[07:59:36] morning@ [08:08:40] joakino: phuedx_ good morning! [08:09:08] we had to revert the MobileFrontend change that introduced support for MediaWikiServices ( was https://gerrit.wikimedia.org/r/#/c/305211/ ) [08:09:29] because on beta cluster update.php broke with some fatal error https://phabricator.wikimedia.org/T143974 [08:09:40] morning hashar [08:09:45] oh let me have a look [08:10:25] the task has some details [08:10:32] there is a fix proposal for mediawiki/core [08:10:42] and apparently archCom is going to talk about it [08:10:58] I will be the one cutting the wmf branch tomorrow around noon CEST [08:11:10] so the MediaWikiServices thing might not land in the next wmf version [08:11:18] can be deployed later on as a SWAT thing though [08:11:26] poke me as needed :] [08:11:31] hashar: there's no rush I believe [08:11:39] it's work phuedx_ and I have been doing to improve code quality [08:11:44] but it is not time sensitive [08:11:55] that is what I thought :] [08:12:13] you might want to loop back with whoever deal with MediaWikiServices system (I have no clue what it is or how it works) [08:12:39] then easier it will land in master and get added in wmf branch next week [08:12:49] or you can get it pushed over a swat deploy slot [11:59:09] joakino, hashar: i'll try to weigh in on FlorianSW's patch [11:59:16] hashar: thanks for the ping [12:00:49] phuedx: if you want that patch to land in the next wmf branch to be deployed this week, I will cut the branch tomorrow around noon CEST [12:01:01] then that can always be cherry picked later or wait for next week :] [12:12:10] there's really no rush [12:12:22] thanks hashar :) [12:32:41] joakino: did you see HaeB's analysis? [12:36:34] my kids are both in bed, so i'm reading email [15:41:29] dbrant: coreyfloyd can you guys make the tech leads meeting? [15:41:46] dr0ptp4kt: 1min [16:03:58] phuedx: kickoff [17:01:52] phuedx: standup [17:01:55] phuedx: sterndup [18:02:01] phuedx: can you send a standup email when you get a chance? [18:29:57] maxbinder: sam is off today if i recall correctly [18:30:07] jdlrobson: ah ok, you're right [18:30:17] phuedx: disregard [19:42:27] Gene Wilder :( [19:46:06] geez [19:46:24] i just got a copy of young frankenstein yesterday! [21:24:29] jdlrobson: can't remember if we talked about it, but anyway the tablesample operator can be handy when querying hive. there's an example at https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Hive/QueryUsingUDF . i haven't gotten to running queries yet, but plan to use this when i do some hive stuff. i usually first check on an hour of data to make sure it [21:24:29] works, then run the bigger query across the appropriate date range. [21:24:50] Hey dr0ptp4kt [21:25:05] the fine people on #wikimedia-analytics are generally happy to give guidance on queries if you have questions, too [21:25:06] i was gonna go through your research this afternoon and see if I could get a dump of the data [21:25:38] jdlrobson: you need anything from me? [21:25:56] dr0ptp4kt: tomorrow probably but not right now. Will ping you if I have any questions. [21:26:09] jdlrobson: cool. catch you later. [22:34:17] jhobs bmansurov: any chance you can save me a swat deploy by reviewing the lazy images math patches? [22:36:05] jdlrobson: i'm already away from home, sorry :( (I still don't have +2 anyways if someone wants to send that email) [22:37:12] jhobs: did you post a patch for the related articles viewport code btw? [22:37:51] jdlrobson: I didn't today; I'll have one up tomorrow though, hopefully before standup [22:38:06] a WIP patch would be much appreciated [22:38:10] if you can [22:38:56] jdlrobson: like I said I'm already away (responding on mobile) :/ I'll try to throw something up this evening maybe [22:41:09] as a best practice please post up anything you have before you leave and comment on the ticket at the end of your day - it's helpful to the team if you are ever sick as someone can just pick up where you left off. [22:43:11] dr0ptp4kt: did you use Open Office to analyse btw? My tsv file is too big for google docs and Open office keeps opening it as a word doc ahahrhg [22:43:12] :) [22:45:33] apparently it needs to be a csv file sigh [22:45:36] jdlrobson: i did use open office, yes. that was indeed on a probably rather smaller file given it was confined to not all the wikis! change the extension to csv [22:46:08] dr0ptp4kt: so i thik the hive query is running - 2016-08-29 22:45:12,302 Stage-1 map = 11%, reduce = 0%, Cumulative CPU 244400.64 sec [22:46:35] jdlrobson: cool [22:46:40] if all goes to plan it will be in a jonquery.tsv file soon [22:46:57] ive got navigation timing data. Just a shame i cant put it in google docs [22:58:27] jdlrobson: looking [22:59:32] dr0ptp4kt: too many rows. doh. [23:07:34] jdlrobson: anything else besides https://gerrit.wikimedia.org/r/#/c/306953/ ? [23:08:02] bmansurov: there are 2 patches. That has a parent - https://gerrit.wikimedia.org/r/#/c/306479/4 [23:20:35] tthanks bmansurov as always