[07:27:38] morning! [07:28:43] hey joakino -- just getting ready for work [07:28:47] back in a little while [07:29:34] hi phuedx ! I'm doing some cljs for a little bit, i'll start with work in ~30m [08:16:08] have coffee [08:16:09] work time [08:17:23] gonna do some sign off while running the browser tests [09:18:18] phuedx: how was the weekend [09:18:29] good thanks [09:18:38] an old friend spent most the weekend with us [09:18:45] curtain shopping for george [09:18:48] that's about it [09:19:01] went to the local farm [09:19:03] you? [09:19:26] nice [09:19:59] i spent it on a house in the countryside with the pool, it was good [09:20:04] too hot, but nice [09:20:22] what did you do at the farm? [09:20:31] buy food? show the animals to the kids? [09:21:13] both [09:21:24] the bacon that they sell there is /amazing/ [09:22:44] i can only imagine :D [09:22:53] phuedx: hey what happened with the browser tests [09:24:02] both MF and gather are super broken [09:36:26] joakino: in the email reports? api calls failing [09:36:48] :( [09:37:08] running them on my machine against labs isn't great but they do work [09:37:15] i think wmflabs is a little borked at [09:37:16] *atm [09:37:34] the login step is failing because centralauth is saying that there's no token blah blah [09:37:43] and the step fails because it doesn't know which link to click [09:38:14] once i've signed off T94713 i'm going to run 'em locally [09:38:17] and report the results [09:57:08] thx phuedx [09:57:32] btw when you have a moment we should talk about this afternoons prioritization/backloggrooming met [09:57:35] meeting [09:57:51] joakino: want to do it now -- i'm in the middle of a reload [09:57:58] --provision [09:58:04] lol [09:58:46] ok let me grab a shirt https://media2.giphy.com/media/8ylZxTWDCJ2AE/200.gif [09:58:58] lol [10:38:34] phuedx, joakino: hey guys, was mobile ever able to edit whole pages? or is that a regression? [10:51:25] bmansurov: by clicking the first edit icon? for as long as i can remember that's edited the first section [10:51:28] or is this another thing? [10:52:08] also, hey [10:52:20] phuedx: yes, the first edit icon. I was hoping that would allow the user to edit the whole page, but it doesn't. [11:00:35] bmansurov: that last review is totally -1 worthy, you're right that more context is needed [11:01:19] phuedx: ok [11:02:06] bmansurov: nice job on the search descriptions, verified locally and on bc in stable/beta [11:02:07] *woo* [11:02:14] yay [11:06:36] bmansurov: i missed the start of the chat at friday's stand up about this T103213 [11:06:46] (and the middle too) [11:07:26] phuedx: basically rob mentioned that the team had already spent too much time on the bug [11:07:52] phuedx: and their fix was good enough so I abandoned the patch [11:08:01] okie poke [11:30:31] going for lunc [11:30:33] *lunch [11:30:41] responding to jonk's email about initial browse data [12:03:12] back [12:22:40] bmansurov: okie, i'm going to resolve T103213 [12:22:44] thanks for the discussion :) [12:22:54] sure, np [12:42:55] going to take george and harry to the park for a while [13:33:21] dbrant: I was looking for you! [13:33:23] https://phabricator.wikimedia.org/T104190 :) [13:33:31] dbrant: also yay for tabs in the app :) [13:33:44] YuviPanda: o/ [13:33:53] awesome drone video, btw :0 [13:34:24] lol just starting to mess around with it... [13:34:59] YuviPanda: that's probably better answered by bearND|afk... [13:54:51] dbrant: alright! [14:26:42] yo [14:53:33] hi benestar [14:53:34] err [14:53:35] sorry benestar [14:53:37] i meant bearND [14:53:47] Hey YuviPanda dbrant: what drone video? [14:54:22] YuviPanda: to answer you're question, we currently still need your alpha build set up [14:55:02] since we don't have the Jenkins apks build with a stable cert and published to a stable location. [14:55:18] It also doesn't have the metadata.json stuff you added [14:56:17] bearND: right. [14:59:17] hay phuedx [14:59:34] yo [14:59:35] kristenlans: [15:02:32] kristenlans: how was yer week? [15:08:21] phuedx: it was pretty good! Busy. [15:09:42] phuedx: joakino I changed the prio meeting this AM/PM to start 1/2 hour earlier so we can catch up on workboard changes and chat about how to kick off Q1 with a BANG. Does that work fro you guys? [15:10:37] kristenlans: i said yes [15:10:58] Yes. Yes you did phuedx. [15:12:22] bearND: can you respond on the ticket? I"ll attempt to rescue that instance, if not I"ll have to spend time rebuilding it :( [15:13:44] kristenlans: just responded 👍 [15:14:48] YuviPanda: will do. I'll also update the ticket which we'll need to eventually replace that setup. [15:18:10] morning coreyfloyd! Were you takgin off any time around 4th of July long weekend? [15:18:40] bearND: alright. I'll try to rebuild it this week. can you also have pointers to your CI setup? I might be able to steal stuff from there [15:21:40] kristenlans: hey - just the friday holiday [15:21:55] why whats up? [15:25:23] bearND: hey, question: [15:25:34] coreyfloyd: just trying to figure out schedulign for estimation mtgs, since Friday's a holiday [15:25:38] bearND: if I'm making modifications to java-mwapi, what's our current procedure for using the working copy of java-mwapi for testing in the app? [15:27:34] dbrant: you'd change the version in the pom file of java-mwapi, and make the corresponding change in our build.gradle file [15:30:52] bearND: got it, thanks [15:36:54] YuviPanda: I've just updated https://phabricator.wikimedia.org/T99115. Is this what you need? [15:37:28] YuviPanda: When you rebuild, are you doing that from a backup or does that mean from scratch? [15:37:36] don't know yet :) [15:37:50] worst case from scratch although the code is probably in a git repo somewhere [15:38:16] YuviPanda: I'm wondering if that means a new debug keystore (-> we need to tell users to uninstall the old alpha APKs) [15:38:30] bearND: indeed, that is possible - if I can't find the old keystore... [15:38:36] YuviPanda: yes the code is in a repo [15:38:44] I might be able to - this feels like a solid day of debugging / diving into stuff... [15:38:47] .me is looking for it [15:38:54] and I"m unsure when I can do that - atm everything works as well :D [15:40:12] YuviPanda: the repo is here: https://gerrit.wikimedia.org/r/#/admin/projects/labs/tools/wikipedia-android-builds [15:41:00] YuviPanda: I just hope we've pushed everything up there [15:41:35] heh, me too :) [15:42:10] YuviPanda: Good thing is that we don't really need the update Android SDK stuff anymore since we're using a Gradle plugin for that [15:42:35] bearND: aha! I might've found a way to revive it! stand by [15:46:38] dr0ptp4kt: ping (test) [15:55:09] dr0ptp4kt: ping (test again) [15:56:15] dr0ptp4kt: ping again (test) [15:58:58] dr0ptp4kt: ping one more time (test) [15:59:43] dr0ptp4kt: ping again (test) [16:00:22] dr0ptp4kt: final ping (test) [16:30:02] kristenlans: i updated the showcase description. think positive thoughts re: hangouts on air working [16:35:09] thanks dr0ptp4kt ;-) [16:35:57] kristenlans: i'm living in the future. i have an irc client with a bouncer that uses desktop notifications, and i can live broadcast events. it's a good day. it's a good day. [16:40:38] bearND: whops, thought I could recover it but apparenlty not :( [16:40:50] bearND: sorry! it looks like it'll have to be rebuilt. [16:41:13] YuviPanda: so we would lose the keystore file? Or can you still access it? [16:41:19] not sure. [16:41:50] bearND: ah, I can recover the keystore [16:42:07] YuviPanda: awesome [16:42:11] bearND: just everything else needs rebuilding tho :( [16:42:33] YuviPanda: how long would it take? [16:42:38] no idea [16:42:46] the first time it took me a couple of days but that was different... [16:42:55] and now you say a gradle build is just clone and build... [16:45:22] YuviPanda: just so i understand the impact: Looks like we cannot ssh into it. Since when are we experiencing issues? Are the builds still working? (We haven't merged anything since Saturday.) [16:47:05] YuviPanda: If the builds are still working I'd say we should focus on https://phabricator.wikimedia.org/T99115 and https://phabricator.wikimedia.org/T104207 instead. [16:47:27] bearND: ah, so I had made it a self hosted puppetmaster, and then not followed up some recent well announced labs changes because I forgot about it [17:01:34] rmoen: standup [17:01:54] kristenlans: oops [17:16:30] technical difficulties joining standup, be there asap [17:18:29] phuedx: are you signing off or are you available to run through In analysis column in 12 minutes? :-) [17:28:45] i'm gonna go to the pool for a bit before the late night meeting [17:37:53] joakino: the pool!! [17:45:09] jdlrobson: fyi https://phabricator.wikimedia.org/T104212 [17:54:55] mdholloway: niedzielski: bearND: can we have a quick huddle in batcave? [17:55:06] dbrant: omw [17:56:06] dbrant: sure, can you PM me the link again? i'm on a different computer than usual [18:11:54] dbrant: sorry, missed the message earlier [18:31:06] dbrant: I'll just be a minute. [18:50:10] dr0ptp4kt: let me know if you want to do any pre hangout testing for showcase (if that's even helpful) [18:50:28] kristenlans: ok, in about 5 mins [18:52:13] niedzielski: just rememberign that a saw a question from you about the reading showcase on Friday :P Did you find a satisfactory answer? [18:53:29] kristenlans: oh, it sounds like it's a free for all! are there any forbidden demos? [18:53:52] niedzielski: only the demo that can not be named [18:54:53] kristenlans: you mean... volDEMOrt [18:55:35] niedzielski: LOLOLOLOL you Android dementor [18:59:19] for reading folks: I'll be watching the showcase/staff meeting from the youtube stream to help with hangout size :) [19:00:49] kristenlans: can you share the link to the hangout? [19:01:44] kristenlans: got it [19:06:31] i'm viewing the meeting even though it looks like i don't appear to be [19:07:52] kristenlans: just tried joining and it's full [19:08:13] had some things to tidy up before i joined [19:08:20] am i being punished for tardyness? [19:08:51] phuedx you may have to watch vi th elink we're at max capacity [19:10:21] phuedx youtube link in invite [19:10:27] kristenlans: you got all quiet on the Youtube link for some reason btw [19:10:58] phuedx if you are presentign I can ask if someone who isn't presenting will swap a spot with you. [19:11:01] kristenlans: you were fine at the start but the last couple times you talked it was barely audible compared to everyone else :/ [19:11:11] jhobs I will scream when I come back [19:11:12] kristenlans: nope -- just watching for now [19:11:53] jhobs that actually may have corresponded with me being "presenter" to not beign presenter.... [19:12:25] kristenlans: maybe... but SF comes through fine while stephen is presenting [19:12:51] jhobs: hmm curious [19:13:49] jhobs: how was that? Did you hear me announce the time? [19:14:05] kristenlans: still very quiet :/ [19:14:31] niedzielski: painfully awesome, you mean [19:14:42] it's weird because you were like 300% louder before the 12:30 vs. 12:40 discussion [19:17:27] yeah still quiet even with nobody presenting... weird [19:20:36] Do I sound super quiet to peopl eon the hangout as well? bmansurov [19:20:47] kristenlans: no, you're good [19:21:13] kristenlans: in fact your mic is so good that i can hear you breath too [19:21:54] bmansurov: haha oh no creepy! [19:22:02] does phuedx have the same problem I do on the youtube link? [19:22:08] lol [19:22:27] jhobs: i just get "it's full" [19:22:36] bmansurov: glad this hangout is being recorded :P [19:22:44] phuedx: the youtube link is full? [19:22:45] phuedx: sorry, context: kristen (and only kristen) is very quiet for me on the youtube link [19:22:58] oh wait [19:23:02] i see now [19:23:27] what's weird(er) to me is that she wasn't quiet at the beginning of the meeting [19:28:27] kristenlans: well, regardless, I can make out what you're saying, so it's nbd :) [19:29:18] ok jhobs I'll troubleshoot w/IT afterwards for next time [19:29:49] kristenlans: hopefully (or maybe not so) the recording will be able to shed some light [19:44:18] kristenlans: it'd be great if non-staffers did drop off, for those of us who can't get in (and consequently don't have much of a voice) [19:44:23] non-staffers? [19:44:27] non-reading folk [19:46:49] phuedx: join [19:46:57] phuedx: jdlrobson dropped off [19:47:22] phuedx: jdlrobson dropped off if you want to join the hangout [19:47:29] phuedx: and there you are [19:47:37] as if by magic [19:48:06] I forgot to thank anomie (Brad Jorsch), he makes teh MediaWIki API go [20:16:43] dr0ptp4kt: it's great havign these showcases recorded. I'm watchign and pickign up a few things I missed the first tiem around. [20:17:08] thanks for wrangling [20:21:06] jdlrobson: with who and when is your interview scheduled with a ux candidate? [20:30:04] kristenlans: coreyfloyd mhurd bgerstle etonkovidova iOS kanban meeting time [20:37:23] dr0ptp4kt: do you know if we have data on how much hamburger menu is clicked compared to how much it is not clicked? [20:38:07] dr0ptp4kt: on mobile web [20:51:53] bmansurov:https://gerrit.wikimedia.org/r/#/c/220627/5 please note "Fixes to regressions should be as small as possible imo" can you pull out the fix into a separate patch? [20:52:09] the refactoring can happen post branch cut [20:53:24] jdlrobson: ok [20:54:04] kaity: i believe this *may* be captured via Schema:MobileWebClickTracking - https://meta.wikimedia.org/wiki/Schema_talk:MobileWebClickTracking#Main_Menu_click_tracking . i believe several queries would need to be run to figure out (1) how many pageviews there were and then (2) adjusting for whatever sampling rate is used, capturing the number of [20:54:04] "hamburger-home" events (plus visits to https://en.m.wikipedia.org/wiki/Special:MobileMenu , which is where user agents like Opera Mini go) [20:54:37] kaity: we only have clicks not non-clicks we cant measure at that scale [20:54:38] http://mobile-reportcard.wmflabs.org/#other-graphs-tab [20:54:50] the graphs you never use have useful information ^ :P [20:55:04] see page-ui-daily [21:02:23] jdlrobson: how do you even measure non-clicks if you could? page views - clicks? [21:03:08] bmansurov: only feasible would be to sample and log a page-impression event [21:15:09] dr0ptp4kt: hey adam, you coming to 1:1? [21:15:35] mdholloway: thanks for checking. apologies. here i come [21:15:41] no prob! [21:30:29] MaxSem: vagrant still exploding [21:30:36] failed on wikidata updating [21:42:57] oh oops, forgot to change my nick back [21:46:58] MaxSem: sigh still wikidata issues [21:47:39] ebernhardson: you managed to setup Cindy I see :D [21:48:21] https://gerrit.wikimedia.org/r/#/c/218993/ < would love to hear back from you in say a weeks time [21:51:36] mbinder: Hello:) are you in the office today? [21:52:51] etonkovidova: I am. Hangout or tea? [21:53:19] mbinder: Tea! Are you on the third floor? [21:53:34] etonkovidova: yes, next to vibha's desk, I'll be there shortly [21:53:58] mbinder: I will be there [22:08:55] mbinder: fan of eddie ezzard? [22:09:11] www.youtube.com/watch?v=BNjcuZ-LiSY [22:09:16] izzard* [22:30:38] bgerstle: I looooove eddie izzard [22:31:27] mbinder: just saw your comment to elena above, and made me think of that skit ^ [22:31:44] "cake or death" / "i'll have death—I MEAN CAKE!" / "nope, you said death first!" [22:32:10] heh, I love the star wars and imperialism bits too [22:32:51] "The force is strong with this one." "'ow strong?" "Well, about as strong as a small pony." [22:37:45] rmoen: your patch looks fine but i think we should add some browser tests given it is an important workflow [22:43:19] ok [22:43:45] jdlrobson: it will be good for me to add some tests [22:53:07] Hi! Is there a standard way for client-side JS to determine whether we're on the desktop or mobile site? [22:53:46] jdlrobson: rmoen: ;MaxSem? [22:53:49] THx in advance! [22:55:43] On a related note, is there any way in JS to check some header or something set by Varnish to determine which device (class) we're on (i.e., android, iphone, etc.)? We're just peeking at options for ditching this: https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FCentralNotice/8135c1783023c854d27a49000534b6f060cb254c/modules%2Fext.centralNotice.bannerController%2Fmobile%2Fdevice.js [22:57:57] AndyRussG, MF uses device-independent scripts/styles and Varnish doesn't detect device type anymore [22:58:01] AndyRussG: so you can do something like isMobile = mw.config.get('skin') === 'minerva'; [22:58:13] AndyRussG: I'm trying to hold off the urge of making a flippant remark with something about writing it yourself is the best way (aka reinvent the wheel) [22:58:20] Not a dig at you though :) [22:58:55] Reedy: rmoen: MaxSem: thanks! Reedy, heh, that's what it looks like someone did once [22:59:23] P.S. I'm cool w/ flippand remarks :) [23:00:11] K I'm going w/ rmoen's suggestion. I'm refactoring but will take more or less the code from that module (linked above ^). The problem I had is that there's no "desktop" device in that module, since it's only loaded on mobile [23:00:20] rmoen: Reedy: MaxSem: [23:00:23] ^ [23:04:00] MaxSem: rmoen: Reedy: BTW, while I have your attention: https://phabricator.wikimedia.org/T103695 :) [23:12:51] AndyRussG: hey [23:13:02] jdlrobson: hey ho! [23:13:06] so... [23:13:11] what if i told you.. [23:13:14] there was no mobile. [23:13:21] ^ :) [23:13:38] heheheh ;) we're in MS Surface heaven [23:13:39] Whatever you do for mobile should also benefit desktop [23:13:54] so i think the real question here is why do you want to know if you are in the mobile skin [23:13:54] jdlrobson: yes, save that thought...... [23:14:08] for the _real_ CentralNotice refactor [23:14:16] Yes, correct [23:14:49] I'm prettifying much CN code and making a new bannerController API while preserving current functionality [23:14:59] AndyRussG: so.. you can abuse targets for this [23:15:12] addModules( 'centralnotice', 'centralnotice.mobile' ) [23:15:18] For now banners know whether they can show on desktop, iphone, ipad, android or unknown [23:15:43] if the first module centralnotice has target=>'desktop' and centralnotice.mobile=>'mobile' only one of these will get loaded depending on whether you are on mobile or not [23:16:05] but only do that if you are aspiring to have centralnotice.mobile become the future version of centralnotice with targets=>desktop,mobile [23:16:32] jdlrobson: that is the way the existing code works. But the current plan is to deprecate it, yeah one module rules all sites [23:18:17] post-makeover (WIP): [23:18:17] function getDeviceCode() { [23:18:17] var ua; [23:18:17] // If we're on the desktop site, all your devices are DESKTOP [23:18:17] // TODO Fix this if ever there's a better way to seewhich site we're on [23:18:17] if ( mw.config.get('skin') !== 'minerva' ) { [23:18:19] return DEVICE_CODE.DESKTOP; [23:18:21] } [23:18:23] ua = navigator.userAgent; [23:18:26] if ( ua.match( /iphone/i ) ) { [23:18:28] return DEVICE_CODE.IPHONE; [23:18:30] } [23:18:32] do you want to hang out AndyRussG ? [23:18:32] if ( ua.match( /ipad/i ) ) { [23:18:34] return DEVICE.IPAD; [23:18:37] } [23:18:39] if ( ua.match( /android/i ) ) { [23:18:40] it might be easier to understand there what you are trying to do [23:18:41] return DEVICE.ANDROID; [23:18:43] } [23:18:45] return DEVICE.UNKNOWN; [23:18:48] } [23:18:52] * jdlrobson segmentation faults [23:19:21] jdlrobson: sure! gimme a second or two or three [23:19:29] AndyRussG: why not just make DEVICE_CODE.DESKTOP more explicit? [23:19:46] ? [23:19:48] or use window.innerWidth [23:20:00] * MaxSem slaps AndyRussG around with a pastebin [23:20:00] and assume everything over 768px should get desktop banners [23:20:36] MaxSem: due to lack of pastebin or untoward code? [23:21:00] jdlrobson: that'd be a change to current functionality that may or may not be desired, but it's not my decision... [23:21:15] thou shall not paste code in the chat [23:21:15] why not? you are the dev who has to maintain this [23:21:19] also, Java detected :P [23:21:31] you should argue that it makes more sense - i'm not sure why that would be a change in current functionality [23:21:34] lol MaxSem [23:21:52] rmoen: ahhhhh yeah I thought Monday afternons in June were an exemption [23:22:22] AndyRussG: i guess the only impact that would have is that some UNKNOWN devices will now be recognised as DESKTOP [23:22:51] (I don't know of any browsers with Android, iPad or iPhone in their user agent that are over 768px window width) [23:23:22] and i can't see how that would be catastrophic [23:23:24] MaxSem: yes hahah.... https://tools.wmflabs.org/paste/view/7a00b9bb [23:23:31] but no more mw.config.get('skin') !== 'minerva' please [23:23:39] unless you are doing skin specific banners [23:23:51] (which isn't a bad thing) but you shouldn't mix it with device [23:23:57] jdlrobson: hmmm... that's a good point [23:23:58] Desktop users can still be using Minerva [23:24:05] (i do for example) [23:24:56] jdlrobson: How is the mobile/desktop RL target distinction made on the server? It's still just a skin thing, no? [23:29:28] jdlrobson: if you want to hang out I can do so any time :) pls ping here if it looks if I'm not answering a call or something, my hangout setup is... odd... thx! [23:30:00] AndyRussG: oh we can do it now yeh [23:51:52] jdlrobson: this is the crazy task I mentioned: https://phabricator.wikimedia.org/T103695 [23:51:55] jdlrobson: thanks again!