[00:05:19] rmoen: there? [00:07:09] rmoen: just curious how the browser environment variables get setup in your automation script [00:07:13] i guess ill just read the code :) [15:22:26] dbrant, bearND: Did you guys get your otrs spam issues fixed? [15:22:30] niedzielski, what was going wrong on gerrit? [15:23:06] Krenair: yep! DanielB was very helpful [15:23:25] Krenair: sorry, i was just whining. sometimes gerrit doesn't add everyone in the gerrit group to a patch review. the issue is very infrequent [15:24:17] niedzielski, is gerrit supposed to add everyone..? [15:24:31] I don't think I've seen it add people to review things before [15:25:13] Krenair: everyone mentioned in the relevant group https://www.mediawiki.org/wiki/Git/Reviewers [15:26:21] niedzielski, that's a standalone bot run by someone that connects to gerrit and adds people as reviewers [15:26:57] not gerrit itself [15:27:23] Krenair: ah ok [15:29:27] Krenair: since we're chatting, would you happen to know when we're switching to differential? [15:30:12] I think some teams were being allowed to beta test it? [15:30:47] Krenair: oh wow! if that's the case, i didn't realize we were so far along [15:31:03] Krenair: i haven't used it before but i'm excited to try it [15:31:22] I don't think it's entirely ready: https://phabricator.wikimedia.org/tag/gerrit-migration/ [15:48:15] hallo [15:48:55] I remember that when the Italian Wikipedia allowed mobile anonymous editing, there were statistics that showed that it didn't increase vandalism or bad edits. [15:49:04] Based on that it was enabled in other sites. [15:49:11] Are there published statistics about this? [15:49:58] I mean, after enabling it on other sites. [15:50:22] dbrant: i was curious if you feel the standard FAB button size is too large even on a tablet like device (not a loaded question, just curious) /cc mdholloway [15:50:44] niedzielski: maybe not a tablet. [15:51:20] jonkatz, dbrant ^ [15:51:43] oh sorry, dbrant is apps [15:52:16] dbrant: hm, i'm not hung up on this but if i made a follow on patch to make it larger on resource qualified devices, would you consider it? /cc mdholloway [15:52:21] niedzielski: dbrant: i felt that way at first, but quickly got over it; that said, dbrant makes a good point that we have special concerns about obscuring content. [15:52:53] i think a lot of apps probably have more leeway to kind of design around that problem [15:52:59] niedzielski: yea sure, i'd endorse that [15:53:10] niedzielski: that's cool with me! [15:53:25] * mdholloway remembers having a working tablet... [15:53:51] mdholloway dbrant: ok, i'll give that a spin when i have a little free time. let me know what you really think about it when i do [15:53:55] mdholloway: LOL [15:54:53] mdholloway: i have a 2012 nexus 7. it's fine but the USB connector is garbage so i normally have to adb over the network. i'm looking forward to getting a new one but wanted to wait for the next gen [15:55:34] yeah, i wonder how soon the next new hotness is coming out -- i was kind of thinking about that when i bought the nexus 9 [15:57:55] mdholloway: i've no inkling. i'm in a similar boat with my phone. i'm using a tracfone right now and it seems to be only be able to hold one app in memory at a time making it quite slow [16:02:09] joakino are you up for pairing/caring meet or are you eating too many mangoes and tequilas (that is, honoring your vacation)? [16:05:25] I've been duped. kristenlans, aside from me its an empty hangout. [16:07:09] rmoen: curious, i was in a hangout w jon and baha, we're going to rschedule for early next week anyway when joakino and sam can attend. ;- 0 [16:07:26] no idea what that emoticon was 0_o [16:07:53] kristenlans: oh lol, i guess i was in the wrong hangout ;/ [16:08:04] rmoen: it happens :-) [16:08:13] i think i joined standup [16:13:08] rmoen: on a different topic, heads up that I did change standup to M/W/Th, effective today. So standup today, no standup tomorrow. [16:18:30] bearND: niedzielski: mdholloway: so, a couple users have noted that the "correct" behavior of "open in new tab" should be to create the new tab *behind* the current tab, and remain on the current page. What are your thoughts on that? [16:20:16] dbrant bearND mdholloway: this is not the new tab button but when long pressing a link? [16:20:25] dbrant: niedzielski: bearND: when you mention it, that does better align with how I typically use tabs -- for something I'm going to read a bit later -- and how it works in (afaik) all browsers [16:20:41] that is, the focus remains on the original tab [16:21:05] niedzielski: yes, i think so [16:21:20] mdholloway bearND dbrant: to me, open in new tab means "read later," so yes, i want the tabs to open in the background but i need a strong visual cue for that too [16:21:28] dbrant: open link in new tab should keep the current tab in the foreground and the new tab in the background. [16:23:25] dr0ptp4kt: are you seeing the web standup invites now? [16:24:42] bearND: niedzielski: mdholloway: well then, i'll add a task for that! [16:24:52] thanks! [16:25:01] jdlrobson: are you still GSoCing? [16:25:48] dbrant: niedzielski: bearND: while we're chatting, how would you guys feel about disabling ACRA for dev builds? [16:26:36] kristenlans: thanks for the heads up. I saw the standup pop in the calendar and thought I was seeing things. [16:26:55] mdholloway: what's the motivation? [16:27:48] niedzielski: i'd rather deal with crashes/stack traces directly in AS [16:28:19] dbrant: thanks, IMO that's the one thing that would hold me off voting for promoting the tab feature to stable. Some ideas for visual clue: Snackbar message. If we had an indicator in the toolbar for the number of tabs open then that would be good to see an update there as well. But since we don't have that even a ripple on the toolbar item for the tab [16:28:19] chooser would be good. [16:29:49] mdholloway: i don't think i've had any problems with that in my workflow but if the change is just flipping a switch on / off, i wouldn't object to it. i use pidcat most of the time and just follow the stacktrace [16:30:10] bearND: well, currently when we open a new tab, we have an animation that shows the current tab stack. We can keep this animation, while putting the new tab behing the current tab. Would that be acceptable? [16:31:15] dbrant bearND: i recommend against using snackbar for indicating a new tab has been made. i think it's too heavy. i think i'd actually prefer a toast over a snackbar and an animation of some sort over both [16:31:37] *behind [16:31:58] dbrant bearND: the other thing dbrant mentioned is trying to preserve tab state before going to production. i agree with this since no one likes losing tabs [16:32:40] niedzielski: oh, yes, that one, too. [16:34:15] kristenlans: yes. yay! [16:34:35] dbrant: I guess that could work. Sounds like a simpler change. So, I'm cool with that. [16:35:37] kristenlans: yup still GSocing [16:35:44] any reason you ask? [16:37:56] jdlrobson: was looking to move pairing is caring to tuesday, but conlficts with your gsoc catchup. I can find a better time :-) [16:38:10] dbrant: bearND: niedzielski: yeah, i like the idea of modifying the current animation [16:38:13] kristenlans: yeh it's usually wednesday mornings but he could only do tuesday that week [16:38:36] rmoen: https://gerrit.wikimedia.org/r/#/c/223176/ < can you help me get that merged today. I'd really like it to be on the BC for a few days [16:38:48] and i really don't want to wait till next deployment [16:38:51] jdlrobson: yes [16:38:55] thanks dude [16:40:09] jdlrobson: btw thanks for all the feedback on the automation script. Looking into resolving that now [16:40:20] rmoen: my plan is to help you with that today if i can [16:40:40] But i could switch, and review your patch now instead [16:40:47] i'm just confused why language links are not working on http://mf-browser-tests.wmflabs.org/wiki/Selenium_language_test_page [16:40:58] rmoen: would be cool to wrap the border box stuff up [16:41:05] so we can exclusively focus on labs [16:41:10] yeah [16:41:12] context switching hurts me at the moment [16:42:10] jdlrobson: yeah, totally. Been trying to learn puppet. now back to bash [16:43:10] bgerstle_afk: i want to second what paravoid said on the mailing list, please don't use GH. [16:46:41] mhurd coreyfloyd: would you mind reviewing this guy? https://gerrit.wikimedia.org/r/#/c/222369/ [16:49:49] niedzielski: that's interesting... ACRA doesn't intercept your stack traces and keep them out of logcat? [16:50:11] (for crashes, that is?) [16:50:45] mdholloway: it did on my first few days? and i was going to make a patch for it but then it started working perfectly so i never did. i'm not sure what the original issue was [16:51:03] mdholloway: i thought it was just me [16:51:33] niedzielski: it's always been the case for me. [16:51:57] mdholloway: weird. i don't use the ide logcat so that might be one difference [16:52:13] niedzielski: and emailing stack traces to myself every time seems like a poor way of handling things. [16:52:22] niedzielski: hmm, that could be [16:52:23] mdholloway: yeah, that would not be good [16:52:54] jdlrobson, bmansurov: Just went through the border-box patch. Not seeing the issues pointed out in the screenshots. Are they resolved ? [16:53:05] rmoen: should be [16:53:46] Cool, I tested and seems good now. [16:54:01] rmoen: try resizing the browser window [16:54:28] dbrant: too many reviews! [16:55:15] bmansurov: i see. There is a point where the text is left aligned when resizing [16:55:28] matanya hey! sorry i missed you. was there anything you wanted to add to the comments or just echoing the same reasons? [16:55:43] (not trying to be obnoxious, just want to give you a chance to share your thoughts) [16:57:31] mdholloway: niedzielski: I usually click Cancel on the ACRA dialog, and the stack trace of our app appears in regular logcat inside AS. You may have to scroll up a little bit higher since I tend to get another stack trace from ACRA itself [16:58:29] niedzielski: bearND: ah, never mind, i figured it out. has to do with pids changing before the crash report is sent, looks like. just needed to tweak the controls for the AS logcat viewer. [17:00:18] mdholloway: oh, forgot to mention that when I want to see the stack traces I either turn off the logcat filters or chose one just for the Error level that doesn't have an app (PID) filter [17:01:46] mdholloway bearND: hm, pidcat seems to work fine without hitting cancel. i highly recommend it [17:03:26] niedzielski: thanks for the tip. Looks pretty cool for just browsing logcat. I do prefer the links to our code I get in AS logcat tab. [17:08:57] bearND: yeah, no links :( [17:21:49] kaldari: rmoen bmansurov those issues should be fixed with border box [17:21:55] issue was that Page was adding border box class [17:22:13] checking [17:22:19] would be great if you could re-review and help me get it merged before lunchtime to give us ample chance to test :) [17:23:19] jdlrobson: oh yeah that's much better. Fixes some other things with the content as well [17:24:20] jdlrobson: all good by me. bmansurov, want to confirm ? [17:24:29] looking [17:24:40] bgerstle: hmm, mostly echo, but adding that community volunteers don't always have GH account, and don't always want to share private stuff with a third party provider. they trust WMF, but not GH with the same level of trust [17:25:18] + one more tool to learn and remember and adopt [17:25:54] matanya you'll still be able to submit patches via gerrit [17:26:15] though we'd much prefer GH, since that's where most of our CI will be setup [17:26:35] so my last point is a bit less valid but the trust one is still very much valid [17:26:36] we'll be updating syncing gerrit master after every release [17:26:58] matanya that's ok, you don't have to touch GH if you don't want to [17:27:59] i personally have no problem with it, not to talk about the fact i don't get close to i-name-of-device-here [17:28:21] just sharing some concerns i heard over time [17:29:32] matanya: gotchya, thanks for sharing [17:29:46] matanya: what do you work on most of the time? [17:30:51] bmansurov: sure that's my patch? [17:30:52] bgerstle: eng wise or in general ? [17:31:10] let's start w/ eng wise ;-) [17:31:16] jdlrobson: i think so [17:31:50] bgerstle: https://gerrit.wikimedia.org/r/#/q/owner:matanya%2540foss.co.il+status:merged,n,z would be a good answer i guess :) [17:32:16] bmansurov: can you check without it? [17:32:21] ok [17:32:27] as i think it might be another patch i just merged [17:32:36] matanya: so ops, then? [17:32:39] yes [17:32:42] cool [17:32:47] i've done a little dabbling w/ puppet [17:33:05] mostly https://github.com/bgerstle/my-boxen [17:33:11] enjoyed ? [17:33:17] (fork of GitHub's boxen project to manage my machine configuration) [17:33:25] (really, really over-engineered dotfiles, essentially :-P) [17:33:48] jdlrobson: yes, it's happening in master too [17:33:49] but TL;DR;, puppet + puppet-librarian + lotsa batteries for OSX [17:34:00] jdlrobson: i guess i'll merge your patch [17:34:11] oh wait 1s [17:34:20] looks nice bgerstle esp. if you are a mac person [17:34:24] bmansurov: think i found an issue with the pointer overlay [17:34:43] jdlrobson: ok, i'll wait for your fix [17:35:11] matanya: that's the main reason to use it (if you're on a mac). they say it's interoperable w/ linux, but i've not had much need for that [17:35:24] anywho, got some work to do, but nice chatting [17:35:35] yes, indded. thanks for your time [17:35:40] anytime :-) [17:55:46] bmansurov: rmoen new patch up [17:56:02] k [17:58:18] dbrant niedzielski mdholloway|lunch dr0ptp4kt etonkovidova bearND see you shortly for Android THC :) [18:07:56] jdlrobson: the overlay header icon is touching the screen borders http://imgur.com/g7ClKDB [18:08:16] bmansurov: in stable? [18:09:05] jdlrobson: no, beta [18:10:17] mmm i bet it's those ghastly -1em hacks [18:14:23] bmansurov: yup that fixed it [18:14:37] not sure if the positioning in beta is now right or not but i think we can deal with that in a follow up [18:15:47] jdlrobson: also the pointer overlay is not wide enough, but that's a minor issue: http://imgur.com/NCeEElZ [18:16:21] yeh i just noticed that.. looks a bit tricky. Seems to only be in beta thought. [18:19:12] jdlrobson: i think we should keep the -1ems because according to the design the icon (without it's left padding) should vertically align with everything else on the page/overlay [18:19:25] jdlrobson: let me see if I can find the mock-up [18:19:39] bmansurov: well i'm not sure how to do this but i really don't want to block this on changes in beta that there is no guarantee we are finishing any time soon [18:20:03] let me know if you know a quicker cleaner way to fix [18:20:07] jdlrobson: ok, i'll test a little more and merge [18:20:11] cos i'm stumped with these hacky -1ems [18:20:20] i don't understand why they exist and where the -1em comes from [18:20:36] jdlrobson: 1 sec, I'll find some reference [18:20:59] if they were calculated from another value we wouldn't have this issue :/ [18:21:24] jdlrobson: they are based on the icon gutter [18:21:33] jdlrobson: i don't think they are hardcoded [18:21:41] the iconGutter is hardcoded to 1em [18:21:41] jdlrobson: https://phabricator.wikimedia.org/T96317 see the design screenshot [18:21:48] but what does this mean? Is this half the icon size? [18:21:52] (if so then the value is wrong) [18:22:15] jdlrobson: yeah i agree [18:22:32] jdlrobson: it's also weird that the icon is not implemented as a square [18:22:39] anyway if we are resorting to a negative absolute position we are probably doing something wrong somewhere [18:23:38] anyway what will it take to get this merged [18:23:54] jdlrobson: some more testing [18:23:54] given stable is fine and that should be the main goal here [18:24:23] (note those mocks don't show how the editor is supposed to look) [18:31:16] thanks bmansurov let's try and work out how the beta is meant to behave and try and make it clearer what those gutters are for [18:31:38] (in theory we shouldn't need gutters - we can just increase/decrease padding on appropriate elements) [18:32:30] jdlrobson: it would be nice to see new icon designs that conform to apple human interface guidelines (cc kaity|away) [18:34:42] bmansurov: they /should/ do [18:34:57] violetto_ spent a long time with the ux standardisation team doing this sort of thing [18:35:06] if we have non-standard icon they are probably from the old days [18:35:15] or have been built incorrectly and need updating [18:35:18] jdlrobson: all icons have left and right gutters but not top and bottom [18:35:25] if they are pngs the have definitely wrong [18:35:46] i think icons at least should be 44x44px [18:36:01] bmansurov: there's a bug around this i think nirzar was trying to change the size [18:36:14] dip into the user standardisation project - you'll find all sorts of disagreements there :) [18:36:33] oh no, not another disagreement [18:36:45] bmansurov: the icons are made on a 14px x 14px canvas as requested by trevor. But they're svgs, so you can use them at 44px if you prefer [18:37:24] we've redone the icons 3-4 times now at different sizes, but trevor wanted it at a specific size at 14px [18:37:24] violetto_: ok, thanks we should change the implementation then, jdlrobson [18:38:00] bmansurov: to us it doesn't matter since it's svg [18:38:10] makes sense [18:38:19] bmansurov: who's /we/ ? We use a number given by core [18:38:40] i really don't want to go through icon crap again.. :) Luckily the ui standardisation group is apparently rebooting soon [18:38:57] so we should filter any problems through there. I'd rather we worked to all use the same thing. [18:39:27] all the updated icons are already in core [18:39:32] under oojs [18:40:52] Anyway raise a task being careful to explain motivations and problems with current approach and we can prioritise it accordingly. Right now I'm really concerned about sucking more of the teams energy on anything around this without good reason [18:45:47] bmansurov: you were just saying the touch points should be larger, yea? [18:45:57] not the icon file specifically [18:47:51] violetto_: yes [18:48:35] ah that's what i thought, cus it's resizable unless you're using png bmansurov [18:49:02] ;) [18:54:04] mbinder: Monte is out sick today… also Kaity and VIbha are gone as well. You want to reschedule again? [19:18:27] coreyfloyd: seems apt. I'll look at the sched, if it's too tough, we may be better off waiting until the next one. Kind of like dosing a prescription :) [19:22:39] lol [19:22:48] mbinder: ok sounds good [19:23:09] jdlrobson: so i just hosed my instance while trying to test. Going to re-spawn and continue after lunch [19:23:11] mbinder: also - Kaity and Vibha are out tomorrow as well - FYI [19:24:17] coreyfloyd: yep thx [20:04:01] bmansurov: rmoen|lunch: yo i'm back btw. sorry my message saying i was heading to lunch didnt show [20:04:36] codezee: there? [20:04:41] there is a mail [Wikitech-l] Changes require Verified+2 to be merged by CI [20:04:49] apparently Jenkins will be back to complete normal Friday AM european time [20:04:58] jdlrobson: yes [20:05:03] codezee: did you see there is a problem with the latest wikidata ? [20:05:09] it's causing me problems testing your patches :( [20:05:36] k [20:05:40] jdlrobson: yes, I'm looking into it, but I'm unable to figure out, meanwhile I'm using local files or commons for testing [20:07:26] jdlrobson: these two https://gerrit.wikimedia.org/r/#/c/226341/ and https://gerrit.wikimedia.org/r/#/c/226303/ have +2 but not V+2, should verified be given? [20:08:17] dr0ptp4kt: I'm getting blocked while sshing into labs to verify analytics. May be because of my public wifi situation. Anyway way you could log in and check for some events for me? [21:20:36] jdlrobson: 1:1 in 10 minutes instead of at 1535 sound good? 1535's also ok [21:21:51] 1535 would be better if that's still ok. I'm just wrapping some stuff up right now.. [21:22:28] jdlrobson: cool [21:22:59] niedzielski bearND one of you guys have time to help me verify up some logs in beta event logging? My ssh is broken, maybe because of insecure wifi… [21:28:23] coreyfloyd: yeah i can help out [21:28:34] coreyfloyd: sorry, i didn't see your message until now [21:28:44] niedzielski: np [21:28:48] sweet! [21:29:14] coreyfloyd: is there a schema you're looking for or do you know your device id? [21:29:27] MobileWikiAppNavMenu [21:29:31] shod be the only ones… [21:30:13] niedzielski: ^ [21:30:28] niedzielski: send about 4:06PM EDT [21:30:31] sent [21:31:01] niedzielski: 16 events [21:31:41] coreyfloyd: 15 events [21:31:51] hmmm… let me recount [21:31:57] coreyfloyd: ok to post the last one? [21:32:45] niedzielski: ? [21:33:02] coreyfloyd: er, i was going to copy and paste one here for your viewing pleasure [21:33:06] k [21:33:22] coreyfloyd: {"clientIp": "940214c0b215a53dfaec2ca5ff052f6a1a7167d6", "event": {"action": "select", "appInstallID": "656965BC-6554-4B69-B9FB-BA28BC93CA6F", "menuItem": "Random", "sessionToken": "7803A776-3BC0-49B1-A6E7-EC42CD683972"}, "recvFrom": "deployment-cache-text02.deployment-prep.eqiad.wmflabs", "revision": 12732211, "schema": "MobileWikiAppNavMenu", "seqId": 221352, "timestamp": 1437682053, "userAgent": "Wikip [21:33:22] ediaApp/4.1.6.150 (iPhone OS 8.4; Phone)\n", "uuid": "9289a28377805923b48bcb0604e9cefd", "wiki": "enwiki"} [21:34:07] niedzielski: yeah thats the last one [21:34:22] niedzielski: is there w way to find the one that is "lost" [21:34:44] i forget - isn't there a table they are sitting in? [21:35:31] coreyfloyd: hm, here's what i executed: ssh eventlogging grep -c MobileWikiAppNavMenu /var/log/eventlogging/all-events.log . there is a log for messages that didn't make it. let me see if i can find it [21:36:00] niedzielski: thanks for the help [21:36:51] np, i need to know this stuff. i think i have it, i just need to find my note. just a sec [21:38:33] niedzielski: yeah its a bit arcane… i have notes for it too… got it working a while back… its broken now, and I feel like it might be my local network… i just don't want to bother troubleshooting while I am on a rogue network. [21:38:46] coreyfloyd: hm, i thought the file was /var/log/upstart/eventlogging_processor-client-side-events.log but i'm not seeing it. [21:38:58] coreyfloyd: can you run your test again? [21:40:13] sure [21:40:16] one sec [21:42:11] niedzielski: ok - done [21:42:39] coreyfloyd: now there are 23 events total [21:43:17] niedzielski: ok math time… [21:43:42] niedzielski: can you paste me all the logs? [21:43:51] i think i am missing 1 again [21:44:58] coreyfloyd: i sent you an email [21:45:46] coreyfloyd: now i am showing 26 entries [21:46:01] oh - that may be right… [21:46:04] maybe a delay there [21:46:19] coreyfloyd: the event logging doesn't seem to propagate until a subsequent event "pushes" it through in my (limited) experience [21:46:38] coreyfloyd: and there's a small delay but usually not much [21:46:44] niedzielski: thanks - i should be able to figure it out then with the logs… [21:46:53] niedzielski: i think i just didn't send the last log [21:46:57] from the first batch [21:47:08] let me resend that mail to make sure you got the full payload [21:47:31] niedzielski: cool [22:10:24] * bearND niedzielski: any concerns merging https://gerrit.wikimedia.org/r/#/c/226130/? [22:11:05] bearND: no, sorry i was going to +2 that when i got sidetracked on some jenkins stuff [22:11:42] dbrant: ^^ meant to add that i appreciate your responsive to my feedback and hope it came off as constructive rather than forcing my beliefs on the method