[07:34:22] do you rank outdated statement as deprecated? [07:34:45] that is, statements which were true in the past but are no longer true [07:38:05] or, should it be done in the other way - marking all current statements as preferred? [07:48:30] liangent: when I brought it up a few weeks ago, I was told that deprecated is only supposed to be for things which weren't true in the past either [07:49:13] nikki, ya I found it just now [07:49:23] what about the other solution? [07:50:44] I'm not sure that marking *all* current values as preferred is right, but it depends on the context I guess [07:51:48] like I would expect there to be multiple preferred people in a list of members of a band, but only one preferred population of a city [07:52:44] nikki, what I tried to do is https://www.wikidata.org/w/index.php?title=Q13662685&type=revision&diff=279487993&oldid=260824188 [07:53:13] to move this division from one "upper" division to another [07:55:18] nikki, another related change is https://www.wikidata.org/w/index.php?title=Q10870043&diff=prev&oldid=279488831 [07:55:34] to remove it from the old upper division [07:56:04] there are still many subdivisions in the old upper division though [07:57:03] yeah, I was told it's wrong to use deprecated for those and the current code/division should be marked as preferred instead [07:57:13] I don't think it's very intuitive though :( [07:57:48] I think the word "deprecated" is a bit problematic [08:00:56] yeah, I think people are more likely to try and edit a historical statement to say that it's historical and shouldn't be used (unless you specifically want historical data), and then "deprecated" seems like the right option [08:06:55] but I can't think of any wording that would work. even if you call it "wrong", you could interpret it as "this statement is now wrong" and use it to mark something as deprecated that shouldn't be marked as deprecated. the only thing that seems like it would stop people from marking historical statements as deprecated is a separate option for historical [08:11:23] nikki, yeah and people (including me) were just trying to "deprecate" it from Wikipedia display [08:11:34] * nikki nods [08:22:35] list( , , ) = /* … */ [08:24:32] https://git.php.net/?p=php-src.git;a=commitdiff;h=e9068cefb4d97483823f8ee4dc384da7adf0c124 [08:34:04] https://gerrit.wikimedia.org/r/#/c/256900/ [08:39:59] but why did we do list( , , ) = /* … */ ? [08:40:19] So that it looks consistent in the file [09:38:18] addshore: what item has the most statements and how many statements is that? [09:38:25] do we know? [09:51:04] I imagine some of the items on https://www.wikidata.org/wiki/Special:LongPages must be near the top of the list for most statements [09:52:58] nikki: probably [10:01:23] https://grafana.wikimedia.org/dashboard/db/wikidata-datamodel has it on it :P [10:01:36] thanks addshore :) [10:44:21] Jenkins is going to fail a bit sorry :( [10:44:29] been messing up the MediaWiki config [10:48:08] :D [10:48:31] aude: when is the next branch of mediawiki core? [10:48:38] addshore: tuesday [10:48:42] yaya :) [10:52:06] aude: evening SWAT on a Friday is allowed config changes? [10:52:27] there is no swat on friday [10:52:41] https://usercontent.irccloud-cdn.com/file/55PzRlfP/ [10:52:55] oh wait, I guess thats thursday evening right? :P US time [10:53:03] oh, that is thursday in SF [10:53:08] cool! [10:53:10] 1 am for us [10:53:31] if you want to be awake at that time :) [10:54:35] ooooh, well, I want to try and enable catwatch on dewiki I think, and thats the only chance this week as the last fixes are in wmf8 [10:54:43] or I just wait until monday 14th! [10:54:57] i see [10:55:17] we might be able to ask greg and/or mukunda to do it after the train [10:57:01] ok, actually aude apparently it is waiting until january now ;) [10:58:06] sounds ok to me [11:02:18] Adrian_WMDE: hey, got a question for the isInitialValue method [11:02:28] Where? [11:02:42] I'm working on allowing to edit all languages in the entitytermsforlanguagelistview and change it to work on a Fingerprint object [11:03:11] now I wonder how to properly implement the isInitialValue method for that [11:03:28] should it just walk through all languages shown in the ui? [11:03:35] what if I edit some and then hide them? [11:04:20] Basically, isInitialValue is `this.options.value.equals( this.value() );` [11:04:42] Is this.value implemented? [11:04:49] Is Fingerprint::equals implemented? [11:06:48] Adrian_WMDE: looks like the page load tests are being merged :D [11:06:56] Whuuh [11:07:03] addshore: \o/ [11:09:10] Adrian_WMDE: oh, I realize I did a mistake in this.value [11:09:35] only added the edited terms but than all other terms would be removed :S [11:09:38] is there something like copying or cloning objects in our datamodel? [11:10:08] Good question [11:10:47] or can we say that this.value() just returns a Fingerprint containing all changed terms? [11:10:55] * benestar thinks that wouldn't be a good idea perhaps [11:24:20] Adrian_WMDE: do you have any further clues regarding https://phabricator.wikimedia.org/T117886 (qunit tests on the build with ULS enabled?) [11:24:50] it was some setting that we wanted to change [11:27:48] aude: Jan didn't like changing settings [11:28:19] Adrian_WMDE: btw how do I execute the qunit tests? [11:28:31] http://localhost/index.php?title=Special%3AJavaScriptTest%2Fqunit [11:28:41] cool, thanks [11:28:46] Adrian_WMDE: i know but trying to understadn the problem [11:29:04] ULS adds some animations and qunit tests check for animation queue empty [11:29:06] now ContentTranslation (which depends on "Wikidata") has qunit tests fail [11:29:09] and they have ULS [11:31:54] Adrian_WMDE: do I need to enable that page somehow? [11:32:00] Yeah [11:32:26] $wgEnableJavaScriptTest = true; [11:33:04] hah, that's why I didn't find it :P [11:34:22] Adrian_WMDE: https://phabricator.wikimedia.org/T92532 [11:34:36] apparently this is not new and i had this before (probably logged in with uls enabled) [11:34:54] * aude *sighs* [11:35:05] That also [11:47:16] Adrian_WMDE: all the page load tests are now merged :) now I guess we just have to wait for data, I'm guessing at most 24 hours! [11:47:20] I think they run daily [11:47:35] gonna be hitting my F5 button through the keyboard over the weekend [11:47:37] oh, or maybe every 3 hours... [11:47:59] we could have them within the next hour ;) [12:02:58] Not there yet [12:03:22] xD [12:41:19] Not there yet [12:46:09] Thiemo_WMDE1: https://gerrit.wikimedia.org/r/256932 [13:00:30] Not there yet [13:21:06] Lydia_WMDE: we have to many open bugs:p [13:36:14] Adrian_WMDE: https://gerrit.wikimedia.org/r/#/c/256936/ (i'm not so happy with this but might be ok in short term) [13:36:38] then would like to know if my suggestion there is reasonable to do [13:37:13] which maybe i can help work on for monday [13:42:04] Sure, it is [13:42:27] ok [13:42:43] we can't just use contentlanguages since that is missing some languages [13:42:49] apparently [13:43:45] We definitly should use ContentLanguages [13:46:49] You can easily create another ContentLanguages implementation that uses mw.config.get( 'wgULSLanguages' ) and $.uls.data.languages [13:47:20] :q [13:47:28] haha :D [13:48:33] It's there [13:49:27] aude: https://gerrit.wikimedia.org/r/#/c/256936/2/view/tests/qunit/wikibase/wikibase.getLanguageNameByCode.tests.js,unified [13:49:30] yeah that is ugly :-} [13:49:45] then if it fix the integration test [13:49:56] and given there is a task tracking the ugliness, maybe it is acceptable [15:33:36] hashar: made the task and might poke at it during the weekend or next week :) [15:34:34] and see we have a new wikidata build already [15:47:28] aude: has your ugly fix ( https://gerrit.wikimedia.org/r/#/c/256936/2 ) fixed the ContentTranslation qunit test ? [15:47:38] thoughI guess we need a new wikidata build [15:47:44] we made a new build [15:47:59] so let's recheck one of their patches [15:48:08] apparently it is fine now [15:48:13] ok [15:48:39] the whole code that cause the bug is ugly and i would definitely like to fix [15:49:01] yeah Jan has rechecked a few patches apparently [15:49:05] ok [15:50:15] aude: will close their bug / assign it to you and reference the task and gerrit ugly patch [15:50:23] ok, thanks [15:50:27] aude: thanks a ton for the fix ! [15:50:43] sorry for the trouble (and curious how this was only a problem now) [15:54:53] filled a summary https://phabricator.wikimedia.org/T120349#1852469 [15:55:11] aude: hopefully there will be no production impact whenever we push the new wikidata [15:55:24] I am assuming it is an oddity that only occurs in tests [15:55:47] hashar: it is [15:56:03] afaik and has been this way forever [15:56:37] ohh [15:56:37] so the test is actually valid ? [15:56:58] anyway, it is known and tracked. So I am not too worry :-) [15:58:17] without cldr, the "other languages" box says Deutsch, francais, etc for me [15:58:25] on wikidata, it says German, etc [16:00:18] "German" is nicer imho but we dont' want a hard dependency on everything [20:15:28] i know about a voluntary organization which works verygood [20:15:55] i want to write about them . I dont know how to write in wikidata [20:18:22] how top write in wikidata .. any reference [21:17:23] Doing PHP 5 after being used to PHP 7 is weird [21:22:17] JeroenDeDauw, As a Python-person. What is the difference? [21:33:31] tobias47n9e_: http://php.net/migration70 highlights the differences [21:36:15] hashar, Read over a couple of changes. Weird :P [21:36:35] ditto [21:36:48] assert() now throw exceptions [21:37:13] and can be made to have zero cost (i.e. turn them off) [21:37:20] sumOfInts(2, '3', 4.1) = 9, not sure if I would ever use that. [21:38:29] print (int)(2 + '3' + 4.1); [21:38:29] 9 [21:39:26] I would rather get a typeerror on that one, no? [21:39:36] ohhhh [21:39:47] function sumOfInts(int ...$ints) [21:39:58] magically cast all $ints parameters to int... [21:40:08] I can imagine how many bugs that one can cause :} [21:42:37] hashar, I somehow want to learn PHP just to see for myself if it is really as weird as people say. [21:43:07] tobias47n9e_: I think it used to be a terrible language for beginners [21:43:14] cause it gives you terrible practices [21:43:33] you got to be aware of some of its limitation, but overall it is not that wierd [21:44:17] I prefer python syntax (lighter) and python has a shit ton of built-in modules that are very convenient for sysadmin [21:44:25] I learned Python first, so I program everything like it is Python :P [22:04:19] addshore: what do you think of this value object? https://github.com/JeroenDeDauw/QueryrAPI/blob/master/src/ResponseModel/SimpleStatement.php#L3 [22:10:37] *looks* [22:11:10] it looks like a value object :P [22:11:23] safelyGet :p [22:11:48] addshore: very different approach than we use in Wikibase [22:15:21] addshore: I've extracted the infrastructore code out into a trait: https://gist.github.com/JeroenDeDauw/8cbe6888e63408571564 [22:15:55] I figure it fits together with https://github.com/JeroenDeDauw/QueryrAPI/blob/master/src/NoNullableReturnTypesException.php [22:16:03] Perhaps I should make a library for those [22:29:41] tobias47n9e_: PHP is not all bad tho. Its gradual typing is very nice compared to the absolusit stance of languages such as Java or Python on that topic [22:29:56] Still, overall I think Python is a much nicer language [22:30:16] Non-broken collections, named parameters, packages... [22:30:34] All those are huge when it comes to code quality and maintainbility [22:30:39] So are type hints tho :( [22:31:04] Oh, and non-string refferences to functions [22:31:09] That is also huge [22:47:17] JeroenDeDauw I removed another 1 whole file from lib ;) [22:48:42] JeroenDeDauw, It would be nice if Mediawiki would migrate to Python or Javascript. I think that would also make the pool of developers larger. [23:03:17] tobias47n9e_: I doubt it would [23:03:23] In any case it's not really feasable [23:03:47] Besides, the language is not the problem in case of MediaWiki and almost all code out there [23:04:12] addshore: yay for removing code from lib [23:04:21] actually, 3 classes gone [23:04:23] 4 [23:04:27] addshore: we should recreate the graph though. I bet it went up a bunch since we moved out DMS [23:04:29] well, with tests 6 [23:04:39] addshore: rate my new awesome PHP lib https://github.com/JeroenDeDauw/OhMyPhp [23:04:45] yeh, you have a spreadsheet or soemthing? [23:05:13] idk [23:05:18] think bene made something [23:05:22] should find the email [23:05:29] yeh! [23:08:32] I found it [23:08:37] not sure how to edit it.... [23:09:11] also not sure if it is classes or files [23:09:13] benestar|cloud: ^^ [23:11:07] if it is files in includes we went from 102 to 119 :P in 3.5 months [23:24:42] benestar|cloud: gimmeh teh scriptz! [23:24:54] probably one line of magical git stuff