[09:11:50] yo [15:23:46] jdlrobson: yt? [15:37:54] niedzielski: mdholloway|afk: bearND: any objection to pushing to production? hockeyapp looks clean enough... [15:38:14] dbrant: none, tsg report came back pretty clean [15:38:31] oops, wrong status... no objections here! [15:39:07] sounds good to me [16:02:28] i'm out! [16:04:10] mdholloway: o/ i've been enjoying your micro patch spree! one quick question: are you planning to re-add the pronunciation button in portrait mode? i just noticed it's only available in landscape [16:06:05] niedzielski: good catch. i hadn't noticed that. [16:06:36] mdholloway: oh well it seems to work great in landscape so i think portrait should be easy [16:54:25] hey joakino [16:54:28] did i miss you? [16:54:37] boo [17:40:34] https://phabricator.wikimedia.org/T143437 did we change anything here [17:59:55] phuedx, joakino, dr0ptp4kt: any new insights yet regarding the hovercards instrumentation (https://phabricator.wikimedia.org/T141922 )? let me know if i can help by exploring the data in additional directions, or if there are questions about the problem description (which has become a bit unwieldy, i guess ;) [18:02:38] HaeB: phuedx has been looking at it. we were brainstorming a bit today. one thought that occurred was: is it possible the missing lefthand side of the bell for the test group actually just got moved to the righthand side of the bell? it seems self evident that the missing lefthand side of the bell is missing. obviously that doesn't help analyzing root cause [18:03:22] but what forces would have moved it? ;) (i.e. what made people click later) [18:06:53] HaeB: it screams race condition or time-of-use/time-of-check, but i was just curious if there was a statistical way to infer shifting of the lefthand side of the curve. like is the peak too big or whatever. [18:06:56] HaeB: gotta go [18:11:24] dr0ptp4kt: we can compare size of the green area overall (or more pertinently, the average ratio of link opens per link interaction). my sense from the other results is that it did decrease (i.e. that the bell didn't just get moved) but it seems we haven't calculated that directly yet [18:11:45] i will try to do that soon [23:43:49] hey guys, question: in https://meta.wikimedia.org/wiki/Schema:MobileWebLanguageSwitcher, what does "languageButtonImpression" [23:43:50] measure? Is this for the new button? I'm looking at dates since we launched the new language button and the total is way lower than the "pageLoaded" total which doesn't make sense since the button is at the top of the page...