[07:53:11] morning! [09:42:29] hey joakino [09:42:40] hi phuedx [09:46:30] joakino: how would you feel about me resolving some pure tech issues that are in sign off? [09:47:26] or are you precious about the column? [09:47:59] phuedx: that would be awesome [09:48:04] thanks! [15:14:39] ugh [15:14:40] builders [15:25:33] preparing dinner/playing with the kids [15:48:51] dbrant bearND: hey! did we want to send that second freeze email today or wait a bit? [15:51:23] niedzielski: good call; let's send it later today, while I add some of this info to the app FAQ page. [15:51:54] dbrant sounds good [16:25:15] hey FlorianSW ! :) hey phuedx|afk etc [16:25:35] hi jdlrobson :) Btw.: I'm trying to fix https://phabricator.wikimedia.org/T110436 ;) [16:25:55] FlorianSW: what did you think of my suggestion about making it configurable? [16:26:01] it seems like the safest thing to do right now [16:26:10] really we should re-evaluate our section collapsing code [16:26:35] jdlrobson: actually I try to make the selection of possible h tags more restrictive, instead of simply look for " i think that would definitely be better on the long term - i think most section headings are collapsed [16:27:03] i suspect this will be 2 patches [16:27:19] 1 to workaround the problem and then 1 which will need some thorough testing and carefulness [16:28:22] sounds resonable. Let me upload a change to make the list of possible h-tags configurable, that shouldn't take too much effort :) [16:33:46] if anyone is trying to get on the the reading strategy meeting right now, lemme know and I can drop (we're at the cap) [16:53:20] yo [16:54:28] gonna get lunch and catch up with the google doc, if anyone wants in on that strategy meeting [17:03:36] dr0ptp4kt, jdlrobson: is the hangout full? [17:03:42] i can join for the last 20 minutes [17:04:02] phuedx: join now [17:04:10] i just dropped out so you can have my spot [17:05:04] jdlrobson: ta <3 [17:06:49] FlorianSW: tests are unhappy :/ [17:08:12] looking... [17:22:55] jdlrobson: do you have any idea why this failed https://integration.wikimedia.org/ci/job/mediawiki-extensions-hhvm/28837/consoleFull? [17:23:12] all resource tests pass for me locally [17:23:23] The module 'ext.visualEditor.mobileViewTarget' required by 'mobile.editor.ve' must exist 13:25:18 Failed asserting that an array contains 'ext.visualEditor.mobileViewTarget'. [17:23:32] ... [17:23:35] thanks [17:23:41] i couldn't read the console log ;) [17:24:00] what was the patch? [17:24:18] https://gerrit.wikimedia.org/r/#/c/234267/ [17:24:37] btw [17:24:46] the browser tests caught a genuine error on my part \o/ [17:25:47] the extension test suite also passes for me locally [17:27:44] \o/ [17:28:39] phuedx: there's a case to be made that wikipedia is just a platform and we let other people suck our content [17:28:52] (i'm not going to get a chance to mention this in the meeting :)) [17:28:57] i was just about to say that [17:29:01] say it! [17:30:37] phuedx: that's for the next meeting [17:30:41] jdlrobson: allowing people to consume our content and derive value from it is inline with our mission [17:31:13] jdlrobson: they were wrapping up [17:31:44] out to supermarket real quick [17:32:56] * even if they are an all seeing, all knowing gatekeeper to the internet ;) [17:33:10] jdlrobson: really can't get the -hhvm and -qunit tests to fail locally :( [17:37:22] phuedx: sorry, just saw your message. saw you managed to get in for part [17:37:57] getting the kids to bed [17:53:26] joakino, bmansurov, rmoen: test failures are nothing to do with my patch: https://gerrit.wikimedia.org/r/#/c/234322/ :/ [17:53:41] new builds on quicksurveys might be failing all the time? [17:53:45] gtg [17:53:45] boo [17:53:51] george is ready for bed and shouting [17:54:00] (daddy! story time! etc etc) [17:57:57] bmansurov: I've responded to your feedback on the QuickSurveys bucketing. I'll resubmit in a few minutes. [18:10:45] rmoen, cool, i'll take a look in about 6 hours when I'm online again [18:10:56] hopefully it will be merged by then [18:23:42] phuedx|afk: when you get back would be great to know what's going on with https://phabricator.wikimedia.org/T107603 as i'm keen to wrap that up [18:26:31] FlorianSW: to wrap up https://gerrit.wikimedia.org/r/#/c/230295/ - any idea why this doesn't work [18:26:40] is this issue documented on the web somewhere? [18:28:32] vibha: hey you have a few minutes to wrap some more styling tweaks? [18:31:08] jdlrobson: I totally doesn't have any idea, sorry. I trusted Esanders (remember, that I haven't any device that could emulate iOS, nor does I have any iOS device, so I can't test this) :( [18:36:04] edsanders: any idea what's going on with that ios toolbar? Happy to merge i'd just rather we have confidence about whether it's an upstream bug or a problem with our code [18:36:29] also if it's super urgent for you guys and you don't care feel free to +2 when i'm not looking ;-) [18:37:36] you have some grunt issues here btw FlorianSW > https://gerrit.wikimedia.org/r/#/c/233481/ [18:42:58] jdlrobson: yeah, that's because Gather would need a change too :) [18:43:02] FlorianSW: sent you my early thoughts on the heading formatting tuff. Definitely something we should think more about [18:43:17] FlorianSW: ohhh. we should do that first then i guess? [18:43:21] did i miss a patch [18:43:56] jdlrobson: there is no change now. But: It would fail, too :) that's one of these changes where you need to merge one of both changes without jenkins (unfortunately) [18:44:28] there's always a way.. just sometimes seems an unnecessary pain but master should always be deployable! [18:46:13] jdlrobson: if we would use 3 changes, instead of 2. But I'm not sure, if this is really needed... [18:46:37] it's wise. Remember we sometimes cherry pick - so we can never guarantee the same code will roll out together [18:47:01] it seems here that you could introduce the mobile.search.api, keep it in mobile.search [18:47:06] do the gather fix which will be needed anyway [18:47:10] and then remove it from mobile.search [18:47:50] jdlrobson: that was the idea :) [19:28:20] jdlrobson, I'm not sure why it get's stuck, but basically viewport size reporting goes crazy once you open the keyboard [19:30:21] either way it will always be hidden when the keyboard is open because there's no way to measure the remaining viewport. the techniques we use in VE to have all toolbars at the top and switch them based on context (document focus) are much more reliable and a better use of screen estate [19:30:49] edsanders: so would it be better to simply hide in on 'input' event? [19:31:18] we toggle the editing bar on focus/blur [19:32:21] if we want to keep it as is, FlorianSW should probably update the inline comment to mention it's only an issue with keyboard open [19:32:27] i just wanted to fully understand the problem [19:33:19] we can reference the phab task I guess [19:42:30] niedzielski: go ahead with the EOL email when you're ready [19:44:18] dbrant: will do momentarily. by the byte, i think that cahce issue we see is just on newer devices and has to do with Application.getFilesDir() returning a path like /data/user/0/org.wikipedia.dev/files/savedpages/... instead of /data/data/org.wikipedia.dev/files/savedpages/... [19:45:06] niedzielski: aha, curious... [19:46:30] dbrant: for the public email, should this just go to mobile-l or something larger? [19:47:20] niedzielski: i think mobile-l is good [19:47:31] dbrant: ok here goes [20:56:38] niedzielski: srcsets! of course! [20:57:00] dbrant: yeah, i was barking up the wrong tree with that url [21:01:36] dbrant: sounds like you're onto something [23:13:31] dbrant|bbl: do we have event logging around wiki zero? i'm not seeing any funnels in the code or tables in the database with "zero" in the name. i'm trying to field this question about old api usage for zero users but i don't think we have that data [23:28:24] niedzielski: dr0ptp4kt may know, since he was the maintainer of that code for the longest time. [23:29:03] Deskana: thanks :) [23:52:24] niedzielski: i'm on it. header enrichment in varnish tags zero-rated traffic, and that's reflected in hadoop. [23:57:09] jdlrobson, rmoen o/ [23:57:28] hey bmansurov just finishing up for the day [23:57:29] hello bmansurov [23:57:31] dr0ptp4kt: thanks! just to recap, if they're android 4.0.2 (API 14) or lower, they'd be considered legacy. i think the question is specifically about legacy vs modern in the pool of wiki zero users, but for sake of comparing apples to apples, it might be helpful to know the total number of legacy users and modern users as recorded in hadoop (if this isn't too much trouble) [23:57:42] but lots of movement. just waiting for new patches from you and rmoen to wrap up the quck survey work [23:57:54] cool, anything important I should take care of? [23:58:07] jdlrobson: https://gerrit.wikimedia.org/r/#/c/233763/ [23:58:29] rmoen: move the card otherwise i won't get the notification :) [23:58:36] will look at it now [23:58:37] Finally, tests are passing. I still cannot get them running locally. Though i have tried [23:58:39] i will [23:59:20] jdlrobson: sorry forgot to move the card. Moved it a bunch of times today ;/