[00:02:01] New patchset: JGonera; "Story 487: Footer update" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61474 [00:02:09] awjr, ^ [00:05:05] awjr: replied on bug and gerrit [00:06:05] thanks a lot Krinkle [00:06:29] thanks jgonera, i'll take a look in a minute, im currently working with someone to try and figure out wtf is wrong with the card transitions in mingle [00:06:39] ok, sure [00:07:10] jgonera: If you can test that jquery.tablesorter works without problems on mobile and propose a change to enable the mobile target for it, I'd be happy to merge/deploy that swiftly. [00:07:21] in time for your Tuesday deployment window [00:12:43] ok mingle should be doing transitions correctly now on the story wall [00:13:06] some crucial option somewhere had been unset [00:15:03] Krinkle, I think Jon used tablesorter as an example but we don't need it atm [00:15:30] yeah i believe tablesorter was what triggered the first time we saw that behavior [00:15:33] jgonera: isn't it throwing uncaught exceptions right now? Causing other code not to run. [00:16:04] If it [00:16:06] If it [00:16:17] Krinkle, I hope it doesn't, the only thing we're mocking now is mediawiki.action.edit [00:16:30] but we better ask Jon tomorrow [00:16:53] I mean, this bug is not anything crucial, as far as I'm concerned [00:17:03] If it's not an issue, then no problem by me either. But I imagine it causes troubles and awjr mentioning that it is to be done before/during the next deployment window confirmed it for me. [00:17:32] I'm here if you need me. Let me know whether or not it is critical, I'll let you (plural) be the judge of that. [00:17:35] awjr, where have you seen tablesorter triggering it? [00:18:08] doh we didn't record it in the bug [00:18:29] i might be able to find it with some random clicking around [00:18:42] I think it's been like that for some time, so I guess we can wait for Jon's input [00:19:25] yeah, it has apparently been a problem for a while [00:20:40] hmm i imagine any article with a sortable table should do it... [00:21:43] surprisingly, https://en.m.wikipedia.org/wiki/Table_(information) doesn't have one [00:22:15] ahha [00:22:17] https://en.m.wikipedia.org/wiki/Wikipedia:Sortable_table [00:22:21] jgonera, Krinkle ^ [00:22:23] there's an example [00:23:37] ok, should I check if it works on mobile now and prepare a patch to enable it? [00:24:14] awjr, ^ [00:26:33] sure jgonera [00:28:54] huh, seems like Special:Import is broken in core's master [00:30:38] Exception thrown by mediawiki.page.ready: Unknown dependency: jquery.tablesorter [00:30:38] yeah [00:30:55] so that looks fairly high priority. You shouldn't have run-time exceptions :) [00:31:26] yeah, unfortunately it seems that tablesorter messes up tables with our CSS [00:31:46] I doubt I'll have time to figure that out before tomorrow's deployment [00:37:20] New patchset: JGonera; "Story 487: Footer update" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61474 [11:06:40] New patchset: Jdlrobson; "Bug 47857: Deprecate #mw-mf-searchForm, allow search term clearing" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61553 [11:30:37] New review: Jdlrobson; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) C: -2; - https://gerrit.wikimedia.org/r/61474 [12:54:00] New review: Jdlrobson; "Since the nearby page fully relies on javascript, I've been thinking of the nearby page as a mini ap..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/60684 [12:55:21] New review: Jdlrobson; "Also you forget the add image to this page CTA only shows to logged in users. Non-logged in users (w..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/60684 [12:57:45] New review: Jdlrobson; "jgonera: got it - page.js is part of mobile.beta - and that wasn't listed as a dependency. Should no..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/60684 [12:57:50] New patchset: Jdlrobson; "Story 425: Nearby preview" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/60684 [12:57:50] New patchset: Jdlrobson; "Alpha: Generate dynamic sections using templates" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/59785 [13:08:27] New patchset: Jdlrobson; "Bug 47598 Make Special:Nearby work on desktop" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61496 [13:45:41] New patchset: Jdlrobson; "Alpha: Generate dynamic sections using templates" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/59785 [13:50:19] New patchset: Jdlrobson; "Story 436: Use file name for images with descriptions with templates" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56613 [14:12:30] New patchset: Jdlrobson; "It's fail not error" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61570 [14:12:31] New patchset: Jdlrobson; "Story 425 Nearby: Cache last known location/nearby places" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61571 [14:35:19] hey guys :) [14:35:24] mobile apps question [14:35:45] for iOS, where in the code can I find the user agent string being built? [14:35:53] I don't speak Objective C but this looked like a good place to start: https://github.com/wikimedia/WikipediaMobile/blob/master/Wikipedia-iOS/Classes/AppDelegate.m#L39 [15:25:01] New review: Jdlrobson; "recheck" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61571 [15:38:17] New patchset: Jdlrobson; "Beta: Lazy load sections in beta" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61580 [15:49:39] New patchset: Krinkle; "nearby.js: Promise fail(), not error()" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61570 [15:49:56] New review: Krinkle; "(retrigger Jenkins job as it was lost)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61570 [15:50:00] New patchset: Krinkle; "Story 425 Nearby: Cache last known location/nearby places" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61571 [17:02:44] hmm... why's everything so quiet ? [17:02:49] * awjr shouts [17:03:04] * brion MOAR COWBELL [17:03:05] pulls earplugs out [17:03:19] * yurik puts them back in [17:03:39] hehe [17:05:54] i think everyone is stuck in tax limbo [17:06:09] tax limbo? [17:06:35] i finished my taxes on time! april 14th. :P [17:07:00] bbiab [17:30:39] mhurd: is tomasz in? [17:31:16] tfinc: mhurd i'm on the hangout [17:31:23] heh [17:31:36] brion: ya [17:39:59] hey Maryana [17:41:29] brion: heh [17:41:32] brion: do we have to submit separate wikimania proposals ourselves? [17:42:03] i'll ask tomasz [17:42:27] alright [17:42:32] did you do one? [17:42:47] I should get out of IRC before oril finds me [17:43:00] YuviPanda: if you're talking about the app then merge it with tomasz's [17:43:05] if you have something else then make anew one [17:43:09] but hurry today's the deadline [17:43:30] brion: i'll be happy to just do it with tomasz [17:43:35] but he's got it coverd i'd think [17:43:53] yeah [17:44:03] YuviPanda: do you have the link to his preso page? feel free to add to it :D [17:44:12] https://wikimania2013.wikimedia.org/wiki/Submissions/Why_Apps,_Commons,_and_Wikimedia_go_great_together [17:44:16] i'll have to decide if i want to do another preso on my research projects [17:44:19] maybe svgedit [17:44:24] brion: 25 mins is not a lot... [17:44:26] perfect [17:44:31] talk fast :D [17:44:34] plus I want to somehow work in the Picard / Riker double facepalm... [17:44:42] brion: oooh, just like last time, when tfinc went 2x over time? :P [17:44:47] hehe [17:45:49] ok i'm going to head into the office in a few. catch y'all in 45-60. YuviPanda don't forget to study for your exams and stay offline :D [18:02:42] yurik, okay if i update https://www.mediawiki.org/wiki/Requests_for_comment/Zero_Architecture to account for the 'partnered' field, replace 'bannerStyle' with 'background' + 'foreground' + 'fontSize', explain that whitelistedLangs needs an explicitly empty array if going for '(all)'. if you'd prefer to update it, that's cool, too, but i'm happy to do it! [18:02:49] tfinc: yes, but I'm here to handle Wikimania [18:03:03] i'll be out before ori spots me again, I swear. [18:03:13] dr0ptp4kt, i could do it, script is the most important :) [18:06:55] New review: JGonera; "I will consult this with Vibha. It seemed a bit sloppy and unordered for me, that's why I tried this..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61474 [18:08:39] dr0ptp4kt & yurik, need anything deployed today? [18:08:50] MaxSem, nope, not yet [18:09:07] MaxSem, dr0ptp4kt, lets plan next week? [18:09:07] YuviPanda: :) [18:09:19] YuviPanda: just iterate on the talk that i sent [18:09:22] change what you want [18:09:39] this is is clearly aimed at commons, andriod, ios, apps [18:10:02] tfinc: I want to do one deep diving into campaigns [18:10:11] YuviPanda: nice [18:10:20] tfinc: almost done with it, lemme give you a link in a while [18:10:28] YuviPanda: can you rev on mine as well ? [18:10:34] sure [18:11:34] jesus how am I gonna get this thing past 300 words... [18:21:01] yurik, i'd also like to add a field called 'geographiesToSuppressZeroUnsupportedWarning' to the . this is to account for dfoy's card "Create ability to optionally suppress 'Your carrier does not support Zero' warning banner in a specific country'. today, we won't use it, but i think it would be nice in the rfc. [18:21:21] yurik, agreed with you on planning for next week with MaxSem for any pushes. [18:21:47] dr0ptp4kt, we could add it, but lets do it after we deploy the first version [18:22:00] adding such values is very easy - just one more item [18:22:10] in the validation function [18:22:32] yurik, so should i just set a reminder to add the verbiage to the rfc, or go ahead and add it to the rfc now? [18:22:52] reminder is fine [18:23:08] we will have to figure out the country (which we don't have yet) [18:23:57] yurik, sounds good. shall i modify the 'partnerId' and 'bannerStyle'-related, and 'whitelistLangs' parts of the rfc now, or just set a reminder for that later, too? [18:24:33] dr0ptp4kt, reminder is fine - i will modify the RFC with all the changes [18:24:49] yurik, sounds good, will add you to the calendar item. [18:25:52] MaxSem: just looked over https://gerrit.wikimedia.org/r/#/c/61539/ - i think it's good to go [18:26:12] cool [18:35:02] man no tfinc [18:36:52] MaxSem: is the only way we determine whether or not to display WML by examining Accept headers? [18:37:18] (aside from ?useformat) [18:37:42] well [18:38:28] if accept contains vnd.wap.wml AND user-agent does not match any known device [18:38:54] ah yeah [18:38:55] ok [18:43:45] =hey brion [18:43:46] see https://wikimania2013.wikimedia.org/wiki/Submissions/Guided_Mobile_Contributions_to_Commons_via_Campaigns [18:43:56] ooh [18:44:02] nice [18:44:07] give it a read? [18:44:18] english is not my native language, etc [18:44:54] brion: are you also looking at mobile-feedback-l commons feedback? There was one earlier today [18:45:09] brion: and one at Commons_talk:Mobile_app [18:45:15] hey tfinc [18:45:16] see https://wikimania2013.wikimedia.org/wiki/Submissions/Guided_Mobile_Contributions_to_Commons_via_Campaigns [18:45:17] i'll catch up ;D [18:45:31] ok :) [18:45:34] YuviPanda: will take a look at in the afternoon [18:45:35] thanks [18:45:42] YuviPanda: just got my dell XPS linux laptop \o/ [18:45:43] tfinc: alright, email me comments? i'll be off ow. [18:45:45] i'll set up my android stuff on it [18:45:48] brion: man i'm jealous :P [18:45:49] YuviPanda: go study [18:45:51] go study :D [18:45:53] hehe [18:45:56] yes boss! [18:55:44] yuvi: I replied to Brianna from mobile feedback. her followup is relevant, though. [18:56:01] ^ Brion [18:59:35] oh noes. kaldari just started scapping. this might take more than an hour:) [18:59:45] ruhroh [19:00:15] oh is this for echo deployment? [19:00:23] that could definitely take a while [19:01:59] good news is that we don't need to scap ourselves [19:01:59] \o/ [19:02:19] MaxSem: dunno if jon mentioned but i think he had at least one config change to go out [19:03:45] MaxSem: https://gerrit.wikimedia.org/r/#/q/operations/mediawiki-config+(owner:jdlrobson+OR+owner:maxsem+OR+owner:jgonera+OR+owner:awjrichards)+is:open,n,z [19:05:22] uhhh [19:05:37] first noticeable scratch on my phone's screen [19:05:52] :( [19:15:33] ragesoss: on the commons page? i'll look in a sec [19:16:27] Brion: no, the feedback mailing list [19:16:45] ok [19:20:26] ragesoss: remind me which list that is? [19:20:46] i don't see mobile-feedback on our listserv [19:21:43] brion: I forwarded it to you. [19:21:52] ah it's hidden [19:21:57] i must not be on it [19:22:37] moderated. tfinc can you approve my subscription to mobile-feedback-l? thx :D [19:22:52] ragesoss: thanks, looking [19:23:12] it's 50% spam, 45% bugs and feature requests for Wikipedia app. [19:23:26] hehe [19:23:38] ragesoss: ok good feedback from brianna, i'll put some notes in our planning board [19:24:58] the license bug was from her first email. [19:26:35] yeah, already put that on the board :D thx [19:26:46] we have an indicator on iOS, but that should be more prominent too [19:27:19] and you'll see in your inbox her report of laggy input. [19:28:06] also, completed uploads notification suggestion that has a bug already. [19:28:23] laggy input? booooo :( [19:30:07] see followup. could have been just during initial sync. [19:31:00] yeah [19:31:04] that might be slowing things down [19:31:28] i'll send her some replies in a bit [19:31:40] I'll test uploading immediately next time I install it fresh on my 2.3 device. [19:32:50] okay, now you have all the emails I exchanged with her. [19:36:27] ragesoss: thanks that's super valuable feedback [19:44:07] Maryana_brb, did you get lunch? [19:44:14] was just going to do that [19:46:00] heya mobile peeps, who does your deployments, usually? [19:46:23] eitherway, it looks like kaldari's scap is going to run over his timeslot, he'll either let you know when he's done, or you can ping him [19:46:26] * MaxSem  [19:46:41] I already noticed that:P [19:58:42] MaxSem: good deal :) [20:02:00] haha the windows phone 8 device i got to test with can't connect to the main WMF network [20:02:07] i can connect to WMF-Guest though [20:02:27] dfoy, dr0ptp4kt meeting? [20:02:57] yurik, i'm in the room. will join hangout [20:08:38] awjr, https://gerrit.wikimedia.org/r/61629 [20:09:07] i feel like there was a reason for that... [20:09:09] let me think [20:09:32] the reason was me being paranoid [20:11:25] MaxSem: yeah i think that should be ok, esp since you're checking shouldDisplayMobileView() in the setup function [20:11:57] i cant think of a reason why you'd care if $wgMobileUrlTemplate were set [20:12:43] brion: mhurd zz_YuviPanda : fancy http://blog.trello.com/introducing-business-class/ [20:13:11] ooh [20:13:23] Maryana_brb: who "owns" our account to upgrade to this http://blog.trello.com/introducing-business-class/ ? [20:13:30] upsell ftw :D [20:13:36] tfinc, not sure.. [20:13:40] ask stevenw? [20:14:03] "Hey there! You must be an admin of the organization to get Business Class." [20:14:10] Maryana: is stevenw our admin ? [20:14:14] awjr, jcmish, jgonera, Maryana - we're live on testwiki, please test:) [20:14:26] woohoo [20:14:28] on it [20:14:54] thanks MaxSem [20:22:03] everything's looking good for me so far [20:22:25] stable and beta looking good [20:22:30] on to alpha [20:22:42] deployed all the config changes [20:22:48] ...on test [20:23:04] Missing_wiki fun! [20:23:25] jcmish, do all tests pass? because there was a commit that removed an id I still see used in Cucumber tests (tests/acceptance/features/support/pages/home_page.rb:35: form(:search_form, id: 'mw-mf-searchForm')) [20:23:52] jcmish, i have procured a kindle! [20:23:54] the search for one failed but it just warns for that [20:24:10] and I do it again as more extensively part of watchlist [20:24:22] ok [20:24:22] it uses random and different workflow [20:24:30] meh [20:24:40] sweet thanks Maryana [20:25:00] Maxsem so the Missing wiki bug is still there? [20:25:07] yes [20:25:07] or I can't test that yet? [20:25:43] jgonera: the footer changes never got merged right? [20:25:43] after more thinking, I think I'll avoid deploying it today [20:26:05] I looked at my list and it still wasn't on the git logs but I thought I better ask again [20:26:09] :-) [20:26:10] jcmish, right, not merged [20:26:17] k cool [20:26:35] awjr: the Meta(tm) thing [20:26:42] eh? [20:26:44] do I have to wait until we're on production? [20:26:52] hold please I shall grab the ticket [20:27:11] jcmish: yeah; you'd only see that change on meat (assuming it works right) [20:27:11] er [20:27:12] meta [20:27:23] gotcha [20:27:34] hmm, i think the kindle bug is still not fixed :-/ [20:27:38] I made a whole bunch of notes to make sure I went over these [20:27:48] also, how in the world does anybody use this stupid keyboard thing?? [20:27:52] HAHHA [20:27:59] Maryana: I've never used a kindle [20:28:20] going back to looking at alpha [20:31:43] Maryana: a kindle keyboard? [20:32:04] or a touch keyboard on some other kindle? [20:33:27] ok, looks like old bug is fixed but new bug is new on kindle [20:33:35] :( [20:36:15] we're also seeing the app banner in beta [20:37:30] that is expected, right Maryana? [20:37:44] it appears the campaign is turned off on meta, so it shouldn't actually go live once we push to production [20:38:05] ok, there's something weird with upload count for me [20:38:08] not sure if this is new [20:38:14] yeah- but it is functional (though there's still a problematic bug with CN) [20:38:23] awjr, right [20:38:26] I open uploads dashboard and I saw for a second that I have 6 uploads [20:38:43] then the count disappeared and the photo upload tutorial showed up [20:38:55] that is weird. [20:39:01] so I uploaded a photo and suddenly it says that I have 7 uploads (but only 1 upload visible) [20:39:18] confirmed ^ [20:39:23] same thing just happened to me [20:39:23] then I refreshed the page and it says 6 uploads again (only 1 photo which I've just uploaded there) [20:39:35] that is not happening for me - device? [20:39:42] i'm on iphone 4 [20:39:54] Android browser 4.1.2 If i remember correctly [20:41:13] awjr, Android browser 4.2.2 [20:42:42] i just noticed something else weird - chrome on android 4.2.2, when i tap the search box, i seem to get zoomed in [20:43:16] and when i tap the arrow to exit the search overlay, i am indeed zoomed in a bit (watchlist start no longer fits on the screen, for instance) [20:43:25] but i can zoom out and all is well [20:43:29] well, that's Chrome's behavior [20:43:30] this is on stable. [20:43:45] really? wow i have never noticed that before [20:43:47] awjr: is that in alpha? [20:43:56] jcmish__: stable [20:44:17] I told Jon several times that we should disable the viewport but he insists that people should be able to zoom the page when they want (which I kind of agree with, but not if it makes the page zoom itself on Chrome) [20:44:18] I saw that on alpha for ipad and safari with ipad but it only happened in alpha [20:44:33] and Jon said he couldn't repro it [20:44:56] probably something changed after the search overlay rewrite [20:45:03] finished the testing with the opera mini and making sure Nearby behaves correcty and it does [20:45:09] something small but apparently important enough for Chrome to start zooming [20:45:13] i just noticed it login too, jgonera, after tapping the username field i got zoomed in a bit [20:45:16] or maybe a new version of Chrome [20:45:39] it's less obvious on login tho since there arent really many other page elements [20:45:41] yep, it will do it with 90% of input fields unless it "thinks" that it's big enough without zooming [20:46:08] that's one of Chrome's features I don't like ;) [20:46:11] sheesh. [20:46:20] +1 [20:46:29] jgonera, Maryana, was the weird upload behavior in stable? [20:46:41] it doesn't do it on mine phone though, possibly because of a different screen size or resolution [20:46:50] i just saw it in beta [20:46:53] haven't checked stable [20:46:55] awjr, it was alpha, let me check what it shows in stable [20:46:55] * Maryana looks [20:47:25] yeah, stable, too [20:47:26] awjr ^ [20:47:30] :( [20:47:35] doesn't happen for me in beta or stable on desktop chrome, desktop chrome pretending to be android, or chrome on android 4.2.2 [20:47:48] I can't get it to happen either [20:47:57] on my ipad and on the emulators [20:47:59] on either [20:48:06] awjr, it does zoom for me in login, not in the search bar though, but as I said it depends on the device's screen and Chrome's algorithm for that [20:48:27] stable shows 6 uploads with only one photo present [20:49:27] I guess the tutorial checks if there are images on the page instead of using the upload count, that's why it popped out [20:50:05] Arrrrrrrrr [20:50:08] huh, i dont see the problem on android browser on andoird 4.2. [20:50:19] is it a race condition, jgonera? [20:50:27] probably depends on something Maryana and I did on our accounts [20:50:39] the upload count? I don't think so [20:50:49] probably something unexpected in DB [20:50:57] i can try with my staff account [20:51:02] i was using my volunteer account to test [20:51:15] hmm aren't we pulling the count directly from the db on the backend? [20:51:25] awjr, I think we are [20:51:29] oh [20:51:39] is it the deleted images thing? [20:52:13] I've got a 1st day problem again [20:52:16] same for my staff account [20:52:16] deleted images thing? [20:52:19] oh no MaxSem|android! [20:52:27] need me to finish the deployment? [20:52:31] it might be, but still, I'm not sure why it would not increase to 7 permanently (it increased only because JS increases it on the page after a successful upload) [20:52:53] awjr, so when admins delete our images, the count doesn't decrease, does it? [20:53:22] im… not sure, actually jgonera - MaxSem|android might know better than me [20:53:37] if we're pulling from the image table, then yes [20:53:47] in fact, we must be [20:54:03] there's no easy way to detect deleted upload count otherwise, i don't think [20:54:25] so yeah, when you have images deleted, they're subtracted from your uploads count [20:54:47] but, that still doesn't explain why we're all suddenly seeing a first-time uploader tutorial :) [20:55:05] all my uploads are still there on desktop [20:56:11] i blame echo… somehow [20:56:17] oh, let me check if I can see my uploads on desktop [20:56:52] so yeah, upload count comes directly from the image table in the db [20:57:23] we pull from the db slaves, so consistency maybe an issue [20:58:23] outside of the upload issues that you guys jgonera and Maryana are experiencing every thing else looks good to me I'll just hang out until I hear for you guys [20:58:25] huh, yes, most of my uploads are still visible on desktop, but not on mobile testwiki [20:58:36] woah, i think something's up w/the commonswiki db [20:58:46] i can't ssh in [20:59:36] you shouldn't be able to directly ssh in to the commonswiki db - are you going through fenari, Maryana? [20:59:45] stat1 [20:59:54] ah that i dont know about [21:00:50] something weird's going on. maybe unrelated [21:00:52] I had some issues when connecting to EL DB today, it would disconnect me when running some queries [21:00:56] but i can't access the commonswiki db at all [21:01:12] hmm [21:01:21] were you able to before Maryana? what is the specific machine you're trying to access? [21:02:27] yeah, this just started today. trying to log into the researcher mysql account on db1047 [21:02:40] i'll check from fenari [21:03:09] im able to get into the commonswiki db from fenari (though that connects directly to a master) [21:03:23] so, not sure about specific mysql hosts you may have direct access to [21:03:37] maybe ops decided to boot me. heh. [21:03:44] aww :( [21:04:50] so jgonera were you saying htat you first saw your correct upload count briefly, before the tutorial popped up and your upload count went to 0? [21:06:15] awjr, I saw my correct upload count from commons, the tutorial showed up (upload count disappeared), I uploaded, upload count was increased by 1 by JS, I refresehd, upload count was back to 6 (the amount of images I have on commons) [21:06:20] I think I know what's going on [21:06:47] we're probably checking uploads count from commons, but for some reason, we're no longer uploading to commons from testwiki [21:07:08] because my new upload doesn't show up on commons, only on mobile testwiki [21:07:30] yes, when I click it to see upload details, I'm still on testwiki [21:07:50] that means a small bug we should fix: upload count should use the same endpoint as upload themselves [21:07:54] yeah, i just uploaded and checked the api URL - it's testwiki [21:08:00] this won't have any effect in production [21:08:23] ahh ok [21:08:27] (shouldn't have ;) [21:08:46] jgonera: does all of that happen in uploads.js or is some of it handled elsewhere? [21:09:03] the upload count is generated server-side I guess [21:10:27] so the backend should be pulling upload count for whatever $wgMFPhotoUploadWiki [21:10:34] i figured out the the slave db issue. unrelated but being fixed. [21:10:35] hm... [21:11:20] which for testwiki, is '', which means it should be querying the local wiki [21:11:38] it seems, there's a separate config option [21:11:43] yeah [21:11:44] wait [21:11:45] ah [21:11:50] the photo upload url endpoint or whatever [21:12:15] there's wgMFPhotoUploadEndpoint and wgMFPhotoUploadWiki [21:12:16] which for testwiki currently points at commons [21:12:22] yeah [21:12:38] which one points at commons? the -WIki one? [21:13:00] http://pastie.org/7744957 [21:13:05] so sorry, got that backwards [21:13:24] oh wait, no, im tripping [21:13:27] jeez i cant read [21:13:34] they are *both* pointing to commons [21:13:41] test2wiki != testwiki [21:13:53] @_@ [21:13:53] then I don't understand... [21:14:09] Hey, didn't the WLM app last year have maps of some form? [21:14:39] mw.config.get( 'wgMFPhotoUploadEndpoint' ) on testwiki returns "" [21:14:55] hmm [21:15:11] marktraceur: yes [21:15:54] awjr: Would that solution be general enough to include it in UploadWizard? [21:16:06] jgonera: ok then either there's a bug or there's a hack in produciton - lemme double check production config [21:16:27] marktraceur: in the middle of debugging during deployment - can we chat about it later? [21:16:39] No problem :) [21:16:39] anyway, I'd proceed with the deployment if this is the only thing stopping us, and figure it out later [21:16:43] :) [21:17:55] jgonera: there must be a bug somewhere; there's no live hack or anything weird in production config... [21:18:02] hm [21:19:12] Its intentionally set this way on test [21:19:23] btw awjr, MaxSem|android, I'm not sure why SpecialUploads::getUserUploadCount() fetches all the rows instead of just doing select count(1) [21:19:54] back in business [21:19:59] nice MaxSem [21:20:00] \o/ [21:20:23] jgonera: there's a reason, probably documented in the patchset review comments i think [21:21:09] 'wmgMFPhotoUploadEndpoint' => array( [21:21:10] 'default' => '//commons.wikimedia.org/w/api.php', [21:21:10] 'commonswiki' => '', # bug 45256 [21:21:10] 'test2wiki' => '', [21:25:03] awjr, MaxSem, can't see anything that would explain not using select count(1) at https://gerrit.wikimedia.org/r/#/c/48978 [21:25:12] ok i take it back then [21:25:57] unless that's because of some MW weirdness that I'm not aware of ;) [21:26:26] i dont think so, but i vaguely recall a converstaion with MaxSem about it [21:26:39] jgonera, imagine someone with 10000s of uploads [21:27:02] count() can be expensive [21:27:07] oh, ok, that's about iterating? [21:27:11] thus, we restrict it to 501 rows [21:27:19] ok, now I recall, I'll just add a comment there then [21:28:39] so i dont understand why this is borked on testwiki and i am concerned about it [21:28:48] iirc last week it was uploading to commonswiki [21:29:51] and considering config shows it should be using commonswiki, but it is not in practice… [21:30:22] New patchset: JGonera; "Add a comment explaining why SQL's count is not used" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61694 [21:30:23] MaxSem, is it overwritten somewhere else? [21:30:36] mhm [21:30:59] 'wmgMFPhotoUploadWiki' => array( [21:30:59] 'default' => 'commonswiki', [21:30:59] 'test2wiki' => '', [21:30:59] ), [21:31:19] that is correct [21:31:42] as is what's configured for wmgMFPhotoUploadEndpoint [21:31:44] anyway, we need to deploy or rollback [21:32:00] no scap, right MaxSem? [21:32:04] yes [21:34:02] let's do it :) [21:34:50] i dunno jgonera, my local wiki is configured with an endpoint other than itself, and mw.config.get('wgMFPhotoUploadEndpoint') works [21:35:09] Maryana: you are ok deploying with the possibility that the dashboard is wonky? [21:35:09] let me just check one more thing [21:35:38] yeah, since it seems like we're still not 100% sure if it's a testwiki issue or not [21:36:39] before we go [21:36:42] let me try one more thing locally [21:37:53] awjr, nothing changed in the code that sets this var since Feb 28 [21:38:01] at least that's what git blame says [21:38:15] yeah ok, and i still havent been able to repro locally so... [21:38:16] let's do it [21:38:29] ok [21:38:50] pushing to wmf3 [21:39:24] exciting times! [21:40:11] today, wmf3 = mediawiki, test2wiki, testwiki [21:44:10] [Commons-iOS] montehurd opened pull request #54: Test mode toggle now refreshes images (master...testModeClearData) http://git.io/OFbJaw [21:44:26] brb grabbing coffee from the kitchen [21:47:45] grr [21:47:59] Gerrit plays slowpoke today [21:48:04] uhoh [21:58:08] MaxSem: looks like we're live everywhere else now? [21:59:01] yes, pushing wmf-config [21:59:07] cool [22:00:06] done [22:00:49] uploads looks fine to me now in production [22:00:58] styles are looking whack for me [22:01:08] maybe need to wait for cache to expire [22:05:46] page html hasn't changed at all, has it? [22:06:39] im seeing something worrying - main pages of enwiki and m.mediawiki look broken in a fresh session (i am hitting varnish cache) but if i bypass cache (eg by opening settings page and returning to main page), pages look fine [22:06:39] enwiki looks fine, but i just went to wikivoyage to check the nearby bug, and it's all fubared [22:06:53] Maryana: try enwiki with a fresh session [22:07:00] (no cookies) [22:07:08] * Maryana tries [22:07:44] gaah [22:07:49] yeah, that's what i was seeing on wikivoyage [22:07:51] :-/ [22:08:00] MaxSem, jgonera has page html changed at all with this deployment? [22:08:02] weird floaty hamburger & star [22:08:09] shoudn't have [22:08:13] yeah that's what im seeing too and i think it's stuck in the varnish cache [22:09:06] &debug=true fixes it [22:09:28] until you go back to the URL with no &debug=true :p [22:09:36] MaxSem: are you seeing this too? [22:09:55] awjr and Maryana I am [22:09:59] no [22:10:14] MaxSem: try with a fresh session (no cookies) [22:10:26] to ensure you're hitting varnish cache [22:10:26] right [22:11:30] oh [22:11:33] well shucks then [22:12:09] but we are seeing it, and i am feeling fairly certain this is the result of stuff cached in mobile varnish (RL caches would have expired by now) [22:13:30] fresh session screnshot: http://imgur.com/vY20Xqf [22:13:32] jgonera isn't seeing it on his phone, but it's showing up on my iphone and a test ipod [22:14:45] varnish cache bypass: http://imgur.com/NwgJUag [22:15:09] same for me awjr [22:15:21] first for regular mobile site, second for debug=true [22:15:27] yeah [22:15:38] :-/ [22:16:22] MaxSem: any thoughts? [22:17:16] basically, it's revert vs. flush [22:18:43] well, we'd have to flush if we revert too now [22:18:59] because by now pages would be cached with presumably no html [22:19:04] s/no/new [22:19:22] awjr, the html might have changed slightly because of the chrome header changes [22:19:34] jgonera: chrome header changes? [22:19:48] I mean, the header in our chrome, not chrome the browser [22:19:57] oh [22:19:57] ha [22:20:12] do you have a patchset handy jgonera? [22:20:20] let me see [22:21:11] awjr, https://gerrit.wikimedia.org/r/#/c/60579/ [22:21:36] so yes, some changes in CSS classes and ids [22:22:00] sigh [22:22:08] that is no bueno [22:22:21] why? [22:22:40] we can't have page html change like that because it requires a cache flush [22:23:02] html gets cached in mobile varnish for like.. 30 days i think [22:23:07] New patchset: MaxSem; "Revert "Fix header styling"" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61705 [22:23:18] ^-^ [22:23:19] hm, so we need to support both html versions for some time if we need to make changes to html? [22:23:23] :( [22:23:31] jgonera, yes [22:24:25] :/ [22:24:27] hm [22:24:30] jgonera: yeah, or otherwise we need to discuss it first and communicate with ops, etc [22:24:43] I see, I haven't thought about that [22:24:52] MaxSem: even if we push that, we'll still need to flush :( [22:25:05] awjr, a few pages is ok [22:25:16] flush is the thing we need to schedule with ops? awjr and MaxSem ? [22:25:24] yes [22:25:30] jcmish: well, something we need to avoid like the plague, really [22:25:31] but yes [22:25:46] :( [22:26:12] so i did look at everything else and we're looking good [22:26:25] including Nearby not showing up on Opera Mini [22:26:26] jgonera: will reverting that change cause any other issues? [22:26:42] I'm not sure... probably not [22:27:13] you should revert this one too: https://gerrit.wikimedia.org/r/#/c/60919/ [22:27:34] I don't see any other patchsets that could depend on it [22:27:34] yeah [22:28:00] ok - well, MaxSem i guess we can try it - worst case we'll have to do a flush anyway, best case we avoid it for now [22:28:21] New patchset: awjrichards; "Revert "Change remaining #mw-mf-header occurrences to .header"" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61707 [22:28:49] awjr, whee - conflicted [22:28:54] sigh [22:29:08] ugh, will have to figure out how to tackle this change later [22:29:30] MaxSem: are you already dealing with the conflict or shall i? [22:29:31] hm, really? I can't imagine a conflict there [22:30:52] there were only two other commits after that and they don't change much, mainly the hooks [22:31:31] oh it looks like it's probably due to minerva stuff [22:34:39] MaxSem: ? [22:34:49] trying.... [22:34:57] dammit [22:36:00] frak [22:37:24] testing... [22:43:56] brion: mhurd MaxSem : this comment is really interesting. letting GPS coordinates attached meta data for commons uploads [22:43:57] http://blog.wikimedia.org/2013/04/29/announcing-the-official-commons-app-for-ios-and-android/#comment-242729 [22:44:07] zz_YuviPanda: --^ [22:44:38] imagine being in front of 149 new montgomery, taking a photo, and then having a selection of meta data to choose from that comes from GeoData and/or WikiData [22:44:38] tfinc: the gps one? [22:44:41] yeah [22:44:47] yeah [22:44:51] definitely one we'd like to add [22:45:02] just let us know how fast to prioritize it :D [22:45:16] it would make each image have a ton more data from the beginning which will make commons reviewers much happier [22:45:20] yeah [22:45:33] wow xchat does *not* like me bumping up the system font size [22:45:39] brion: :) yup. i'm going to noodle on it for a bit. thanks for respecting priorities and not just jumping on it [22:45:59] hehe [22:46:17] ah, restarting xchat makes it much happier [22:46:41] New patchset: Yurik; "Zero config editor & visualizer" [mediawiki/extensions/ZeroRatedMobileAccess] (master) - https://gerrit.wikimedia.org/r/60775 [22:46:48] jgonera, I tried reverting it - now looks as it does now in prod:( [22:47:10] MaxSem, let me see [22:47:24] committting shortly [22:47:48] New patchset: MaxSem; "Revert "Fix header styling"" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61705 [22:48:00] jgonera, ^^ [22:48:24] MaxSem, it looks ok for me in prod [22:48:30] what is you're experiencing? [22:48:53] are you sure it's reverted? [22:49:04] jgonera, try ^^^ [22:49:15] review you mean? [22:49:22] test it [22:50:09] uh, it looks broken [22:50:57] MaxSem, there's something wrong with the CSS [22:51:03] in that reverted commit [22:51:15] MaxSem: do we need to just roll back? we're awfully close to lightning deploy [22:51:18] magmificent [22:52:06] awjr, I guess :( [22:52:36] can we roll back? [22:52:45] yeah [22:52:52] let's just do that, I'm not sure what's going on at this point [22:52:58] regardless, the longer we take the more pages are gonna get cached with the new html :-| [22:53:35] can we flush only the cache entires that were updated in the last few hours? [22:54:11] entries* [22:54:11] no [22:54:16] i dont think so jgonera [22:54:31] not cool :/ [22:54:36] from how it's been explained to me before it's kind of a one-off, or all-or-nothing thing [22:55:13] ok, so do we roll-back or should I try to find the problem with this revert commit? [22:55:38] i vote roll back, this is taking too long [22:56:02] I vote roll back too, no supercrucial new feature got deployed today [22:56:14] yeah same here with MUCH sadness [22:56:20] already prepared the commits, standing in the line [22:56:30] cool [22:56:36] yeah wtf is up with gerrit? [22:57:00] DoD [22:57:03] *DoS [22:57:07] orly? [22:57:08] :( [22:57:10] can we, after that, revert those revert commits? ;) because they didn't exactly get us to the previous state [22:57:14] hehehe [22:57:24] revert the reverts! [22:58:35] that's gonna give me lots of crappy work tomorrow, fixing this + avoiding the same story after merging the footer commit [22:59:11] the main revert wasn't merged [22:59:27] main revert? [22:59:29] hmm we should've thought of that when discussing the footer changes [22:59:37] :( [22:59:37] oh, ok, I'll just check what state master is at [22:59:55] :( [23:00:01] * Maryana gives everyone a cookie [23:00:09] ( :: )))))))) [23:00:17] hehe [23:00:27] I want alcohol [23:00:36] * brion wonders if the cookies expire  [23:00:37] hahaha [23:00:40] lulz [23:00:43] * Maryana gives everyone a cookie and jcmish a whiskey [23:00:47] those are not sticky cookies, i promise [23:00:50] YAY [23:00:54] on the bright side, I'm sure I'll learn some new git tricks trying to work with a repository state which is half-new/half-old (trying to fix new CSS so that it still works with old HTML) [23:01:08] way to stay positive, jgonera! [23:01:18] lol [23:01:45] well, i dont think a mobile varnish cache flush is completely out of the question (especially now that our cache is going to be somewhat wonky anyway due to this deployment) [23:02:01] and if we need to make those footer changes in part for legal reasons... [23:02:06] hm, that would make things easier [23:02:14] it's something we can probably work out with ops [23:02:18] ok [23:02:41] but if we do that, we should be really careful about it and possibly stack up any other potential html changes along with it [23:02:47] (so we don't have to do this again like, next month) [23:02:50] or [23:03:01] [02:47:53] just ban Baidu from our network:P [23:03:21] can't deploy, Gerrit is down [23:03:28] sigh [23:04:09] jgonera, Maryana: that said, if the footer changes are not super urgent, we can go ahead and deploy them and then just let the caches do their thing - it will take up to 31 days for the changes to be ubiquitous, but maybe that's not a big deal [23:05:12] Hello [23:05:45] Quim Gil mentioned that he had asked somebody from wikimedia-mobile to co-mentor my Mobilize wikidata project? [23:06:00] Is there any update on that? [23:07:02] pragunbhutani, I think it might jdlrobson that you're looking for, but he's not here atm [23:07:33] okay, thanks :) [23:07:51] awjr.. hmm, that doesn't sound ideal. if different people start seeing different footer/legal text, that could get confusing [23:10:15] sigh [23:11:20] [Commons-iOS] brion pushed 3 new commits to master: http://git.io/rcH8jQ [23:11:20] Commons-iOS/master ed3eb01 Monte Hurd: NSLog now prints method and line number... [23:11:20] Commons-iOS/master bed00a3 Monte Hurd: Debug toggle now causes images to be refreshed... [23:11:20] Commons-iOS/master 5990235 Brion Vibber: Merge branch 'testModeClearData' [23:13:07] Maryana: the only thing really different between the footers (other than layout) is that about/disclaimers is removed from the new one [23:13:22] but otherwise the same info is there [23:13:31] but it would be weird seeing two different footers for a month [23:14:00] yeah... [23:14:14] i don't think legal would be very happy about that, either [23:14:38] although it would be a great opportunity for an a/b test. see how many people click the links when there are more/less of them :) [23:15:07] hahaha [23:16:40] back in one sec [23:16:47] starting dinner for the boy [23:17:52] how about an a/b test to see what sets off legal? [23:18:47] reverted [23:20:14] ok thanks MaxSem [23:20:30] tfinc: i added a GPS card to backlog, feel free to add ideas or break it down into more actionable pieces - https://trello.com/card/ios-android-some-sort-of-gps-magic/517b115c243fd8d361000faa/17 [23:20:39] k [23:20:51] MaxSem, jcmish, jgonera, Maryana i can confirm things are looking better at least on some of the affected main pages i saw earlier [23:21:00] looking now [23:22:06] looks better now [23:22:46] * MaxSem tries to recall if he has any alcohol left [23:22:48] awjr, i have a question for you about mobile login when you have a sec. still busy w/deployment? [23:22:50] much better [23:23:19] Maryana: sure what's up? [23:23:26] marktraceur, now that lulz are over, we can talk about maps:) [23:23:35] Ah, good [23:23:38] * awjr hands MaxSem some vodka [23:23:44] we have a new clean field in the dbs that logs last time a user logged in [23:23:50] There was a proposal...https://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects#UploadWizard:_OSM_map_embedding [23:23:58] i'm begging and pleading w/platform/ops to add a login method param :) [23:24:12] I was wondering if that could be partly solved with the existing work [23:24:22] 'login method' like… mobile vs desktop? [23:24:27] yep [23:24:36] i'll let pgehres explain [23:24:43] hey awjr [23:24:44] interesting - could we do that though with event logging and correlate? [23:24:46] hi pgehres [23:24:47] long time no IRC [23:24:50] heh [23:24:52] :) [23:25:17] so http://www.mediawiki.org/wiki/Extension:AccountAudit fires on UserLoginComplete [23:25:35] marktraceur, we had monuments API for last year - it searched for monuments around and by some criteria [23:25:45] dyk of a way to distinguish mobile from desktop? [23:25:58] it was completely scary and needs to be rewritten virtually from scratch [23:26:21] pgehres: in what sense? programatically? [23:26:24] pgehres, MobileContext::singleton()->shouldDisplayMobileView() [23:26:35] if so, what max said :) [23:26:43] MaxSem: So...no [23:26:49] mostly from within the hook that I am already using [23:26:51] Noted [23:26:55] yeah that would do it pgehres [23:27:03] cool [23:27:07] w00t [23:27:09] (it would be nice to move that stuff to core someday) [23:27:15] well then, I will go ahead and write up the alter [23:27:17] user_touched_self_mobile here we come! [23:27:24] er [23:27:33] long story :) [23:27:35] i thought this was a family-friendly channel [23:27:39] haha [23:30:45] New patchset: JGonera; "Story 487: Footer update" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61474 [23:31:00] marktraceur, you also asked about maps? [23:32:36] MaxSem: Yeah, did you see the proposal I linked to above? [23:33:06] it's very short [23:33:51] marktraceur, so if you want to use OSM, you're not alone:) [23:34:04] True [23:34:15] MaxSem: Maybe someone from mobile can help mentor? [23:34:42] who's the student? [23:35:00] MaxSem: AFAICT there isn't one yet, but WLM wants to campaign to find one [23:35:58] then it's too early to discuss this - the deadline is approaching fast and the student must decide themselves how it could look [23:36:27] *nod* [23:36:37] MaxSem: We'll see what happens, then [23:36:45] And I'll come back and bug y'all then [23:36:51] sure [23:40:00] Change merged: jenkins-bot; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61707 [23:48:55] doh that should be reverted [23:49:05] New patchset: awjrichards; "Ahem. Revert "Revert "Change remaining #mw-mf-header occurrences to .header""" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61727 [23:52:19] maryana_brb: should we try and reschedule the deployment or are you ok with waiting til next week to get today's changes out? [23:52:20] https://gerrit.wikimedia.org/r/#/c/61705/ needs to be abandoned too? [23:52:47] Change abandoned: awjrichards; "(no reason)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/61705 [23:52:50] yah [23:57:25] awjr: maryana_brb can I take off? [23:57:30] or do I need to stick around? [23:57:38] i think we're all set jcmish, thanks [23:58:10] k thanks [23:58:14] I'll keep my email