[00:26:31] New patchset: Krinkle; "Implement mw-doc-gen.sh tool" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/39212 [00:33:24] Krinkle: hi :) We got a bug in Zuul [00:33:36] Krinkle: so that script is pending the bug fix :-D thanks for the fix ! [00:58:08] hashar: bug fix? [00:58:21] should I be worried [00:58:24] what bug [01:04:07] hashar: thanks, for your work on Sartoris today [01:04:34] Krinkle: I was referring to the shell scrip that will be used to generate online doc :-] [01:04:42] preilly: you are welcome :-D [01:04:45] k [01:05:33] preilly: unfortunately I am of no help to write unit tests for sartoris :-D [01:06:27] hashar: no worries… and they're coming soon [01:13:07] preilly: we don't have tox on the contint server yet :/ Will have to package it for Debian [01:13:17] it is not going to take long thanksfully [01:13:45] hashar: okay great [01:22:05] beeed time [01:22:08] * hashar waves [02:01:21] TimStarling: ping [02:03:27] yes? [02:03:53] TimStarling: Did you figure out who misspelled subtle? [02:04:13] no, I didn't bother [02:04:17] k, so that was me. [02:04:19] BUT! [02:04:43] I did not design that style nor implemented this crappy interface [02:04:53] ok [02:05:13] I ported the hardcoded "table.TablePager >crappy selector" from shared.css into a css class and added it to the list [02:05:20] ideally this thing is a separate property or an array [02:05:28] which Html::element etc. support now [02:05:35] as class [02:05:53] TimStarling: btw, I don't remember, why is getTableClass not overridable from a subclass? [02:06:11] it is overridable [02:06:26] $tableClass = htmlspecialchars( $this->getTableClass() ); [02:06:30] $s = "\n"; [02:06:49] you just can't remove mw-datatable by overriding it [02:06:57] Right [02:07:24] what is a datatable anyway? [02:08:01] a table of data [02:08:06] wee [02:08:29] that was my suddle sarcasm [02:08:37] lol [02:08:41] ok, enough of that [02:08:52] wikitable was taken, mw-tablepager seemed inappropriately specific as the style has nothing to do with paging. [02:09:12] right, so just a random name [02:09:19] and replacing it with wikitable was objected at the time. people wanted the tablepager tables to be blue due to the iconography attached to those. [02:09:34] because it still has harcoded buttons [02:09:44] yeah, the whole thing is a mess. [02:10:05] they wanted everything to be blue because the buttons were blue? [02:10:05] ideally it'd just use wikitable and some sortable interface over ajax to page and sort. [02:10:09] ... [02:10:45] that's how it was, and as a nontrivial code cleanup to phase out the legacy stylesheet I didn't feel like facing all that drama so kept the style as-is. [02:11:15] right [02:14:06] !log deleting sysadmin and netadmin roles from ldap [02:17:33] heh. wrong channe [05:27:06] New patchset: Pastakhov; "Add jobs for mw/ext/MultiMaps" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/46495 [05:28:30] siebrand: ^^^ hi, I did the right thing? [07:56:26] hello [08:10:26] so, fun times, the CentralNotice deploy had a bug. unless someone else is doing anything, I am going to push a quick fix [08:14:40] pgehres: I am not doing anything this morning :-] [08:15:22] hashar: lol, it's not quite morning here, anything but, in fact :-) [08:15:35] ah timezones .. [08:15:46] so yeah, EUROPEAN mornings are usually very quiet [08:16:05] though I do sneak some changes from time to time [08:16:26] i am only doing this because it is mwalker's bug and I deployed it earlier [08:16:38] and, since he will be up a while, he can wake me back up if there is an issue [08:17:05] because I decided to be silly and take a 'nap' at 8PM [08:17:19] which turned into a glorious four hour sleep fest [08:17:29] pgehres_: make sure to get the patch in core if you haven't cherry picked it [08:17:42] oh no [08:17:45] that is an extension update [08:17:46] ahah [08:17:48] I am lame :/ [08:17:54] no need to cherry pick, we deploy master ;-) [08:18:12] ... because we're even more silly about CentralNotice than I am about picking sensible sleep times [08:18:33] (because apparently, they haven't decided how to do feature branches on the cluster yet) [08:21:00] and … we are done [08:21:05] pgehres_: if you have any idea, please expose them on eng list or on wiki or something :-] [08:21:10] I guess several people could use support branches [08:21:29] oh, we had a long discussion with demon [08:22:14] something about issues pinning to a branch other than master when they do the bi-weekly releases [08:25:40] pgehres_: we should stop the bi-weekly releases :D [08:25:48] eh, I like them [08:25:51] and just deploy continuously as changes are merged [08:25:56] oic [08:25:59] ;-D [08:26:07] but that is my s3cr3t evil plan [08:26:35] ultimately I would want master to self deploy on the cluster [08:26:47] and from time to time, the ability to merge so called feature branches [08:26:58] that would have their feature disabled by default / not setup [08:27:10] could be a lot of fun [08:27:13] which we could then try out on some servers before merging in master [08:27:32] typical example: a change introducing some evil sql schema changes can't really be deployed automatically [08:27:46] so yeah, maybe next year we will be able to do more stuff automatically :-] [08:31:48] that sounds like a lot of fun to implement [08:33:11] pgehres_: yeah my team has a lot of fun constantly :-] [08:33:27] (mw/core team, same as demon, reedy, tstarling..) [08:44:45] New patchset: Hashar; "ignore PSR2.Classes.PropertyDeclaration.Multiple" [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46499 [08:47:41] New patchset: Hashar; "disable Generic.Strings.UnnecessaryStringConcat.Found" [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46500 [08:47:58] Change merged: Hashar; [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46499 [08:48:05] Change merged: Hashar; [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46500 [09:27:01] New patchset: Hashar; "PHP reserved variables never start with $wg, skip them" [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46502 [09:27:13] Change merged: Hashar; [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/46502 [12:31:03] re [12:37:18] hi hashar [12:51:08] hashar: are you coming to fosdem? [12:52:18] zeljkof: nop [12:52:26] zeljkof: I kind of forgot about it [12:52:43] hashar: too bad, looks like there will a few of us there [12:53:16] yeah should have taken care of it back in december [12:53:31] I am also heading SF at the end of february [12:53:49] hashar: see you in SF then :) [13:05:22] Json exception: It appears you have an extra trailing comma [13:05:34] the number 1 reason I hate JSON [13:05:59] or you're too used to PHP? [13:07:02] perl does the same [13:07:08] I bet python as a similar behavior [13:09:24] aha [13:11:47] what I hate JSON for is that it's too restrictive [13:12:11] for example, all Unicode has for be encoded as \uXXXX [13:12:23] which takes more space [13:12:24] yeah cause you never know how it is encoded [13:12:31] the spec just does not have anyway to handle that [13:12:39] we should write json 2.0 [13:12:57] or migrate the web to YAML (which in turn could use a standard) [13:12:59] or [13:13:01] XML !!!!!!!!!!!!!!!!!! [13:13:18] in PHP 5.4, there's actually an option to json_encode() to not escape Unicode [13:55:46] JSON_PRETTY_PRINT [13:55:47] bahhh [13:55:52] only in > 5.4.0 [14:06:08] ^demon: hey tom [14:06:13] <^demon> Howdy [14:06:22] ^demon: if one want to create a new branch in a repo, how would he does it ? [14:06:26] he === me [14:06:26] ::D [14:06:35] should I get myself the right to push new references [14:06:43] or create the branch in the GUI ? [14:07:25] <^demon> You can just do it in the GUI. [14:07:37] <^demon> But both require "Create Reference" -- which you should have on most repos probably. [14:08:01] niiice [14:08:14] will try out and let you know [14:36:37] Change merged: Erik Zachte; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/41979 [14:53:31] does anyone know when the WMF switched from [[image:example.jpg]] to [[file:example.jpg]] ? [Not urgent] [14:54:33] <^demon> Dragonfly6-7: Dec. 2008. [14:54:36] <^demon> https://bugzilla.wikimedia.org/show_bug.cgi?id=44 [14:56:28] hmm [14:59:42] file, fichier, datei, archivo... how many language-equivalents does mediawiki support? [15:02:29] anyone know? [15:10:43] New patchset: Hashar; "test pipeline for mw/ext/Translate" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/46523 [15:11:14] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/46523 [15:15:47] Dragonfly6-7: how many language do we support? [15:15:57] roughly 365 including variants [15:16:25] so probably less [15:57:48] JeroenDeDauw: the repo is on labs at http://integration-composer.instance-proxy.wmflabs.org [15:57:56] JeroenDeDauw: highly experimental though [15:57:59] will poke it later on [15:59:27] and I am out! [15:59:27] * hashar waves [17:01:05] Bug day time! [17:02:26] Welcome everybody here to our first bugday this year! [17:02:33] More info: http://lists.wikimedia.org/pipermail/wikitech-l/2013-January/065792.html [17:02:39] We're going to use http://etherpad.wmflabs.org/pad/p/BugTriage-2013-01 for documenting [17:02:59] and the full list of items in the bugtracker that everybody is free to take a look at it is here: [17:03:21] https://bugzilla.wikimedia.org/buglist.cgi?bug_severity=blocker&bug_severity=critical&bug_severity=major&bug_severity=normal&bug_severity=minor&bug_severity=trivial&chfieldto=1y&f1=longdescs.count&list_id=176394&o1=lessthaneq&resolution=---&v1=1 [17:03:27] (and it even fits in one line, yay) [17:03:44] guessing a lot of these will become WONTFIX? [17:04:28] If we're really against them, it might happen. [17:04:35] otherwise I'd just set low priority. [17:04:43] ok [17:05:00] some might be even superseded by newer technology, e.g. I closed a ticket today that was "fixed" by the TimedMediaHandler deployment [17:05:30] right. my favorite reason for marking a bug CLOSED is "OVERTAKEN BY HISTORY" :-) [17:09:02] * andre__ updates the documentation for "preparing a bugday", realizing that sending out reminders is something to document [17:09:15] https://www.mediawiki.org/wiki/Bug_management/Meeting_preparations that is, for anybody interested [17:09:32] :-) [17:11:28] * andre__ looking at the long list of reports and sorting them by product and component to drop them in the etherpad at http://etherpad.wmflabs.org/pad/p/BugTriage-2013-01 [17:19:20] hmm, I think I've found an acceptable format for the etherpad [17:19:32] chrismcmahon: around? [17:19:40] hi hexmode [17:20:07] so, q about testing node.js ... any ideas that have come from VE or before? [17:22:29] so I'm looking at https://bugzilla.wikimedia.org/show_bug.cgi?id=12263 , about 1.9, and states "not reproducable on any Wikimedia sites" [17:22:46] it's about a special page that calls delete functions directly. [17:23:18] and I'm not sure if I'd classify that as a "bug". Could be an enhancement as I'm not sure if Special Pages are expected to interact like this. [17:23:54] In any case, I'll ask the reporter if there's a chance to get a minimal testcase I guess [17:23:58] marktraceur: ^ re hexmode's question - know an answer? [17:24:11] Erm [17:24:23] heh... [17:24:28] hexmode: Parsoid's testing with node, what do you want to know? [17:24:37] subbu, CC [17:24:49] marktraceur: do you have tests running in jenkins? [17:25:14] hexmode: We did, but only really hackishly. We set up a labs instance to respond to HTTP requests with a JUnit XML file. [17:25:29] hexmode: They're broken now and I haven't been back to fix them since [17:25:58] hexmode: Though now that the jenkins host has been upgraded we could probably run them more easily.... [17:26:00] ah... guess I'll talk to hashar and see if he has any more ideas [17:26:07] ? [17:26:24] garrr... http://etherpad.wmflabs.org/pad/p/BugTriage-2013-01 says "The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later." [17:26:25] So much for collaborative editing. [17:26:45] hexmode: IIRC the host was running Lucid, which had some insufficient node packages or something [17:26:51] andre__: Oh noes! Allow me. [17:27:18] andre__: WFM [17:27:22] hexmode, are you looking specifically at jenkins? or commandline as well? [17:27:24] hexmode: are you testing node itself or an application using node? [17:28:03] I was looking at testing a node.js application and wanted to see if there was anything particular for jenkins [17:28:14] and also, if it has nUnit testing [17:28:37] ah, etherpad works again now too [17:28:57] ah, i see, this is is a generic node.js qn. not necessarily related to parsoid/ve. as marktraceur mentioned, our jenkins integration is currently broken, so we rely on commandline runs of tests. [17:29:04] andre__: Yeah, I got disconnected earlier. [17:30:10] right, not parsoid/ve... just thought that with you're experience with it, you might have insight [17:30:12] for parsoid specifically, we are focused on parser tests to maintain compatibility with the php parser -- so we have a parser test script that runs the php parser tests - but no other unit tests currently. [17:30:21] hi sumanah [17:30:25] ah [17:30:33] hexmode: quick google showed me this, might be of interest: http://debuggable.com/posts/unit-testing-with-node-js:4b647d40-34e4-435a-a880-3b04cbdd56cb [17:30:37] so, we haven't looked into unit testing setup, but i would be surprised if there weren't any already [17:30:41] subbu: Also the RT testing. [17:30:47] Not unit testing, but testing [17:30:51] hi Nikerabbit [17:30:52] ah, there you go :) chrismcmahon has an answer for you. [17:31:19] marktraceur, right. [17:32:02] marktraceur: right, RT as well... [17:32:29] chrismcmahon: ty, looks like I suspected [17:32:33] so yeah, testing a node app you'd want something listening on a port to check that the server is sending expected stuff [17:32:37] Hm, lingo ambiguity. [17:32:57] hexmode: lingo.parsoid.RT = "round-trip" [17:33:16] * andre__ looking at datasets bug https://bugzilla.wikimedia.org/show_bug.cgi?id=27115 but having problems to follow the steps [17:33:18] heh [17:33:19] lingo.ops.RT = "request tracker" [17:33:22] regression testing [17:33:29] marktraceur: language about testing is awful. and the Wikipedia pages about software testing are more awful. [17:34:35] {{sofixit}} [17:41:02] do you need help creating more bugs? ;) [17:41:05] valeriej, Re bug 25327: https://en.wikipedia.org/wiki/Special:Preferences#mw-prefsection-gadgets [17:41:16] There you can enable WikEd, I think [17:41:24] Nikerabbit, yes, spam us! ;) [17:42:46] Ah, thanks andre__! [17:43:12] valeriej, I found that info in the infobox on the extension page of WikEd [17:44:40] andre__: Oh, I see it. [17:47:32] Trying to reproduce https://bugzilla.wikimedia.org/show_bug.cgi?id=32144 and succeeding [17:48:01] When I log in via HTTP and log out via HTTPS, I'm still logged in via HTTP. Heh. [17:48:20] * andre__ updating the report by confirming and mentioning the current MediaWiki version (1.21wmf8) [17:49:04] but wondering if that's really a MediaWiki bug or rather a Wikimedia bug. Hmm. [17:51:25] andre__, just realized that I didn't advertize Bug Day in social media. Do you still think it's worth doing it right now? [17:51:48] qgil, we're still here for five hours, so feel free to go ahead [17:51:51] andre__, also, a reminder to wikitech-l? (or was it sent? I'm a bit behind email lately) [17:51:57] I've sent it [17:52:01] ah ok :) [17:52:08] (on the other hand as this is the first bugday for a while, I don't expect many people to pop up) [17:52:37] plus I'm just trying to think aloud here on this channel, to make it a bit more transparent what we're doing and what triaging means [17:54:11] wondering if fixing bug 32144 depends on bug 20643 [17:54:32] (still on the HTTPS login vs HTTP cookie issue) [17:56:54] andre__: Those bugs link to bugzilla.mozilla.org, btw. [17:57:05] uh? [17:57:09] which ones? [17:57:39] andre__: Oh, I bet it's my IRC client, nevermind. [17:58:11] I just wrote "bug XYZ" here (replace XYZ by numbers). Heh, probably, yeah [17:59:26] andre__, valeriej Twitter, FB, G+ and identi.ca covered. Now going for mediawiki.org home. Your RT / share welcome [17:59:40] qgil: Thanks! [17:59:41] qgil, thanks! [18:00:45] Also remember: anybody interest in the Bug Day could watch / join / promote https://www.mediawiki.org/wiki/Groups/Proposals/Bug_Squad [18:01:00] this way every Bug Day will be simpler to promote and reach critical mass than the prevous one :) [18:01:21] yurik, around? [18:01:47] good point! [18:02:22] ^demon: BTW, thank you very much for the repos. [18:02:32] <^demon> yw [18:06:02] * andre__ takes a look at https://bugzilla.wikimedia.org/show_bug.cgi?id=22333 and clarifies the summary [18:06:49] andre__, valerie, now at the http://mediawiki.org homepage as well [18:06:59] yay [18:07:29] andre__, valeriej maybe a better landing place needs to be prepared for next Bug Days. Just a thought, and I'm not even sure about it. Not a discussion for today. :) [18:07:48] qgil, I wrote that done today in the meeting, as I loved your VE testing landing page. [18:08:16] s/done/down [18:08:36] good [18:10:20] hmm, so if somebody sets custom stuff in LocalSettings.php and forms don't get submitted anymore, is that a valid bug? [18:10:26] * andre__ wondering looking at bug 22333 [18:13:21] 29472 looks more like an enhancement to me, resetting severity [18:13:47] andre__: I enabled WikEd in preferences, but the toolbar doesn't show up when I edit a page. I disable WikiEditor as well, and it still doesn't show up. I guess I'll just ask bug 25327 is still an issue. [18:14:11] hello there! [18:14:23] valeriej, I'd mention these steps that didn't work, and also how to test it. [18:14:27] heja matanya! [18:14:29] welcome :) [18:14:49] * andre__ once again throws http://etherpad.wmflabs.org/pad/p/BugTriage-2013-01 into the room for those that know it's a bugday [18:14:54] thank you, glad to meet you [18:15:13] andre__: Will do. matanya: Hello! [18:15:21] hi valeriej [18:15:37] any thing for me to look at? [18:15:55] matanya, haha, a lot :) [18:15:58] though I done my bug stuff for today :) [18:16:08] see 44470... :P [18:16:18] anything from that list in etherpad that looks interesting to you - free choice! [18:16:42] matanya: You can add your name in the etherpad by clicking the person-icon in the upper-right corner. [18:16:44] matanya, yeah, saw that report. Thanks for spotting & reporting [18:17:21] done. you should thank mathinus [18:17:26] he found it [18:17:45] but he is afraid from bugzilla [18:18:25] matanya, meh. I'd love to know what makes mathinus afraid, because I'd love to make Bugzilla less scary in the long run. [18:18:58] catch a talk with him, he is often on #wikimedia-stewards [18:19:15] ah, great [18:19:28] and I misspelled his name [18:19:34] it is Mathonius [18:20:05] very friendly fellow (and running for stewardship) [18:21:39] * andre__ resets 29472 to "enhancement" and moves it to "Parser" component [18:22:17] I'll be back in 30min, time for dinner [18:25:21] bone appetite [18:27:08] hi matanya - nice to see you! [18:27:18] hi sumanah :) [18:27:25] glad to see you again [18:36:35] hey notpeter how are you? [18:38:13] <^demon> qchris: So, I played around with https://gerrit-review.googlesource.com/#/c/41760/ earlier. Good news: I was able to delete a group member. Bad news: I wasn't able to delete any more group members. [18:38:30] <^demon> So, this is obviously *touching* my problem, if not fixing it. [18:41:24] sumanah: I'm alright. what's up? [18:41:45] ^demon: great! Thanks. [18:42:04] notpeter: today I aim to finish my draft of https://meta.wikimedia.org/wiki/Wikimedia_Blog/Drafts/Ops,_2011-2013 which I've divided into 2 parts, 1 more about the puppetization, and 1 more about profiling and monitoring [18:42:25] just wanted to give you a heads-up - I hope to get it to you, binasher and the rest of Ops (and Guillaume) for review! [18:42:35] and congrats again on the basically flawless dc migration [18:43:03] sure! I'm taking a redeye tonight, so I will have way too many hours to read and comment, and then tomorrow I can make those comments make sense [18:43:04] thanks! [18:43:31] much credit to mark and binasher for flawless execution, and whole ops team for the.... years of work getting us here :) [18:43:53] <^demon> sumanah: "One of the servers" in that picture caption should probably be "Some of the servers" [18:43:53] TOTALLY. [18:43:57] <^demon> Lots of servers in that pic :) [18:44:32] ^demon: Fixed. [18:45:03] ok, now, someplace outside the home, for a change of scenery. [18:46:03] I just watched VE erroneously nuke a whole chunk of text, but I'm not finding a consistent repro, phooey. [18:49:34] Enable screen recording by default? :P [18:51:00] * andre__ triaging https://bugzilla.wikimedia.org/show_bug.cgi?id=21944 [18:52:53] andre__: I couldn't reproduce this bug: https://bugzilla.wikimedia.org/show_bug.cgi?id=16005 Could I have just closed it as WORKSFORME, or it's best to give the reporter the chance to test it? [18:54:32] valeriej: I'd give the reporter two weeks or so to answer (though it's always unlikely with such old tickets) and without answer I'd go ahead and close it as WFM [18:55:29] andre__: Good to know, thanks. So after this bugday, we'll go through any reports like that about 2 weeks from now and close all non-answers? [18:56:11] Sounds like a plan, yes. 2 or 3 weeks [18:56:26] Ok. [18:56:27] other projects say 6 weeks, but Wikimedia doesn't seem to have a rule for this [19:03:24] wondering how to retest morebots i18n bug 21944 - probably by writing "!log á" in #wikimedia-operations but I don't dare :) [19:06:11] anred__: I'm interested in taking on that bug, or I can set up a test bot for you if you want to experiment [19:06:20] *andre_ that is [19:07:11] andrewbogott: I'm only curious if that's still an issue, so depends if it's worth the trouble for you [19:07:30] going through old bug reports that have not been touched for more than a year as part of a bugday [19:07:54] Lemme see if I still have a test system running... [19:12:12] andre__: Meet me in #wikimedia-bottest [19:17:05] looking at https://bugzilla.wikimedia.org/show_bug.cgi?id=22356 - I guess that's a case for support instead, PHP XML parser issue for a DOM depth > 256? [19:17:47] does not look like something Wikimedia specific but PHP's XML parser, however it's triggered by a WM datadump so wondering whether something could be fixed in creation of datadumps. [19:18:55] <^demon> Not really, no. [19:19:55] <^demon> We could possibly add an option to MW core to use XML_PARSE_HUGE, but you probably wouldn't want it on by default. [19:21:08] https://bugzilla.wikimedia.org/show_bug.cgi?id=12221 mentions the error happens on systems not using Tidy. From what I can gather MediaWiki can be configured to use Tidy. How should MW handle bad HTML? [19:22:17] <^demon> valeriej: That's why mediawiki has tidy support, to deal with bad html. [19:23:18] ^demon: That's what I thought. So should I just close this as INVALID or WONTFIX? [19:23:31] <^demon> I'd say WONTFIX. [19:23:33] I think that's WONTFIX. [19:23:51] ^demon, andre__ : Ok, thanks. [19:24:58] <^demon> andre__: On the subject of IRC.... [19:25:00] <^demon> "Freenode Cloak/channel requests etc should go to #wikimedia-ops or to meta." [19:25:07] <^demon> Why would we send cloak requests to ops? [19:25:10] <^demon> That's always meta. [19:25:30] ^demon: haven't you got a plan to offload gerrit by moving anonymous git access to a secondary host ? something like git.wikimedia.org [19:25:37] ^demon, I have no idea. Let me fix that description. [19:26:10] <^demon> hashar: I've thought about it, but never made any real plans. [19:26:28] ^demon: ok was wondering. [19:26:54] <^demon> If we only replicated heads and tags, we could possibly do something. [19:27:19] well I am working on integrating Composer [19:27:34] which takes care of everything for us granted the composer.json properly describe tags / branches etc [19:27:43] <^demon> Not terribly concerned though. Load on Gerrit's been fine. [19:28:03] yeah maybe I should just use gerrit for now [19:28:23] <^demon> Or those repos I replicate for you on gallium? ;-) [19:28:24] I am going to use Composer to load up extensions dependencies in jenkins jobs [19:30:42] hashar: any chance of the composer thing working for wikibase? [19:31:03] aude: no way :-] [19:31:07] heh [19:31:19] soon? [19:31:21] aude: I will eventually (like by June) create a basic composer repository [19:31:26] hmmm, ok [19:31:30] aude: and will let extensions author update the composer.json [19:31:37] nice [19:31:46] aude: regarding Wikibase / Jenkins it does fetch the dependencies but that use a lame shell script [19:32:03] yeah [19:32:47] how's the bug triage going? [19:35:26] Hey visual editor people... I've been asked to evaluate this 'upload files from edit page' proposal: http://meta.wikimedia.org/wiki/Grants:IEG/Easy_Media_Uploader [19:35:39] is any of this covered in the new visual editor already? [19:35:57] i.e. does it support file uploading? [19:36:11] James_F: ^ [19:36:37] aude: It's going, ha ha. There are over 200 bugs, so I think we'll only make a small dent. It's better than nothing, though. [19:36:53] we've already gotten 15 out of the list :P [19:37:22] but first bugday is more about learning. Like having a better intro page for next time, and where to announce etc. [19:39:26] * andre__ can reproduce https://bugzilla.wikimedia.org/show_bug.cgi?id=33679 , moving to AFT [19:40:38] \O/ [19:41:07] nice that the bugs are getting some attention :) [19:41:41] Hmm. It's interesting that the user filed an Artivle Feedback bug under "User survey". Makes sense, somehow. [19:42:53] Krinkle: ^ [19:44:32] * andre__ gets sidetracked into looking at open bug reports in the "Wikimedia > User Survey" component and expects more misfiled "Article Feedback" bugs there. [19:44:39] Yeah. Sigh. [19:44:47] kaldari: i'd be surprised if visual editor covers uploading (yet) [19:44:51] * aude hasn't seen it [19:45:01] I haven't either, I just wanted to make sure [19:45:17] how is easy uploader different than upload wizard, or is it a batch uploader thing? [19:45:36] it's an upload interface from the article editing interface [19:45:38] * aude shall read the proposal again [19:45:48] "For issues relating to the general Wikipedian user survey." -- Does that refer to http://meta.wikimedia.org/wiki/General_User_Survey ?? [19:45:59] ^^ question to the elder folks here. [19:46:06] like the add media wizard (though never really polished, etc.) [19:46:57] <^demon> andre__: Likely. I don't remember mucha bout the survey. [19:47:25] maybe it's generally about http://survey.wikimedia.org/ [19:47:48] I love ambiguous unclear Bugzilla component descriptions. [19:47:49] kaldari: Hey. [19:47:59] kaldari: VE currently does nothing with file uploading. [19:48:06] <^demon> andre__: Please try to avoid reminding me that survey.wm.o exists. Limesurvey makes me cry :( [19:48:08] New patchset: Hashar; "jobs for mw/ext/WikiEditor" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/46553 [19:48:15] <^demon> andre__: ;-) [19:48:23] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/46553 [19:48:36] ^demon: Sorry. Not. ;) [19:48:38] kaldari: That said, it's sort-of on our roadmap, but as a "stretch" goal as an extension to our media handler (which we also don't have yet). [19:49:14] James_F: thanks, that's helpful [19:49:37] kaldari: Integrating our UX concepts (and UI artefacts) into their might be a mess, though. [19:49:47] kaldari: But, frankly, it'd be a Nice Problem To Have(tm). [19:50:02] kaldari: TrevorParscal may have some thoughts. :-) [19:51:00] * aude thinks "approx. one month project with two 2-week sprints" is ambitious but who knows.... [19:52:10] Yeah. [19:57:41] Be back in about 30 min. [20:00:09] enjoy your meal. :) [20:00:47] * andre__ getting back to triaging old bug reports, after a trip to ArticleFeedbackTool land [20:06:47] bug 30791 (wrong language link in UploadWizard) feels like a dup to me, now I just need to find it... [20:07:05] MaxSem: yep [20:07:08] !b 30791 [20:07:08] https://bugzilla.wikimedia.org/30791 [20:07:45] yurik, can you give me an aesthetical advise? does https://bugzilla.wikimedia.org/show_bug.cgi?id=43674 sound right? [20:08:00] andre__: Never seen that before. We've dealt with language-wise license links, but I don't know about the old upload form one. [20:08:44] andre__: https://bugzilla.wikimedia.org/show_bug.cgi?id=29791 I stand corrected [20:08:56] MaxSem: reading... [20:09:39] marktraceur: garr, I searched for "old form" in comments. Couldn't find that one as it has "old form" in the summary. Thanks for finding it. [20:09:41] New patchset: Hashar; "triggers for mw/ext/WikiEditor" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/46559 [20:09:51] My pleasure. [20:09:57] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/46559 [20:10:09] * marktraceur decides to keep the highlight on "UploadWizard" for now :) [20:11:01] * chrismcmahon isolates and repros a data-loss bug in VE [20:11:06] * chrismcmahon does a little dance [20:11:45] chrismcmahon: awesome. Congrats. [20:12:13] andre__: I'd been seeing it from time to time, couldn't quite nail it down until just now [20:12:21] MaxSem: yes, looks ugly [20:12:49] yurik, so this bug is worth implementing? [20:12:55] i think so [20:12:59] thanks! [20:13:31] MaxSem: i agree the style is bad, the question is if it is being used :) [20:13:43] sec, checking the stats [20:13:55] well, this extension is not officially live yet:) [20:14:17] so I feel completely confident about breaking b/c [20:14:37] is there a plan to integrate with wikidata? [20:14:52] because this stuff should be stored there, not inside the article [20:14:56] yes, some time in the future [20:15:05] sigh :) [20:15:19] ok, sec, let me see what graphite said [20:15:23] yurik: i'm sure we could grab the same data with the parser function [20:15:44] aude: not sure what you mean [20:15:45] * aude one of the wikidata developers [20:15:45]