[01:40:17] brion, i want to show a video on wmflabs maps demo, like in this example - https://www.mapbox.com/mapbox.js/example/v1.0.0/video/ [01:40:32] what do i need to do to include a video from commons? [09:55:49] o/ [12:26:39] jenkins runs 'mediawiki-extensions-qunit' and 'mwext-MobileFrontend-qunit' on a patch submit to MobileFrontend, can anyone tell what is the former set of tests? [12:30:38] codezee: link? [12:31:09] bmansurov: I've submitted a patch, locally mobilefrontend qunit passes, but here one of them fails - https://gerrit.wikimedia.org/r/#/c/258969/ [12:34:46] looking [12:40:44] codezee: "Error: Module not found: mobile.settings/settings" [12:42:00] bmansurov: I saw that but I don't think I've fiddled anywhere with mobile.settings, but nevertheless I'll look again, I was wondering about the two set of qunit tests [12:42:26] thanks :) [12:42:41] i think the first one just checks general JS stuff, and the second one only runs qunit tests in MF [12:43:48] usually, comming 'recheck' will solve the issue as it may depend on the running environment [12:43:56] commenting ;) [12:44:17] sure [12:46:19] I suppose I'll look into it again, and submit a new one :P [12:49:40] spotted, I removed mobile.drawers dependency, which caused it to fail [12:50:08] good catch [17:00:12] niedzielski: i'm about ready to merge the native toolbar! any objections? [17:00:34] dbrant: nope! [17:05:07] hi there [17:16:53] dbrant: why does the android phone gap app not point to the new shiny one nor say it's not being actively developed? (spurred by https://twitter.com/berkes/status/676313070239686656) [17:18:23] jdlrobson: hmm, good call. We'll need to update that, for sure. [17:18:39] also see there are some open pull requests.. [17:32:08] dr0ptp4kt_: screen is really blurry on youtube [17:32:51] dr0ptp4kt_: i'm watching on YouTube [17:34:09] ...and crystal clear on hangout [17:57:35] bgerstle: exactly, the problem was that there was no a/b test (the site became faster for everyone at the same time) [17:58:03] HaeB: right [17:58:46] jdlrobson: yeah, i think that repo is uber deprecated [17:58:56] jdlrobson: we should probably take it down or mark it as such in the README [17:59:02] at least remove the iOS folder [17:59:25] ...ie. it's an https://en.wikipedia.org/wiki/Observational_study instead of a controlled experiment [17:59:58] sure [18:01:14] HaeB: as far as "lack of evidence," do we at least have qualitative feedback RE: the speed difference? [18:01:48] i know that's not usually something you gather for websites (unless you have one of those dreaded "We'd love your feedback!" full-page interstitials) [18:04:49] bgerstle: well, we got slapped on the back a lot https://twitter.com/mediawiki/status/630865572654264320 https://phabricator.wikimedia.org/T107399#1869984 [18:07:28] HaeB: yeah, curious how much of that is kudos from engineers vs. "omg i love the site now that it's faster" from users [18:07:50] HaeB: e.g. on iOS we can at least see if we have app store reviews that mention targeted perf improvements [18:07:57] HaeB: is there a "website" store? ;-P [18:08:37] HaeB: back to the point about data usage, would be interesting to see page views per byte served [18:08:54] e.g. if we're serving the same number of views w/ an order of magnitude decrease in bytes served [18:09:05] could be a notable accomplishment in terms of overall efficiency [18:16:07] bgerstle: yes i've been thinking about the speed data from user perspective [18:16:33] bgerstle: I dont think most people would notice any difference [18:17:52] bgerstle: I'm wondering if the slower the typical connection speed, the less users have knowledge of what Wikipedia is [18:18:42] good night! [18:20:11] night joakino [18:22:40] jdlrobson: +1'd [18:22:55] 'night king [18:28:02] jhobs: would be great to get a fix for https://phabricator.wikimedia.org/T121152 out (and new release) what do you think? I worry we will lose steam on related articles after the cut [18:29:42] jdlrobson: fine by me. I'm about to head to lunch if you wanted to do it while i'm out, otherwise I can do it when I get back. Either way we should just include it in the 2.1.0 release [18:30:00] jhobs: already merged that but we can do a 2.2 ;-) [18:30:04] yeh i need to head to office for the meeting [18:30:05] oh lol [18:30:37] jdlrobson: yeah new release is fine [18:31:21] worst case if it didn't get in today for some reason we could always leave it merged waiting for january (i don't see why it wouldn't get in today though) [19:16:52] dr0ptp4kt_: the folks on 5 are looking for you [19:35:31] jdlrobson want to demo related articles? [19:36:44] coreyfloyd bgerstle mhurd is now a good time to demo some 5.0 stuff? or are you waiting for beta? [19:38:12] JonKatz: we haven't really talked about it. we've already demo'd the feed [19:38:41] we've mostly been bug-fixing and doing other less-user-facing things (e.g. error reporting) [19:39:01] bgerstle k cool, i thought the search/switch language feature was new and exciting (but i might just not know the product ;)) [19:39:17] JonKatz: it is for iOS, but IIRC android's been doing it for quite some time [19:39:51] bgerstle k, might as well wait and wow everyrone with the overall majesty [19:40:04] yeah, i'd rather wait until we have something we really want to demo :-) [19:40:26] ^ totally reasonable [19:40:58] unless coreyfloyd & mhurd have something they think we should be demo'ing ^^, i'm fine sitting this one out [19:57:59] JonKatz: i was hoping someone else in the team would since i've had little to do with its dev. I really don't want to be the demo guy purely because of the time we have this meeting ... [19:58:31] jdlrobson okay, who would you nominate [19:58:53] Well Baha would be the ideal person but i'm guessing he's not around? [19:59:41] jdlrobson okay we can punt till next one [20:00:00] is the opensearch api the same thing that mobile and apps use for as-you-type search suggestions? [20:00:20] ragesoss: yup [20:00:22] should be [20:00:36] jdlrobson: thanks [20:14:20] mdholloway: looks great [20:23:28] dbrant: thanks! that was actually the version on gerrit from last friday, which i had to fall back to because of various technical difficulties that made it impossible to showcase the latest. (have you had the issue with the genymotion 6.0 preview emulators not injecting our custom CAB menu?) [20:24:02] mdholloway: i haven't been using the 6.0 emulator much... [20:24:18] dbrant: me neither, which i guess is why i never noticed before now. [20:57:49] there are two mobile related (possible) problem reports on WP:VP/T [21:11:24] bearND: is storeOldContentThenEdit() in your patch just to update the cache? [21:12:20] niedzielski: no, it's later used to makes sure the page content has changed [21:13:06] bearND: what's the new FetchSectionWikitextTask needed for i mean? [21:13:45] niedzielski: so I can check what the old value is before editing the page. After the edit is done I make sure we don't have the old content anymore [21:14:11] bearND: ah, i see. thanks! [21:19:07] Would be great if I could get quick reviews on https://gerrit.wikimedia.org/r/257438 (rebasing doesn't make much fun :() :) [21:21:15] niedzielski: which script(s) do you use to check localizations? [21:26:03] bgerstle: we have an android instrumentation test https://phabricator.wikimedia.org/diffusion/APAW/browse/master/app/src/androidTest/java/org/wikipedia/test/TranslationTests.java [21:26:33] hm [21:27:05] but no shell script that "sanitizes" them [21:27:07] ok [21:29:54] bgerstle: no, i was just running sed on the files but that doesn't seem to work for RTL as mhurd found. i was looking into a quick fix [21:30:21] niedzielski: yeah, if there isn't one we can figure it out [21:30:37] we might need to update the guidelines as well [21:48:46] dbrant|brb: mdholloway niedzielski: the 55% RB roll-out for the beta app is operational (https://meta.wikimedia.org/static/current/extensions/MobileApp/config/android.json) [21:48:59] \o/ [21:49:04] excellent [21:52:16] bearND: how did you say you were using inkscape to resolve icon creation issues (separately from niedzielski's patch)? i played around a bit, but no luck. [21:56:16] mdholloway: I launched the Inkscape GUI, load the svg, then rotate 90 degree one way, then 90 degree back, save, got a shiny new svg. The use that svg to run through the convert-icons script [22:05:30] niedzielski: get your change-id's outta my pull request :-P [22:05:41] (by mine, i mean your pull request :-P) [22:06:06] bgerstle: whoopsies. btw, should i be squashing? [22:06:15] niedzielski: no [22:06:17] we love commit history [22:06:30] also, rebasing nukes comment history [22:07:07] bgerstle: if i amend these commits, won't that blow it away? [22:07:22] niedzielski: that's what i'm saying: don't squash [22:07:41] bgerstle: doesn't editing the commit message require amending? [22:07:55] oh niedzielski [22:08:00] you dont' need to edit commit msg [22:08:18] your commits are separate from PR description [22:08:35] bgerstle: oh, so you actually use GitHub.com to browse commits? [22:08:44] oh yeah [22:08:47] and for code review [22:08:50] [22:08:54] niedzielski: https://github.com/wikimedia/wikipedia-ios/pull/323#issue-122125885 [22:09:00] niedzielski: it's awesome, for the most part [22:09:12] we try to keep diffs manageable and commit as atomically as possible [22:09:20] so you can view PR's as a succession of steps (commits) towards the end result [22:09:22] bearND: that's what i thought. just tried again and that still doesn't work for me for some reason. [22:09:29] helps a lot when there's some yak-shaving involved [22:09:56] mdholloway: ok, would you send me the svg? [22:09:57] PRs don't even really have "descriptions" anymore, just a leading comment [22:10:21] bearND: sure, kaity's original or the modified one? [22:10:33] mdholloway: the original [22:10:51] bgerstle: that's great as long as github always exists :) [22:10:54] bearND: you can download here https://phabricator.wikimedia.org/F3069832 ("download file" link is on the right) [22:11:09] bgerstle: alright, i've blasted the description you pointed at [22:11:16] niedzielski: 👍 [22:11:19] mdholloway: ok [22:11:24] and we don't use PR's as documentation, so nbd if it goes away [22:11:33] just to aid w/ code review [22:11:57] niedzielski: thanks for all your help today! next one's on me (i hope!) [22:12:00] mdholloway: which size bucket does this go to? [22:12:06] and i owe you some 🍻 when we're in SF [22:12:11] bgerstle: np, i think next time should be less bumpy [22:12:22] niedzielski: last favor: can you pastebin me the regex you used for the dollar sign stuff? [22:12:25] bearND: 24, i think (it's 24 x 24) [22:12:55] bgerstle: i found this open source pastebin: http://hastebin.com/iyacovubew.hs [22:13:16] purty [22:13:31] oh wow, that's the whole workflow huh? [22:13:35] niedzielski: i can update the wiki w/ this [22:13:56] lookit you using xargs n stuff [22:14:02] we gotta bonafide shell wizard over here [22:14:09] bgerstle: uh, so i don't have a solution in my pocket for that arabic problem... i would like to solve it one day but that day is not today. i hope that it would just work with a real tool like Python regex [22:14:32] i mean, all we need to do is flip the dollar signs around, right? [22:14:39] is there anything more to it than that? [22:14:44] bgerstle: nothing [22:14:56] so, y u need python? [22:15:20] bgerstle: i don't think sed mates with rtl well [22:15:39] bgerstle: feel free to try it. just didn't seem to work for me [22:16:00] bgerstle: i've had the problem previously but that was a year plus ago. maybe it's fixed now [22:16:04] hm, but that sed statement you pasted works? [22:16:20] mdholloway: hmm, the svg seems to work, but I'm not 100% sure it's correct. I see an upper-case A and three horizontal bars [22:16:24] not sure what you mean by RTL [22:16:24] bgerstle: i don't think so. it should work fine for LTR not RTL [22:16:31] bgerstle: oh, right to left language [22:16:35] mdholloway: is there a screenshot of how it's supposed to look? [22:16:43] i know what RTL means, but why does that matter for parsing a strings file? [22:17:17] bearND: nope. i'll take a screenshot of what i get, though. [22:17:34] bgerstle: i think so. feel free to try it. the test would be easy [22:18:01] bgerstle: just make sure your editor properly reads it. everything but atom (including Sublime) failed me [22:18:27] niedzielski: hm, ok [22:19:09] bgerstle: btw, i recommend this recent rtl talk if you haven't seen it already: https://www.youtube.com/watch?v=qmLdHuFRGgM [22:19:27] bearND: https://phabricator.wikimedia.org/T114949#1879304 [22:20:24] mdholloway: wow, that tiny, probably wrong size bucket. how does the png look by itself? [22:20:29] niedzielski: flagged for later, thanks [22:20:29] mdholloway: want me to save out another svg? [22:21:23] bearND: that was my first thought too but i tried various size buckets and they all ended up that way [22:21:26] bearND: slightly bigger or smaller, but always tiny [22:21:48] kaity_: the error seems to be on my end, what bearND is describing sounds like how it should be. [22:22:03] mdholloway: I added it to phab again but forgot to ping you [22:22:14] mdholloway: shoot ok, thought i'd try! [22:22:31] dbrant|brb: thanks for all the nice polish to the menu bar [22:22:33] mdholloway: hmm, ok, the trick I used is for fixing some strange artifacts appearing in the resulting png, but not about resizing [22:22:40] kaity_: can't hurt to try with a new one :) [22:23:24] mdholloway: restart everything? throw computer out the window and get a new one? ;) [22:24:24] kaity_: exactly! [22:25:09] kaity_: just make sure nobody is walking on the street nearby when you throw it out ;) [22:29:36] kaity_: mdholloway : I've posted how the png looks like when I use the svg on https://phabricator.wikimedia.org/T114949 [22:30:12] bearND: that's from the 24 bucket? [22:30:16] bearND: looks correct there! [22:31:13] jhobs: how are you getting on with the scrollign issue? [22:31:36] mdholloway: yes, If you want I can add a standalone patch for this icon or somehow amed it to your patch [22:31:50] jdlrobson: not making much headway :/ I think it's time I throw it up for grabs [22:32:21] i've had some things work, but never consistently enough (always dependent on loading) [22:32:23] jhobs: i could possibly take it over. Wanna grab https://phabricator.wikimedia.org/T121152 to get an immediate sense of accomplishment? [22:32:35] sent what you've done so far and i can take a look [22:33:03] i've got meetings for next hour but should have bandwidth at end of day [22:33:29] jdlrobson: yeah i'll pick that up. And you're better off cloning fresh. I've be wiping my local work everytime something didn't work anyways. [22:33:57] jdlrobson: ping me when you've uploaded a patch tho cuz i'd love to review it [22:34:37] although it'd prob be tomorrow, T121152 will prob be my last task of the day besides legal review [22:39:17] https://en.wikipedia.org/wiki/Blowing_a_raspberry < [22:39:47] bearND: looks like the same thing happening on my personal laptop. would you mind uploading a patch to add the icon? that might be best for now. still wondering what i could be doing wrong. [22:40:37] bearND: just drop the original into the right bucket in /icon-svgs and then 'python convert-icons.py define.svg', right? [22:41:19] mdholloway: that's correct. I'll add a patch for the icon right after the meeting I'm in is over [22:41:25] bearND: thanks. [22:41:40] kaity_: same result on my end with the new svg, by the way, but it seems to work for bearND. [22:41:46] jdlrobson: ok, that only works on desktop but looks more like a mobile web design [22:42:01] bearND: what only works on desktop? [22:42:17] https://en.m.wikipedia.org/wiki/Blowing_a_raspberry?mobileaction=beta [22:42:42] the read more thing. Or, maybe the setting from desktop does not translate to mobile web [22:43:33] jdlrobson: cool, i see it now [22:43:37] bearND: it will show for all beta users unconditionally on mobile [22:43:40] on mobile web, too [22:44:14] bearND: we are noticing 2s delay on loading read more - did you guys hit this problem with cirrus search? [22:44:16] cirrus search is slow [22:48:16] https://phabricator.wikimedia.org/T121254 < bearND bgerstle_mtg [22:59:34] niedzielski: new native bar looks good! [23:01:08] niedzielski: saving an article feels much better now. yay! [23:02:44] kaity_: yay! hey there's a preexisting save page bug i'm working on. it should be a bit nicer after that [23:03:06] mdholloway|afk: https://gerrit.wikimedia.org/r/#/c/259173/ [23:03:25] jdlrobson: are we meeting? [23:05:18] bgerstle mhurde coreyfloyd hey we need you guys on a hangout asap. adam is calling a hangout [23:05:31] bgerstle_mtg ^ mhurd ^ [23:06:15] JonKatz: link? [23:08:37] dr0ptp4kt_: what hangout? [23:09:01] mhurd adam is sending a link, we're in brown if you want to make it a meat party [23:25:45] jdlrobson: not sure why, but related articles isn't working on my vagrant instance anymore... despite it working earlier when i tested your release. Gonna check with a standalone wiki, but may need some assistance later [23:36:46] yeah it's not working in a standalone wiki either :/ [23:36:54] bmansurov, when you get on you might be able to assist as well [23:50:21] bgerstle_mtg: quick bat cave? 60 seconds? :)