[06:17:32] good morning! [08:05:07] morning [08:05:12] phedenskog: ^ [08:06:33] phuedx: morning! [08:07:48] so [08:07:50] many [08:07:51] emails [08:10:27] morning [08:20:19] joakino: hey! [08:20:50] hi phedenskog [08:21:28] let me know when you have time [08:22:23] i can meet now [08:22:25] + feel the joy :) [08:23:28] cool, I'll be in there in 5 min, will just get the laundry [08:23:53] joakino, phedenskog: mind if i sit in? [08:24:05] ok phuedx [08:24:18] i meant ok phedenskog [08:24:35] damn ph* [08:24:40] \o/ [08:26:12] hehe [09:35:18] phedenskog: submitted https://github.com/axemclion/browser-perf/issues/62 [09:35:23] am i missing something else? [09:35:31] (to specify there) [10:14:59] joakino: i'm having problems sshing into the labs bastion host so i can't update the staging server [10:15:17] my key might have been revoked [10:15:19] (maybe) [10:15:34] didn't it update via cron? [10:15:59] i don't have the ssh key here, only on the mb air, let me check [10:16:12] i'm guessing it is bad practice to share keys between machines? [10:19:04] i'm in, what do you need [10:19:11] joakino: the last piece of advice i read was if the machines are very closely related, then /maybe/, but if you travel with one, then no [10:19:28] makes sense [11:09:41] joakino: lovely [11:09:49] hope he can help us [13:39:36] jdlrobson: hey [13:39:43] jdlrobson: tests are failing in master too [13:40:00] jdlrobson: can someone else focus on fixing them? [13:41:04] phuedx, joakino ^ and the patch is https://gerrit.wikimedia.org/r/#/c/289553/6 [13:43:59] bmansurov, jdlrobson: i can take a look at them in a while, when i'm back from my break [13:44:13] i imagine i broke them (though i'm not sure how the breakage was integrated :/ ) [13:44:14] phuedx: coolio [13:44:37] yeah that was weird, or something is wrong with my local environment [14:14:22] bmansurov: do you know what's up with staging? there's /srv/mediawiki-vagrant running without extensions on mediawiki/extensions, and /srv/vagrant with our extensions but not running [14:14:27] i'm trying to provision [14:14:56] joakino: sorry no idea, haven't used it since quite a bit [14:15:10] this is nuts [14:15:12] phuedx: usually used to fix up my mess [14:15:19] lol [14:16:19] mediawiki-vagrant labs docs say it's /srv/mediawiki-vagrant where it should be, and that's what vagrant status says is running, but the extensions on http://reading-web-staging.wmflabs.org/wiki/Special:Version match the ones in /srv/vagrant [14:18:16] oh isn't it labs-vagrant? [14:18:21] not mediawiki-vagrant [14:20:37] wat [14:20:51] nah [14:21:53] wtf [14:21:59] bmansurov: look: [14:22:06] jhernandez@reading-web-staging-2:/srv/mediawiki-vagrant$ vagrant global-status [14:22:06] id name provider state directory [14:22:06] ---------------------------------------------------------------------- [14:22:06] 53bae82 default lxc running /srv/mediawiki-vagrant [14:22:18] now see http://reading-web-staging.wmflabs.org/wiki/Special:Version [14:22:32] docroot: /vagrant/mediawiki/ [14:22:34] dafuq [14:22:45] run labs-vagrant global-status maybe? [14:24:44] bmansurov: The labs-vagrant command is obsolete. Use cd /srv/mediawiki-vagrant; vagrant [up|provision|ssh|...] instead [14:25:00] bmansurov: i've shut down vagrant halt in /srv/mediawiki-vagrant and the site still works [14:25:17] so we have it running somehow else even if we have the labs-vagrant role provisioned and working [14:25:38] i see [14:28:58] i give up, i'll send an email [14:31:31] lol it's a manually installed environment [14:31:35] tcp6 0 0 :::80 :::* LISTEN 1591/apache2 [14:33:22] ok joakino, let's continue discussion via email [14:33:26] i'll log back in later [14:33:33] bb [14:33:35] i'm out too [14:33:41] later [14:55:00] bb [15:27:13] joakino: yt? [15:27:35] have you run the suggested source command? [15:55:23] bmansurov: you there? [15:55:30] yes [15:55:59] bmansurov: does the el send 'unknown' when localstorage is unavailable for preview count bucket? [15:56:14] bmansurov: if not, would you please add that? [15:56:38] dr0ptp4kt: not yet, we're using $.jStorage. ok I'll add that [15:56:38] bmansurov: i see it's not updated in the task description, but it is in the event logging definition (much like with the lang switcher) [15:56:54] bmansurov: thx [15:56:56] bmansurov: also [15:58:29] bmansurov: do i understand correctly that we won't persist stuff in localstorage as a consequence of the event logging? [15:58:52] dr0ptp4kt: do we keep track of the number of previews [15:58:56] we do * [15:58:59] bmansurov: right [15:59:06] bmansurov: sorry, yes, the preview count bucket's okay [15:59:21] and the settings (as before) [15:59:24] bmansurov: but is anything else persisted in localstorage as a consequence of the patch? [15:59:28] no [15:59:43] bmansurov: which settings? whether the user has it on/off? [15:59:52] yes [16:00:05] bmansurov: that's determined by the a/b bucketing and not the el bucketing, right? [16:00:40] cc phuedx ^ [16:00:47] dr0ptp4kt: i'm not sure as I haven't looked at a/b bucketing yet, but that's the functionality we have and yes it's not related to el [16:01:17] bmansurov: ok, so the only thing in your patch(es) that will get additionally persisted into localStorage is the preview count bucket, do i have that right? [16:01:26] yes [16:01:59] bmansurov: great, thanks! [16:02:01] the a/b test persists a token in localstorage, PopupsExperimentID [16:02:25] if you need be, then, while the module exists in the codebase, it can be made to delete that key if the experiment is disabled [16:06:48] bmansurov: all the popups qunit tests are passing for me locally (i also ran them without my additional configuration) [16:06:52] are you still seeing the failure? [16:07:03] phuedx: i haven't tried since [16:07:08] thanks for confirming [16:11:27] phuedx: some tests are still failing. are you testing by going to /Special:JavaScriptTest/qunit?debug=true [16:11:29] ? [16:29:44] dr0ptp4kt: should sampling happen if localStorage is not supported? [16:30:03] dr0ptp4kt: or should I ignore those browsers? [16:30:46] dr0ptp4kt: or should sampling happen on every page load for those browsers? [16:40:22] bmansurov: i see removeTooltips failing with debug=true [16:40:31] is that what you see? [16:40:41] i see thumbnail one failing [16:40:46] and somethign else [16:40:52] browser? [16:41:34] phuedx: firefox, safari [16:41:38] chrome too [16:41:47] all on a mac [16:45:30] yup, i'm only seeing removeTooltip [16:45:39] timing issues? [16:45:45] (flaky tests?) [16:49:31] probably [16:55:04] bmansurov: qunit tests should pass now [16:55:20] jdlrobson: ok, i'm addressing the issues in ps6 [16:55:27] cool [16:55:36] bmansurov: the issue was you were saving a boolean to jStorage [16:55:39] not a string [16:56:04] i see [16:56:05] lets move away from jStorage asap [16:56:16] ill rebase my patch off yours [16:56:20] should make it simpler [16:56:54] ok [16:56:55] ^^^ +1 [16:57:03] alright [16:57:06] bb [17:31:11] niedzielski: do you have that link again? I made the change I thought would work, but it didnt [17:31:59] https://www.mediawiki.org/wiki/Wikimedia_Apps/Team/Release_process#Beta_release [17:32:16] maxbinder: https://phabricator.wikimedia.org/maniphest/task/edit/form/10/?projects=wikipedia-android-app-backlog,Release,Unplanned-Sprint-Work&title=Build,%20test,%20and%20publish%202.1.XXX-2016-XX-XX%20beta&points=1&priority=50&subscribers=dbrant [17:35:32] niedzielski: yup, there are different form types, and the one that allows points is a different category. Should still work, though [17:36:01] mdholloway: i'm going to move that crash check further into the UI lifecycle. probably onresume. can you hold off a moment? [17:36:13] niedzielski: sure [17:36:38] maxbinder: i hope it works. otherwise it's pointless [17:36:44] zing [17:36:51] https://phabricator.wikimedia.org/maniphest/task/edit/form/10/?title=[BUG]&projects=wikipedia-android-app-backlog&points=1&description=%3D%3D%3D+Steps+to+reproduce%0D%0A%23+%0D%0A%23+%0D%0A%23+%0D%0A%0D%0A%3D%3D%3D+Expected+results%0D%0A%0D%0A%3D%3D%3D+Actual+results%0D%0A%0D%0A%3D%3D%3D+Environments+observed%0D%0A**App+version%3A+**+%0D%0A**Android+OS+versi [17:36:51] ons%3A**+%0D%0A**Device+model%3A**+%0D%0A**Device+language%3A**+ [17:36:55] geez [17:37:08] anyway, it works, form 10 instead of 1 [17:37:21] maxbinder: cool [17:43:01] mdholloway bearND: i say we shoot for a new beta monday. it would be nice to get the outstanding patches in but i don't think we're waiting on anything else. [17:43:16] niedzielski: sounds great [17:43:26] niedzielski: bearND: yes, let's plan on it [17:44:31] cool. sounds like a plan. we can confirm with dmitry at standup [17:54:27] niedzielski: hm, would onResume be called after a crash or would onStart be safer? [17:54:49] bearND: niedzielski mdholloway might be a few min late [17:54:59] maxbinder: ok [17:57:41] mdholloway: i'm not opposed to putting it in onStart but it's a little bit earlier in the lifecycle. there's a pretty good diagram here that i use: https://github.com/xxv/android-lifecycle. what do you think? [18:00:27] niedzielski: oh yeah, that diagram is great! but it seems to conflict with the developer docs for Activity.onResume (https://developer.android.com/reference/android/app/Activity.html#onResume()) -- and i'm not sure which i trust more :| [18:01:04] niedzielski: namely the diagram makes it seem like onResume is called in all cases, but the docs don't [18:51:48] bmansurov: sorry for delay. if sampling is based on mw.user.sessionid() you can sample on every page and get the same result [18:52:26] dr0ptp4kt: but how do I know what the previous value of mw.user.sessionid was? [18:52:49] dr0ptp4kt: ohh [18:52:53] dr0ptp4kt: i don't think it's based on that [18:52:59] bmansurov: it stays the same throughout a session as a cookie if i'm not mistaken [18:53:26] dr0ptp4kt: sampling is based on a random token [18:53:56] bmansurov: right, and if you pass mw.user.sessionid as that token, it should be determinstic, right? [18:54:18] bmansurov: got 5 mins to video? [18:54:24] sure [18:54:28] er mw.user.sessionid() a function [18:54:29] tracy? [18:54:32] bmansurov: tracy [19:20:00] hi, is someone working on https://phabricator.wikimedia.org/T135866 ? [19:21:10] jdlrobson: ^ [19:30:18] tgr: i doubt it [19:37:35] bmansurov: i'll be there in about 90 secs [19:37:44] dr0ptp4kt: sure [20:01:03] bearND: just need you and we can start :) [20:01:59] omw [20:02:09] dr0ptp4kt bmansurov: can you document the outcome (i.e. the exact sampling method involving mw.user.sessionid) on the ticket or in the schema/its talk page? [20:37:45] HaeB: ok [22:20:55] have a good weekend, all! [22:21:02] o/