit it's not related to patching [11:08:47] making progress [11:09:16] when i dump the data in SimpleSiteLink::newFromArray, then the new site link is missing badges param [11:09:21] per the exception caught [11:09:26] mhm [11:09:38] so data['links'] is broken [11:10:10] just trying to figure out where this is called [11:10:49] SimpleSiteLink::newFromArray? [11:10:58] it's added by my patch [11:11:05] and called only in two places in Item.php [11:11:15] ok [11:11:25] I'd search where Item is created with data provided [11:11:56] k [11:13:18] huh, my bad... [11:13:24] i reloaded my page and it works now [11:13:26] :D [11:13:51] looks much better [11:14:17] no way :D [11:14:30] ok, then to figure out what's wrong with change handler [11:14:48] I knew this didn't make sense :) [11:14:48] if we want to skip the test then someone can approve https://gerrit.wikimedia.org/r/#/c/81904/ as a temporary measure [11:14:51] DanielK_WMDE: ^ [11:14:52] yeah [11:15:19] ause: WikibaseLib still uses deprecated SiteLink [11:15:23] that's in WikibaseLib [11:15:25] ok [11:16:38] (03CR) 10Daniel Kinzler: [C: 032] "as a temporary fix, ok. This should *not* stay in long." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81904 (owner: 10Aude) [11:16:44] agree [11:16:58] please make sure this doesn't stay in. i'd rather revert than skipping tests for days. [11:16:59] if i can fix by time i go home or by monday [11:17:02] yes [11:17:14] only to not block other people [11:17:31] so, the editing issue was spurious, and we only have the problem with the diff handling in the change handler? [11:17:36] yep [11:17:46] hm, i guess i should have had a look at the old SiteLink, and how that deals with the new situation... [11:18:04] (03Merged) 10jenkins-bot: skip change handler test, until it is fixed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81904 (owner: 10Aude) [11:18:22] i'll need to make sure the old branch and master can also work together with dispatching [11:18:46] or ensure dispatching won't break [11:20:00] Hi! is it possible to execute Module:Taxobox on wikidata from javascript? I would like to have a gadget which shows the current state of a taxon item (a item which has P225). [11:22:24] reimann: i haven't looked into Module:Taxobox, but Wikidata is not a client itself, which means it has no access to the data [11:22:29] Ironic, huh? :) [11:23:07] I know. This is solved with WbHacks triggering the json api... [11:23:49] not nice but working [11:25:37] The problem is that Wikidata community must show the strength of modelling things in Wikidata to convince Wikipedia chapters to use and help extending the data [11:26:16] aude: ChangeHandler in lib, lines around 570 [11:26:41] reimann: so what is the question at hand? [11:26:45] lazowik: yes [11:27:02] the code has numerous issues there [11:27:06] unrelated also [11:28:08] I think that the $siteLinkDiff[ $siteGlobalId ] : null; should be now $siteLinkDiff[ $siteGlobalId ]['name'] : null; [11:28:17] hmmm [11:29:03] lines 572 & 3 [11:29:09] Module:Taxobox is LUA. It works like a template e.g. on talk pages. But I would like to call it also from a Javascript gadget to show it directly on e.g. https://www.wikidata.org/wiki/Q36146 [11:29:17] lazowik: ok [11:29:29] that's for checking which sites are to be updated [11:30:16] is this possible or do i need to reimplement everything in javascript? [11:31:15] aude: also 826 [11:31:22] $diffOp = $siteLinkDiff[$siteGlobalId]; [11:31:53] and from 868 [11:32:57] but, is it possible for old format to get in there? [11:33:19] Denny_WMDE: or shorter: Is it possible to execute a LUA module from Javascript? [11:34:30] reimann: interesting question [11:34:42] reimann: it should be possible to use the parse api module to do that [11:35:05] reimann: and get the HTML back. I have not tried that though, just gut feeling [11:35:14] lazowik: now have just one test failure [11:35:46] aude: have you changed all the 'page' => $diffOp->getNewValue() ? [11:35:53] looking [11:36:06] either add $diffOp = $diffOp['name'] [11:36:19] or 'page' => $diffOp['name']->getNewValue() [11:36:39] and if there can changes in queue with legacy format [11:36:46] reimann: checking the api doc, but it seems that action=parse should be very helpful [11:37:02] you need to make all these changes only when $diffOp is an instatnce of Diff [11:37:07] and figs are my favorite fruits [11:37:07] ok [11:37:38] (03CR) 10Addshore: [C: 04-1] "A marker, see some more work on claim guids" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/81671 (owner: 10Addshore) [11:38:18] DanielK_WMDE: aude have you added things to the weekly summary doc? :) [11:40:10] aude: I'll need to go in 5 mins for an hour max [11:40:47] addshore: do you have the link? [11:40:56] aude: the basic rule is that whenever there is $diffOp = $siteLinkDiff[$siteGlobalId]; it needs to be $diffOp = $siteLinkDiff[$siteGlobalId]['name']; [11:41:04] ok [11:41:09] * aude probably getting lunch [11:41:14] and all variants with foreach and so one [11:41:29] sorry for adding work for you [11:41:33] lazowik: that fixed most of the tests [11:41:50] I'm not still sure where the actual change takes place [11:41:51] it could be a different issue for the one test [11:42:00] that might be the remaining ones [11:43:06] aude: 357 [11:43:08] the isset [11:43:22] all issets also need to check for change in 'name' [11:43:46] otherwise it will think that page was changed when badges were changes actually [11:44:54] aude also isset( $siteLinkDiff[ $siteGlobalId ] ) to isset( $siteLinkDiff[ $siteGlobalId ]['name'] ) | line 572 [11:46:37] aude: RefreshLinksJob [11:47:43] ok :) [11:47:57] but that's in core? [11:47:59] or what [11:48:03] I can't find it [11:48:15] and have to go, will check in in about an hour [11:48:17] * aude needs to look through the entire class [11:48:27] not everything might be covered adequately with tests [11:49:46] yup, core [11:50:35] ./extensions/Wikibase/client/includes/WikiPageUpdater.php: //XXX: use \RefreshLinksJob2 ?! [11:50:35] ./extensions/Wikibase/client/includes/WikiPageUpdater.php: $job = new \RefreshLinksJob( [11:50:35] ./includes/AutoLoader.php: 'RefreshLinksJob' => 'includes/job/jobs/RefreshLinksJob.php', [11:50:37] ./includes/AutoLoader.php: 'RefreshLinksJob2' => 'includes/job/jobs/RefreshLinksJob.php', [11:50:39] ./includes/DefaultSettings.php: 'refreshLinks' => 'RefreshLinksJob', [11:50:41] ./includes/DefaultSettings.php: 'refreshLinks2' => 'RefreshLinksJob2', [11:50:43] ./includes/job/jobs/RefreshLinksJob.php:class RefreshLinksJob extends Job { [11:50:45] ./includes/job/jobs/RefreshLinksJob.php:class RefreshLinksJob2 extends Job { [11:50:47] ./includes/job/jobs/RefreshLinksJob.php: $jobs[] = new RefreshLinksJob2( $this->title, [11:50:49] ./includes/job/jobs/RefreshLinksJob.php: # Roughly speaking, for each page, one of the "RefreshLinksJob" jobs will [11:50:51] ./includes/job/jobs/RefreshLinksJob.php: $jobs[] = new RefreshLinksJob( $title, [11:50:53] ./includes/LinksUpdate.php: $job = new RefreshLinksJob2( [11:50:55] all appearances [11:50:56] pastebin [11:50:57] ! [11:58:28] Is there a preferred way of doing instance-of (is-a) checks in PHP? [11:58:48] (Concerning items being instances of other items) [11:59:58] I used an “is-a” property in my code so far, but since this stopped working after pulling master from Wikibase today, why not do it better if possible [12:04:45] Granjow: the wikibase code should have no influence on that whatsoever. How did it "stop working"? [12:04:56] and how exactly did you use the "is-a" property? [12:07:01] DanielK_WMDE: My extension extends wikibase and relies on its code ;) I don't get any results back anymore (still investigating why and where). No error messages though. [12:08:24] I created a property that is called “instance of“. In order to use it, I atm search for this property by name and then use its PID. [12:11:00] Code is here: https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FLifeWeb/4245d43bfc853925af8c74770d61adebfef29de1/lib%2FLWItem.php [12:11:01] Line 148 [12:26:35] Ah, I guess it changed from EntityId to EntityIdValue? [12:26:50] It is sometimes a bit difficult to follow since changes are not really posted anywhere [12:35:33] DanielK_WMDE: ping? [12:39:06] lazowik: think i have the change handler fixed [12:39:49] i may go ahead and clean up some of this code while i'm working on it [12:39:53] and add more tests [12:40:22] aude: yay :) [12:40:37] what was the rest? [12:40:50] the getSiteLinkDiff() method now returns a $key => Diff [12:41:05] so then i need to access the operations in the diff [12:41:32] change handler is doing some evil stuff [12:41:44] that's what I meant all the time :) [12:41:50] yeah [12:41:51] (not that it's doing evil :D) [12:41:59] your code is not evil [12:42:03] change handler is evil [12:42:20] Denny_WMDE: thank you! this works (putting the module call in a template seems to be required): https://www.wikidata.org/w/api.php?action=parse&text={{Taxobox|qid=Q36146}}&prop=text&contentmodel=wikitext&format=jsonfm [12:42:21] 10[1] 10https://www.wikidata.org/wiki/Template:Taxobox [12:42:33] that code was created at the very beginning of wikidata project [12:42:59] I meant that I meant all the tim (wut?) that getSiteLinkDiff() returns Diff [12:43:05] right [12:43:24] not that I meant that I meant all the time that the change handler is evil :) [12:43:28] i might need to add some evil hacks into the deployment branch so that nothing breaks [12:43:32] heh [12:43:35] Granjow: sorry, back now [12:43:39] yep, totally clear :D [12:43:59] branches always get the minimal changes required to make it compatible [12:44:02] no refactorigns [12:44:13] mhm [12:44:14] think it won't be difficult [12:45:00] but I can't imagaine adding badges changes dispatching without refactorinh [12:45:01] *g [12:45:15] Granjow: yes, line 169 needs to check for EntityIdValue. [12:45:17] ok, off to eat [12:45:20] lazowik: ok [12:45:42] Granjow: though it would probably be better to check the property's type ID [12:46:21] DanielK_WMDE: What's that? [12:46:32] Granjow: also, contruction of EntityId objects has been simplified. Instead of newFromPrefixedId, you can just use the constructor. But you should use ItemId resp. PropertyId, if you know the type. [12:46:56] reimann: glad it worked out [12:47:08] Granjow: every property has a type associated with it. You get it from a PropertyDataTypeLookup. [12:50:51] (03CR) 10Tobias Gritschacher: [C: 032] "+1 already by Zfilipin. This wont harm anything.." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/80063 (owner: 10Tobias Gritschacher) [12:51:24] DanielK_WMDE: Type meaning the data type it refers to? [12:52:17] (03Merged) 10jenkins-bot: Added cucumber tests for entity descriptions [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/80063 (owner: 10Tobias Gritschacher) [12:52:35] DanielK_WMDE: And what is the reason that Entities have had getPrefixedId() changed to getId() but Items did not? [12:53:18] * EntityIds did not, not Items [12:54:05] or ItemIds ... argh. [12:54:34] Granjow: Entities hav had getId() for a long time (it returns an object). [12:54:51] Entity::getPrefixedId was a legacy method. I guess it got phased out during the EntityId refactoring [12:55:18] Granjow: ItemId and PropertyId are now concrete subclasses of EntityId (which will become abstract soon) [12:57:26] DanielK_WMDE: Is there anywhere some kind of data model for the code? It is currently a bit hard for me to always parse all code via /dev/brain and build a model there [12:57:27] I see how the EntityId vs. EntityIdValue thing is confusing. BUt introducing special ID classes for different types of entities shzould not cuase problems, as far as I can see. [12:57:55] Granjow: well, the code is the model :) for the data model component, it would be nice to have a pretty diagram [12:58:12] maybe somebody has time to make one :) [12:58:25] or is there a good tool for making uml diagrams from php code? [13:00:12] http://stackoverflow.com/questions/393603/php-uml-generator [13:02:59] Granjow: the best place would probably be http://wikidata-docs.wikimedia.de/ [13:03:12] but currently, this isn't updated very often, and does not include diagrams [13:03:18] but it seems like it could... [13:03:28] Granjow: feel free to push for that on the mainling list and/or bugzilla :) [13:03:29] needs updating [13:07:22] aude: I'm back, what's the status? [13:07:53] lazowik: cleaning up the code [13:08:09] aude: ok, can I help? [13:08:22] you can help code review [13:08:28]