[11:23:38] yuvipanda: https://bugzilla.wikimedia.org/show_bug.cgi?id=39039 Is a python script executed by Makefile that take relevant i18n entries from javascript i18n file and push them into value-XX/string.xml a good way to solve this bug ? [11:29:27] Tpt: very [11:31:03] would be awesome to get done :) [11:33:49] yuvipanda: I'll try to do it today. [11:50:55] Tpt: much thanks :) [12:30:46] Oh, MaxSem isn't here. :-( [12:51:22] Jarry1250: around? [12:51:42] saw my comment on #24? [12:52:17] yuvipanda: Nope, have now. [12:53:00] Jarry1250: how do you think that'd work? (Parsing history instead of byline to get authors) [12:53:04] leaving aside unlinked authors [12:53:31] Well there's authors and there's authors... [12:53:38] The byline is a deliberate subset [12:53:52] well, both? [12:54:03] byline parses + a 'more' link that shows history data [12:54:14] I think you'll agree that farming out to history is a horrible cop out [12:54:32] Well, I'm not sure how many people actually care about the history [12:54:42] It's a legal requirement, but if it's not going to get clicks [12:55:19] Jarry1250: the legal requirement is that we credit authors, not that we link to history [12:55:26] well, 'contributors' rather [12:55:46] yuvipanda: Yup, but sometimes authors end up being credited in edit summaries [12:56:03] i'm guessing those would also be credited in the byline :) [12:56:12] *shrug* who knows [12:56:24] i'm thinking of byline being 'X, Y and contributors' [12:56:32] with a tap on the 'and contributors' opening up the rest [12:56:51] yuvipanda: I think we have a few unusual bylines though [12:57:19] you do indeed :) they're hard to parse. [12:57:33] Exactly, it sounds like work [12:57:46] So (full page history) [12:57:56] Would certainly be easier. [12:58:11] We must have a mobile friendly skin [12:58:26] Jarry1250:okay, link to history + license text same as Wikipedia mobilesite/app for v1, and let's see what we can do for v2 or v3 [12:58:58] Yuvipanda: Seems reasonable :) Do show the byline though [12:59:09] (if you are already, apologies!) [12:59:15] (don't have my phone in front of me) [12:59:19] Jarry1250: the full byline? [12:59:41] (I am not, am not even showing authors yet (next to do after about page (which I'm working on now))) [13:00:06] yuvipanda: Yes, although you could visually deemphasise it [13:00:15] Hold on, I'll put my phone back together [13:00:18] One sec [13:00:54] Jarry1250: hold on, I shall give you an updated version [13:01:04] Jarry1250: thinking of making RC1 tonight, and perhaps get a mention on signpost :) [13:01:44] Jarry1250: https://dl.dropbox.com/u/8768784/WPSignpost-debug.apk [13:07:33] yuvipanda: Yes, IB I think. And we should start thinking about a full special story. [13:07:41] *downloads* [13:10:05] Jarry1250: RC for the day, release + special in next signpost? [13:11:31] Yes, sounds good. Have you mentioned this to Ed? [13:12:04] Jarry1250: no, I haven't had any contact with Ed, sadly [13:12:11] would I find him on wikipedia-en? [13:12:13] or do I hit the talk page? [13:14:05] yuvipanda: Probably best to email him about something like this. He's not on IRC AFAIK [13:14:10] I'll PM you the email? [13:14:15] *his email [13:14:43] Jarry1250: sure! [13:30:45] Jarry1250: thanks :) [13:31:01] Jarry1250: No problem [13:31:34] talking to yourself? tch tch [13:32:06] Jarry1250: so, that build you have - about page fixes + attribution fixes = RC [13:32:13] so tell me if there's something going wrong there :) [13:32:47] yuvipanda: Look, I've got about 4 Git merge conflicts open, I'm allowed to talk to myself :P [13:32:53] Sure, will do. [13:33:25] haha [13:33:40] [13:54:24] yuvipanda: Works nicely AFAICT. I would have the byline (not in forced caps) below the subtitle, then an
, same size as the prose, maybe in grey? [14:03:16] Jarry1250: yeah, my thinking too [14:10:10] wgMFFeedbackLinks sets the email address to which feedback is sent right? Why am I not finding it in InitializeSettings.php ? [14:11:37] cause it's in mobile/php [14:11:44] mobile.php [14:23:26] New patchset: Siebrand; "Remove duplicate array keys." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20707 [14:29:03] MaxSem: ? [14:29:20] and http://noc.wikimedia.org/conf/highlight.php?file=Mobile.php has a password :o ? [14:30:09] don't worry, your password is also exposed: http://noc.wikimedia.org/conf/highlight.php?file=SrikanthLogic.php :P [14:30:38] ah! [14:31:04] so my interest is how do I change feedback email ID for my project. Shell request? [14:31:37] you're a committer, you can submit a patchset [14:32:25] MaxSem: which file? I don't see mobile.php on MF code [14:32:35] or its in ops [14:32:42] it's in operations/mediawiki-config [14:34:35] got it, thanks! [14:55:12] Change merged: Siebrand; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20707 [15:20:03] yuvipanda: https://github.com/Tpt/Wikipedia/commit/91b6562654aea9fa1c108794921369bd93263a54 [15:23:50] New patchset: Siebrand; "Remove unreachable code." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20722 [15:31:05] Change merged: Siebrand; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20722 [16:30:37] Jarry1250: https://dl.dropbox.com/u/8768784/Screenshot_2012-08-20-21-58-58%20%281%29.png [16:30:42] screenshot of author name [16:30:46] (had some issues that I just fixed) [16:31:40] yuvipanda: Definitely enough room on that arrangement for your standard byline in full I reckon [16:32:22] Jarry1250: the standard byline is just the name and date [16:32:26] I don't see how date is useful there [16:33:12] yuvipanda: Oh, I didn't mean the date [16:33:21] I just meant to need to parse the author list [16:33:25] *no need [16:33:31] in that case I'll need to parse *out* the date [16:42:30] yuvipanda: Sorry, missed your reply [16:43:08] I meant only for the newer format [16:43:44] No point putting lots of effort into perfecting the bylines on 2008 issues really [16:44:08] Jarry1250: hmm, right. [16:44:13] but good enough for an RC :) [17:06:24] Heya. [17:06:28] MaxSem: Are you around? [17:06:57] I have some questions about the article extract API. [17:07:02] sure [17:07:58] How does the content get broken up into sections and such? [17:08:27] I don't entirely follow the regexen. [17:11:37] basically, the API can output several formats. I chose 'raw' here which uses plain text with section titles prefixed with double \ufffd [17:12:10] Basically, we need to be able to split on top-level sections. [17:13:30] we already have get_top_level_sections() [17:13:44] Sorry, I'm juggling a couple of different conversations simultaneously. [17:14:44] np, we could continue later [17:15:06] No, I'm extracting myself from the others. This one's better. :-) [17:15:36] (One of the others has Kul in it. I need to get away from that guy.) :-P [17:16:57] so we just have to reuse it [17:17:15] Right, back in my own hotel room. [17:17:57] where are you?:P [17:17:59] ajerith: ha ha [17:18:45] greetings all [17:18:50] awjr: yo [17:18:51] Delhi, with Dan and Kul. [17:18:53] tfinc: greetings... [17:18:56] Hey tfinc. [17:18:58] hello [17:19:17] awjr: having run through http://www.mediawiki.org/wiki/Wikimedia_Apps/HowToReleaseAnApp#Android ... what else should i expand on ? [17:19:32] greetings tfinc [17:19:38] tfinc: got my hands full right now can we chat about it alteR? [17:19:39] greetings ajerith & preilly, how goes vuim hacking? [17:19:39] er [17:19:40] later [17:19:43] awjr: sure [17:19:51] vumi* [17:20:16] yuvipanda: i see you've been active on the signpost app dev. do you have a first release candidate ? [17:20:48] tfinc: there are a few more updates to install in production [17:20:54] tfinc: ajerith is making really good progress [17:21:26] tfinc: I will in a few minutes. Final testing [17:21:36] tfinc: no push notifications for 1.0. I want to release it and see how it goes [17:21:39] preilly, tfinc: Reality and my code are slowly converging. :-) [17:21:47] ajerith: we'lll be running a hackathon in india in november. will we see you there? it would be good to thank you in person [17:21:48] and see how my puny tiny little VPS handles the load :) [17:22:00] yuvipanda: why don't you host it on labs ? [17:22:08] s/host/test [17:22:14] tfinc: very good question, actually. [17:22:27] I've a nice little VPS with all the stuff setup (uwsgi, nginx, etc) so it was easy putting it there [17:22:32] . [17:22:41] yuvipanda: you wouldn't want to host a production app against labs but if your just testing then go for it [17:22:45] tfinc: I'd like to come. I even have a visa that'll last that long now. :-) [17:22:50] woot! [17:23:05] tfinc: well, this goes into Google Play in a week at most, so I'd consider it production :D [17:23:08] But I left my phones in Kul's room and have to go back and get them now. :-/ [17:23:17] so I think for now keeping it on my VPS would od [17:23:18] *do [17:23:44] ajerith: good luck with that [17:25:15] ajerith: the recent SMS ban affected you guys much? [17:25:41] * yuvipanda notes that everyone in India is now limited to 5 SMS / day at least until end of month, by a Govt. order [17:25:41] Okay, phones rescued. [17:26:07] yuvipanda: This is the first I've heard of it. [17:26:16] yuvipanda, WTF is going on with your govt? [17:26:19] ajerith: MUMBAI: A government directive to telecom operators, asking them not to allow cell phone users to send more than five SMSs or MMSs at one go over the next 15 days not only failed to check spread of threats to people from the northeast but also ended up creating a communication hurdle for many consumers in Mumbai on Saturday. [17:26:40] It's an operator-operated service, though. [17:26:49] well, it is *supposed* to be 5 SMS in 'one go', but it's implemented as '5 SMS per sim per day' [17:27:10] Oh, right. Stupidly misguided "anti-terrorism" measures. [17:27:28] The US doesn't have a /complete/ monopoly on those. ;-) [17:27:30] MaxSem: the usual. It's either these guys or the religious right to come to power. [17:27:52] not much choice, really :) [17:29:08] So, back to vumi-wikipedia stuff. [17:29:39] Jarry1250: I'm ready to cut an RC :) [17:31:03] damn [17:33:40] tfinc: RC1 cut. https://github.com/yuvipanda/WPSignpost/WPSignpost-1.0RC1.apk/qr_code [17:33:45] making screenshots now, will mail out in a bit [17:33:52] MaxSem: It looks like get_top_level_sections() concatenates all the lower-level section content back together. Is that the case? [17:35:04] it returns a list of top-level sections [17:37:28] Oh, right. I see it now. [17:37:40] so looks like there should be a function returning text of section and all its sub-sections [17:37:52] * ajerith is in a bit of a post-dosa slump. [17:38:45] yuvipanda: woot! [17:39:26] MaxSem: At the moment, the menu lists all sections and subsections in a flat menu. [17:40:43] ajerith, now I remember: I was going to implement it once I've received your review of my previous commits. I didn't realise that it would just quietly stick:P [17:41:05] MaxSem: Ah. Code does that from time to time. [17:41:31] so since they were ok, I guess I could continue with it [17:41:40] Do you have time to work on it today, or should I pick it up in the morning when I'm fresh? [17:41:46] once I'm free from that WLM burden [17:42:18] I'll be able to tell it in an hour - after our sprint kickoff [17:42:22] I need it by tomorrow India time, but I'm happy to do the work myself now that I have a better idea of what's going on. [17:43:22] ajerith, would be great as I'll need some time to recall how that hting works and WTF is that Python thingie:( [17:44:23] but if it's needed urgently I'll try to find time [17:44:58] MaxSem: it's sort of urgent [17:45:13] MaxSem: Okay, don't worry about it then. Probably easier for me to do it in the morning. [17:45:42] preilly, we'll discuss the priorities in 15 mins;) [17:45:46] That was pretty much my plan anyway, I just wanted to pick your brain about it. [17:46:14] I probably have a better idea than you do about what we need, but you know the API better. [17:51:15] MaxSem: So the top-level sections are "2"? [17:52:37] awjr: hangout link please [17:53:52] https://plus.google.com/hangouts/_/e4bcffa32d6b34595b61457deb5c184d4034fb0d?authuser=1&hl=en-US [17:54:07] ajerith, ??? [17:54:26] MaxSem: The section level numbering. [17:54:35] hmmm, so far (1) rejection from filing an issue because bug submitter didn't want to sign up for a github account [17:54:43] There's no level 1 as far as I can see. [17:55:03] ajerith, they can be 1 - it's a convention to generally use second level [17:55:20] awjr: i'm in [17:55:34] MaxSem: Ah. So we can't just hardcode it. [17:56:00] yeah [17:56:46] What's a good way to format section headings in SMS? [17:57:37] hmmm [17:57:48] a separate line? [17:58:10] a separate line *like this*? [17:59:02] How do we differentiate second and third level headings? [17:59:24] Or do we not worry about that? [17:59:48] (Not worrying about it is a valid approach, but I'm not sure if it's the right one.) [17:59:58] awjr: can you send the hangout link? [17:59:59] I don't think we have real ways to highlight them [18:00:03] https://plus.google.com/hangouts/_/e4bcffa32d6b34595b61457deb5c184d4034fb0d?authuser=1&hl=en-US [18:00:06] brion ^ [18:00:09] hmm [18:00:17] "Join Google+ by creating your public profile" [18:00:21] cause == like this == would be too geeky:p [18:00:23] that… doesn't seem right :D [18:00:26] MaxSem: No, we only have plain text. [18:00:51] Also, that might be too character-wastey. [18:00:52] what, no equal signs?:P [18:01:41] it seems to be trying to force me to bvibber@wikimedia.org [18:01:48] but i'm already logged in as brion.vibber@gmail.com, wtf? [18:02:26] brion, yeah, google+ picks a random account of yours [18:02:36] i've had weird issues with multi sign-in and google+ [18:02:43] I have to use it in a different browser due to that [18:02:45] ok lemme see if i can fix this [18:03:07] Most of these things like to use the first account you signed in with. [18:03:14] brion if you both accounts associated with eachother, you should be able to click your picture in the upper right hand corner of the interface and select your other account [18:03:24] ok i'm in [18:03:26] The one with /0/ in the URL. [18:16:37] make that 'number of people who didn't want to report issues on github - 0' [18:16:49] github's signup page seems to hide their free tier a litttle bit (not too much) and he just got confused [18:17:00] Jarry1250: sent out email to mobile-l [18:17:03] shall email ed now [18:25:11] tfinc: email sent. Test it out when you can. You're getting push notifications only in the next release :) [18:26:19] yuvipanda: i'm in our sprint kick off meeting. will look at later. [18:26:24] yuvipanda: no push notifications. slacker. [18:26:26] ;) [18:26:51] yuvipanda: if i open up multiple issues will it happen faster ;) [18:32:43] tfinc: :P [18:32:47] tfinc: you can try! [18:57:13] awjr: there? [18:57:26] jdlrobson: one sec [18:57:28] https://mingle.corp.wikimedia.org/projects/wlm_android_app/cards/133 < does seem to meet acceptance criteria so i'm a little confused [18:59:20] jdlrobson: i wasn't presented with the confirm upload/license option after selecting a photo from my gallery to save for later nor when i took a photo using my camera [18:59:30] jdlrobson: iirc it worked fine the first time i did it, but not the second [18:59:57] jdlrobson: also i did not get the save for later confirmation after selecting a photo [19:00:40] lets try replicate awjr as works for me on latest master [19:00:45] I click use my current location [19:00:47] grab a monument [19:00:56] and click photo in top right [19:01:10] i choose "choose from gallery" [19:01:20] i get the confirm upload page [19:01:47] which has license info on it [19:01:51] i click save for later and get confirmation message [19:02:04] does that not happen for you awjr ? [19:02:42] jdlrobson: it did, although i had initially tested from campaign browsing [19:02:50] mm [19:02:52] let me try that then [19:03:15] ooh [19:03:18] yeh looks like a bug [19:03:32] Error in success callback: Camera6 = TypeError: Cannot read property 'config' of undefined [19:03:58] also, what is supposed to happen after a user selects 'save for later'? right now the user is just presented with the select photo for upload dialog, which seems weird. [19:04:16] oh, it looks like that wasn't defined in the acceptance criteria [19:04:18] they get a confirmation message [19:04:34] it says this has been tucked away in your incomplete uploads available from the welcome screen [19:04:34] jdlrobson: right but after they click ok in the confirmation message [19:04:50] yeh that wasn't defined. I assumed it would go back to the screen before as that's what happens elsewhere in the app [19:04:59] i think it's the right behaviour [19:06:12] maybe it should have a link 'back to monument' [19:06:18] but yeh that needs to be defined [19:06:38] philinje can address it if it needs to be changed, but let's not worry about it for now [19:07:05] jdlrobson: sorry i broke the tests :( [19:07:25] jdlrobson: i'll be sure to check the tests on my changes in the future [19:10:23] awjr: it's ok - you know how to run them right? [19:10:27] awjr: can you pass me the link for the previous iteration storywall ? [19:10:28] it happens [19:10:37] i'm a terrible offender :) [19:11:14] jdlrobson: i believe so, i just hit test.html in a web browser right? [19:11:51] tfinc: under the 'team favorites' section on the right side of the interface, there's a link that says 'previous iteration story wall' [19:12:23] yup that's it awjr [19:13:17] awjr: yup. i see it. thanks [19:13:31] np [20:10:47] Go to sleep -> Think 'Check mail once before going to sleep' -> See a few easy issues reported -> Go 'fuck it, let me go fix these!' -> Lose sleep. [20:10:52] sigh [20:11:08] xD [20:11:55] where are you located? India? [20:12:01] Yes. [20:12:25] And it's nearly 2am here, so I really need to get some sleep as well. [20:13:00] Platonides: indeed [20:13:06] and I've to be up at 6:45 tomorrow [20:13:09] yuvipanda: You still need to come visit me in Delhi before I go. [20:13:23] ajerith: get me a flight ticket and I will :) [20:13:29] * yuvipanda notes he's an unemployed student [20:13:47] But Kul's been dropping subtle hints about a hackathon in November, so maybe I'll come to that. [20:14:23] ajerith: I think that's pretty confirmed, really? tfinc was inviting you to that too [20:14:56] yuvipanda: Some of it depends on how much work I have to do at the time. [20:14:56] yuvipanda: i've got a couple exception backtraces for you on the signpost app. crashes on my kindle fire & galaxy tab 10.1 [20:15:08] filed as issues on github [20:15:10] brion: saw that, I don't think ActionBarSherlock supports the Kindle fire [20:15:20] d'oh [20:15:21] brion: those and tyler's issues are what woke me up [20:15:43] fire shouldn't be hard to fix if it's treated like other android 2.3.x [20:15:50] lemme peek at the source for that [20:16:00] i was wrong, it does support it [20:16:57] we're not planning https://mingle.corp.wikimedia.org/projects/wlm_android_app/cards/172 for this iteration? [20:17:03] My boss seems to have appointed me the India person in our dev team, so I'll probably be back in November. [20:17:19] ajerith: sweeet! [20:18:08] We recently hired a full-time person in India on the business side of things, so with luck I'll be here more often in the future. [20:18:31] ajerith: looking forward to meeting you in Bangalore then! [20:18:39] wonderful city except for everything shutting down at 10:30 pm [20:18:40] :) [20:18:55] :-) [20:19:15] [20:21:01] brion: jdlrobson also, fun look at inline styles + table usage example here: https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2012-06-18/Investigative_report [20:21:22] the one with the key yuvipanda ? [20:21:28] yup [20:21:41] we should have classes for this.. key1, key2, key3…key10 [20:21:49] these should be consistent across wikipedia [20:22:11] that's a nice big table whee [20:22:22] * jdlrobson hates signpost's mobile site [20:22:53] it's also slightly crazy and has mismatching number of I think [20:23:14] jdlrobson: this is the first instance inline styles are causing me trouble in the signpost app. I'll scrub them and see how things go tomorrow. [20:23:24] but I can't actually *fully* scrub them [20:23:28] since they'll lose all style and meaning [20:23:45] now that I think about it, I realize I've no idea what I'm supposed to be doing... [20:24:00] I can't really edit the page - what would I replace those with? key1, key2, etc don't exist [20:24:48] yuvipanda: I will solve tables soon [20:24:50] I hope [20:25:30] we all do, jon :) [20:26:52] awjr_lunch / MaxSem : Nice to see names instead of iso code [20:26:56] *codes [20:27:34] jdlrobson: those colors are actually templates, so shouldn't be too horrible to change [20:32:04] okay, now I *really* should sleep [20:32:06] gnite! [20:42:04] multichill: yeah, i was gald to get those done finally! [20:43:26] Did I see directly that only one language is available? [20:43:52] multichill: what do you mean? [20:44:22] http://wlm.wikimedia.org/api/api.php?action=adminlevels&format=json&uselang=nl returns it in Dutch, but lower levels are just one language? [20:45:36] multichill: if mappings are available in dutch, then you will see them in dutch, otherwise it falls back to the default language configured for that country [20:46:12] Ok, now the field says "translated" [20:46:25] multichill, speaking of default languages - do our defaults make sense? [20:46:27] Probably cleaner to also mention what language [20:46:56] Think so, not sure what it is now since the last time I looked at it was some time ago [20:47:27] http://wlm.wikimedia.org/api/api.php?action=adminlevels&format=json&uselang=lb <- no fallback? [20:48:18] nothing to fall back to: it's country's default language [20:50:31] tfinc/philinje: how many people are currently using the latest beta? should the API maintain backwards compatibility for them? [20:50:46] * tfinc is getting ticked off at the chromebox [20:52:01] MaxSem: Huh? It should output *something* [20:52:05] awjr: are you around for a quick hangout test ? [20:52:34] tfinc: sure [20:52:35] multichill, in an unrelated language? [20:52:45] Always better than nothing at all [20:53:03] awjr: just invited you [20:53:18] i'm not convinced that the boxes speakers work [20:53:46] http://wlm.wikimedia.org/api/api.php?action=adminlevels&format=json&admtree=us&uselang=nl <- This doesn't look very Dutch [20:53:49] awjr: FAIL [20:53:59] tfinc: i feel like i remember seeing a pair of small speakers in r31 [20:54:02] awjr: this box is proving to be more fail then win [20:54:05] dunno if they're always in there or not though [20:54:15] tfinc that is too bad :( just in terms of peripheral hook ups? [20:54:20] awjr: i'd need them to be usb speakers [20:54:26] ... [20:54:35] the thing is. this device has an internal speaker [20:54:43] but it stops working after i plug in the yeti mic [20:54:54] le sigh [20:55:20] i'll have to tinker with it more later today [20:55:49] but i do have all the cabeles now [20:55:55] so its purely software now [20:56:15] MaxSem: a small number of people are using the beta. They will re-install with the next beta, and we can mention the API compatibility in the release notes [20:56:46] philinje, I mean can I break the compat right now?] [20:56:52] guess not [20:57:12] tfinc: That was one of the feedback things I got. Please show a warning or something when people are using an old version [20:58:07] MaxSem: better to do it at end of week [20:58:23] ok [20:58:41] MaxSem: could we use something else for testing changes? [20:58:43] awjr: hey, i will change my schedule this week and next to Mon, Wed, Fri [20:59:17] awjr, not test but deploy new features [20:59:21] philinje: ok - are you still able to meet tomorrow to discuss the upcoming iteration? [20:59:24] MaxSem: roger [20:59:37] awjr: yes [20:59:39] I just don't want to keep two FT indexes, TS has enough [20:59:45] philinje: cool [20:59:51] and I think there needs to be another meeting before the estimation meeting on Wed, for discussion of cards with certain engineers [21:00:19] philinje: can it not be done ad-hoc? [21:00:23] it's hard to squeeze that in with weekly iterations [21:00:38] ad hoc is not working for various reasons [21:01:10] philinje: ok what do you propose? [21:02:05] maybe a half hour with Jon, Brion, you and Max before the estimation meeting [21:02:20] like, immediately before, or sometime on tuesday? [21:02:55] well, Wed would be better for my schedule, so immediately before [21:03:08] philinje: purpose? [21:03:46] awjr, what is the number of the current beta? [21:04:01] to clarify the cards and make sure they are suitable, as well as explaining things that are not so well understood [21:04:07] MaxSem: 2 [21:04:15] thanks [21:04:46] philinje: if we do it 30 minutes before estimation, then there's no time to make any improvements for estimation [21:06:10] awjr: i can do the changes on the spot? [21:07:22] philinje: we've basically been using the weds meeting for that purpose anyway and finishing up estimation on monday. i think the problem is that a) new things get introduced between weds and monday and b) the cards aren't getting cleaned up enough before monday [21:09:22] awjr: there is a small amount of clean-up and a small amount of new cards, so the discussion is about whether something could be done, should be done, or how it will be done [21:10:22] philinje: then perhaps we should make that the focus of the weds meeting, and only do any prioritization on weds if it makes sense [21:10:33] there is very little opportunity for ad hoc discussion, even though some of us are int the same location [21:10:36] awjr: ok [21:10:44] philinje does that work for you? [21:10:48] yes [21:10:53] great [21:11:09] prioritization is quick and easy and generally doesn't require discussion [21:11:20] oh, by prioritization i meant estimation [21:11:30] ok [21:11:32] too many 'aitons [21:11:33] er [21:11:34] ations [21:12:05] awjr: by the way, on iteration 2, there are a few cards I am not sure about in terms of acceptance [21:12:18] like 143 [21:12:53] philinje ok what aren't you sure about? [21:12:54] philinje: You want to get https://gerrit.wikimedia.org/r/#/c/20702/ deployed ASAP [21:13:13] People are testing with new accounts and can't upload.... [21:14:17] multichill: oh, Arthur, can you deploy that? [21:14:44] multichill it's possible that regular user accounts had upload disabled for a reason on testwiki - i do not feel comfortable just deploying it without broader discussion first [21:15:12] It's testwiki, just break it [21:15:52] This is default mediawiki setting on the Wikipedia domain [21:16:29] brion: You around? [21:16:32] sounds ok to me [21:16:53] brion: Brio, we need you [21:17:02] sorry, Brion [21:17:04] just a sec, on a call with tfinc [21:17:24] awjr: is 143 done? [21:17:34] multichill: im asking around about that change [21:17:58] philinje: maybe :p i didn't do a good job with that acceptance criteria [21:18:11] philinje im comfortable calling it done, but there are some issues with it [21:19:15] awjr, it's a default setting [21:19:40] then what does the change do? [21:19:42] everywhere but on commons, you need to be autoconfirmed to upload [21:19:53] upload was not disabled specifically on test [21:20:23] if it's a default setting and it is not enabled, my guess is there's a reason it was disabled so i do not feel comfortable just reenabling it without getting more information first [21:20:53] awjr, see https://bugzilla.wikimedia.org/show_bug.cgi?id=12556 for the reason [21:20:56] awjr: he's saying that there is an accepted exception on Commons to deal with uploads [21:21:01] ok what's up [21:21:23] brion: can you take a look at: https://gerrit.wikimedia.org/r/#/c/20702/ [21:21:29] brion, we want to allow users to upload files on testwiki [21:21:34] without being autoconfirmed [21:21:59] sounds good. i can poke that in gerrit but i can't deploy it currently [21:22:24] whoops, i don''t have +2 access in there it looks like :( [21:22:34] really? who can? [21:22:36] brion if you're comfortable enabling uploads to user accounts on testwiki i can get it deployed [21:22:55] awjr: that's what test is for, testing. :) afaik that's just fine [21:22:59] let's make it so [21:23:00] awjr: MaxSem: brion : Re: [Engineering] [SF] Seating/space change is ripe for one of you to talk about what we've been doing to make tele prescence work [21:23:07] yep [21:23:11] tfinc: already responded [21:23:14] awesome [21:23:31] awjr: too bad i don't see your response [21:23:39] awjr, please review: https://gerrit.wikimedia.org/r/20822 [21:23:42] tfinc: per erik's request i responded directly to him [21:24:23] oh i see the thread has grown wildly [21:24:47] MaxSem: sec [21:25:34] awjr: yup, peopel are passionate about this stuff. i'd love to steer it so that we can share our real world use cases [21:26:08] tfinc for sure, sorry i hadn't seen what the thread had evolved into, i just responded to erik's initial email :p [21:27:05] awjr: no worries [21:28:09] Platonides, multichill, philinje, brion i just merged the change - im looking into when we can get it deployed [21:28:15] brion: http://pastebin.com/yBaSXZBM <- when would this happen [21:28:18] woot [21:28:19] i've seen this on 3 devices now [21:28:21] hmm [21:28:46] cannot read property 'config' of undefined <- suspicious [21:28:48] it seems to be a cordova bug and seems to (sometimes) happen when you load up the app and go straight through choose campaign [21:29:05] never saw it before now [21:29:06] jdlrobson: that's reporting an exception in the camera's success handler [21:29:12] * jdlrobson looks [21:29:13] jdlrobson, brion i recall seeing a similar error when working on adding photo descriptions just before upload [21:29:19] this unfortunately leaves the backtrace pointing at cordova code instead of where your actual error was [21:29:22] which is SUPER annoying [21:29:44] grr that is annoying [21:29:44] brion: happens with gallery selection as well [21:29:52] so likely to be a problem in showPhotoConfirmation ? [21:30:33] Platonides, multichill, philinje looks like no one's deploying right now, i'm going to try and push the change [21:30:37] ahhh [21:30:39] think i've got it brion [21:30:52] or maybe not.. [21:31:28] regardless there is a call to $('#continue-save') which never seems to get added to the dom [21:32:26] ah bloody hell i'm searching in the wrong application [21:32:31] that's why i can't find the camera code :P [21:32:37] * brion stabs eclipse [21:32:39] 492c3cdcdd02e0fe262dd57915fb001550fb9274 [21:32:49] i think that deleted a button ti shouldn't have [21:33:26] var text = formatUploadDescription( curMonument, CAMPAIGNS[ curMonument.country ].config, api.userName ); [21:33:38] if there's no CAMPAIGNS[curMonument.country] you'll get this err [21:34:00] Platonides, multichill, philinje - i just deployed that config change. file uploads should work for regular user accounts on testwiki now [21:34:17] yeah i might have removed the wrong button or something in there, double-check :) [21:35:22] yeh checking now [21:35:24] jdlrobson: does the err happen when doing a photo on items from a particular campaign? [21:35:37] brion: not quite sure yet [21:36:13] but quite possibly brion [21:40:55] Dammit, someone just disabled the uploadwizard at Commons as default way of uploading..... [21:42:47] brion: i'm looking into it [21:45:30] brion: looks like you are right formatUploadDescription( [21:46:55] brion: awjr - so what does it mean if a campaign doesn't have a config? i.e. is not in window.CAMPAIGNS [21:47:02] is there a default campaign we could apply? [21:47:30] maybe no default campaign... [21:47:43] … but if there is no matching campaign we should use cc-by-sa plain license i think [21:48:10] So Belarus for example [21:48:22] i'll cut a bug [21:48:38] jdlrobson: That's wlm-by [21:50:52] awjr. greats [21:50:57] *great [21:51:01] :) [21:52:13] * Platonides notifies wikilovesmonuments-l [21:55:13] jdlrobson: you got code ready to apply to replace my temp commit? [21:55:28] brion: ? [21:55:33] https://github.com/wikimedia/WLMMobile/pull/143 <- [21:56:04] * jdlrobson doesn't understand [21:56:23] perhaps the commit message is just tricky [21:56:36] you mean https://github.com/wikimedia/WLMMobile/commit/c5831fce630a82dd582540dafec8b526336d37a8 ? [21:56:46] i think what happened was you committed a stub and then reverted most of it but not all of it [21:57:36] really that commit https://github.com/jdlrobson/WLMMobile/commit/3aa6171f559ed1b8eb7cedf7d3347d7060147d20 should never have made it in as it wasn't related to your story it was mine ;-) [21:58:11] ah i think i see, yeah i merged those in a hurry friday [21:58:27] the new commit makes sense now :D [21:58:34] want me to change the commit message? [21:58:39] add the context [21:58:40] nah [21:58:51] also brion fix on the way for the camera callback issue [21:59:10] excellent [21:59:18] [WLMMobile] brion pushed 2 new commits to master: http://git.io/XJHR4Q [21:59:18] [WLMMobile/master] Revert "Primitive 'save for later' (can't yet upload, but can view in uploads list)" - jrobson [21:59:18] [WLMMobile/master] Merge pull request #143 from jdlrobson/cleanup-revert - Brion Vibber [21:59:35] Project WLMMobile - Nightly builds build #68: SUCCESS in 13 sec: https://integration.mediawiki.org/ci/job/WLMMobile%20-%20Nightly%20builds/68/ [21:59:36] jrobson: Revert "Primitive 'save for later' (can't yet upload, but can view in uploads list)" [22:03:54] brion, awjr, jdlrobson: backend part of search for address will be complete once my last commit will be merged. on the app side you need to replace srname with srquery and remove the initial ~ [22:04:10] \o/ [22:04:53] MaxSem: awesome! so now we just need to wrangle someone into implementing that in the app? [22:05:15] yup [22:08:40] brion, jdlrobson: will one of you have some bandwidth soon to help implement the changes MaxSem is talking about? [22:08:58] MaxSem: i can take care of that - just poke me when needed [22:09:13] excellent [22:09:38] w00t [22:19:22] awjr: noticing Canada still not human readable at second level [22:19:36] * awjr takes a look [22:19:51] also there are still square brackets at bottom level of US [22:21:13] that is a known issue, i thought that was going to get cleaned up at the app-level - i know it got discussed in irc at least once but i dont know if that ever got turned into a story card [22:21:25] (the wikitext in admin level names, i mean) [22:21:32] ok, will add a card [22:22:30] here is another issue: when drilling down in Austria, notice what happens at the top in the parentheses: CHOOSE CAMPAIGN (at-1) [22:23:26] yeah - let's create a card for that as well [22:23:33] philinje: https://github.com/wikimedia/WLMMobile/pull/142 [22:23:37] awjr: ^ no need [22:23:39] i fixed it [22:23:42] \o/ [22:24:03] great [22:24:30] awjr: is the problem in Canada just a bug? [22:24:53] philinje: im investigating [22:26:18] must be a bug, we have the data (at least in english) [22:26:30] philinje: did you notice any other country with the same problem? [22:28:08] not yet [22:29:54] Denmark has some bad data at the top of the monument list [22:31:25] no addresses for Indian monuments [22:33:08] awjr: Malta has same problem [22:34:51] philinje: re the admin levels issues, i'll see if i can figure out what the actual problem is [22:35:34] ok, added a card [22:35:47] i wonder if we should flag data issues to someone? [22:36:31] pihlinje whenever i've had monument data questions i've brought them up with maarten and he's either dealt with it himself or gotten me in touch with someone else who can [22:37:23] ok. UK has same problem [22:37:32] should I create a card? [22:38:07] hmm it looks like UK is not using iso-3166-2 codes for the subdivisions [22:38:22] so that is a data consistency problem [22:39:15] philinje go ahead and create a defect card for the canada issue [22:39:42] philinje as for malta, the iso-3166-2 data set im looking at doesn't have any data for malta, at least not with the 'MT' country code [22:40:32] ok, I'll write an email to Maarten [22:41:34] philinje: as for malta, the problem is likely with the data set i used [22:41:36] or maybe you should? I can ask about the bad data [22:42:43] philinje: i should deal with the canada and malta problems as they appear to be related to issues in the API or with the iso-3166-2 data set i used. [22:43:02] ok [22:43:16] philinje you should deal with UK and India issues if you can, as those appear to be WLM data issues [22:44:39] is UK the same problem as Malta? [22:44:57] philinje no [22:45:12] philinje the UK problem is that they are not using iso-3166-2 codes in the adm1 field [22:45:30] so bad data [22:45:34] yep [22:45:44] got it [22:45:48] unless they really want the zone to be labeled 'sct' :p [22:46:29] awjr: heads up about brions availability change next weet [22:46:41] we'll need to let him know what he has to be around for [22:47:12] tfinc thanks for the heads up. it's tough timing since we'll be prepping the app for launch that week [22:47:56] awjr: brion said he can be around but he'll need a heads up about *when* [22:48:00] i let you guys know as soon as i did [22:48:43] ok cool [22:54:46] [WLMMobile] brion pushed 3 new commits to master: http://git.io/vu_jHg [22:54:46] [WLMMobile/master] throw away countries which are not described in campaigns-data.js - jrobson [22:54:46] [WLMMobile/master] run python scripts/generate_campaigns.py - jrobson [22:54:46] [WLMMobile/master] Merge pull request #144 from jdlrobson/broken-upload-process - Brion Vibber [22:54:55] Project WLMMobile - Nightly builds build #69: SUCCESS in 8.6 sec: https://integration.mediawiki.org/ci/job/WLMMobile%20-%20Nightly%20builds/69/ [22:54:56] * jrobson: throw away countries which are not described in campaigns-data.js [22:54:56] * jrobson: run python scripts/generate_campaigns.py [23:00:29] brion: we got the galaxy s3 in today. it's screen is amazing. 1280x720 [23:00:35] whee [23:00:57] yuvipanda: THERE? [23:01:19] yuvipanda: did you take a look at generating css for https://bugzilla.wikimedia.org/show_bug.cgi?id=39288 ? [23:02:14] brion can you explain https://bugzilla.wikimedia.org/show_bug.cgi?id=39251 [23:02:31] i got stuck in some kind of loop [23:02:36] not sure if i can reproduce it now [23:03:03] awjr: is someone looking into https://bugzilla.wikimedia.org/show_bug.cgi?id=39244 ? [23:03:14] brion: if you can't can you close as WONTFIX - i've never seen that before.. :( [23:03:34] jdlrobson: yup, the square brackets are wikitext [23:03:36] also brion is this still relevant - https://bugzilla.wikimedia.org/show_bug.cgi?id=38290 [23:03:38] phil added a card to cope with it [23:03:43] awjr: great [23:03:52] awjr: what's the link to the card: [23:04:01] ah 176 [23:04:02] hmm i'll check [23:04:03] no worries [23:04:29] https://mingle.corp.wikimedia.org/projects/wlm_android_app/cards/176 [23:04:31] jdlrobson: ^ [23:04:55] awjr: philinje: when are we switching the app to point to commons instead? thats a pretty scary change for us [23:05:01] we may even want to do a one off build soon [23:05:10] to see if anything doesn't work out of the box [23:06:34] good question, we have lots of data issues that could cause problematic uploads, so I suspect next week is the earliest [23:07:28] it would be a good idea to try basic functionality against commons this week so we can unearth anything large and scary before we're in the middle of next ewek [23:08:32] we might even consider having the beta we ship on friday use commons so we can get more real-world testing over the weekend to find any serious issues [23:11:41] jdlrobson: looks relevant still. certainly new photos are being saved into that cache space on my galaxy nexus [23:12:33] k brion [23:12:45] also brion i'm not getting key press events on https://bugzilla.wikimedia.org/show_bug.cgi?id=38896 anymore :( [23:13:03] jdlrobson: tried 'input' event? [23:13:08] yup [23:13:12] nothing seems to be firing :( [23:14:04] maybe z-index issue [23:14:06] lemme see what i get [23:14:16] brb [23:17:02] jdlrobson: on #filter-monuments, I see hits on the 'input' event [23:17:09] however there's an error that happens when searching [23:17:23] 08-20 16:16:26.720: D/CordovaLog(26215): file:///android_asset/www/js/monuments.js: Line 82 : Uncaught TypeError: Object [object DOMWindow] has no method 'trimBoundingBox' [23:19:21] mm [23:19:25] #filter-campaigns seems to work fine with bind('input', ... [23:19:37] i'm not getting anything on my phone.. [23:20:37] it never enters the callback.. [23:20:38] bizarre [23:21:06] weird. definitely works for me... [23:21:25] galaxy nexus (verizon), 4.0.4 [23:21:44] mm [23:22:08] let me uninstall and try again [23:22:50] D/CordovaWebView(26631): KeyDown has been triggered on the view [23:22:51] D/CordovaWebView(26631): KeyUp has been triggered on the view [23:22:56] but no event handler rrgh [23:23:41] weeeeeeird man [23:23:44] yup [23:24:09] publish the branch and i'll try it, just in case there's something different in there somehow [23:29:01] ahhhh [23:29:38] i'm beign stupid [23:29:49] confused #filter-campaign with #filter-monuments [23:29:59] aha [23:30:02] \o/ [23:37:14] awjr, I've got a scary thought: what if some monuments belong just to an admin level while others belong to its children? [23:38:37] MaxSem: i think i asked that same question to maarten once [23:38:57] and i *think* he told me it was ok [23:39:13] i'd have to dig through logs/email… although it might've happened on the phone too [23:39:24] but with our drilldown procedure user will never see them [23:39:39] heads up to brion awjr and philinje > http://pastebin.com/hjTWpQTs < certain campaigns are not being returned by http://commons.wikimedia.org/wiki/Commons:Monuments_database/Campaign_names?action=raw [23:39:44] correct, but i believe that it should not happen [23:40:22] so currently Belgium campaign for example has disappeared from app [23:40:30] hmm [23:40:50] be-bru, be-vlg, be-wal but not be [23:40:50] jdlrobson let me add it to my stack of things-to-investicate [23:40:52] er [23:40:54] investigate [23:41:07] jdlrobson what do you mean 'be-bru, be-vlg, be-wal but not be' [23:41:20] so previously we had regions of belgium as 'countries' [23:41:40] * awjr is confused [23:41:43] the app currently assumes when uploading a photo that there are rules for how to upload photos - e.g. license information, template for the top level country [23:41:59] country != campaign [23:42:11] so when uploading a photo anywhere in Belgium it is expecting an entry for "be" in campaigns-data.js [23:42:28] that is not good - i do not think there is a 'be' campaign [23:42:48] yeah, be, de, dk, and it are all split [23:43:07] so i think https://github.com/wikimedia/WLMMobile/pull/144 needs more work [23:43:10] it needs to be looking up the campaign, not assuming the top-level country is a campaign [23:43:39] i believe campaign info is associated with the monument record [23:43:40] ok i'll revert and revisit https://github.com/wikimedia/WLMMobile/pull/144 [23:44:15] jdlrobson: do you need me to investigate anything, or did that answer the question? [23:44:29] the problem is basically some monuments are associated with campaigns that do not exist in campaigns-data [23:44:40] i'm trying to prevent the user ever seeing a monument in this situation [23:44:52] oh that sounds like a different problem that countries vs campaigns [23:44:56] [WLMMobile] brion pushed 2 new commits to master: http://git.io/DCKa3w [23:44:56] [WLMMobile/master] fix search for android 4.0.4 (bug 38896) - jrobson [23:44:56] [WLMMobile/master] Merge pull request #145 from jdlrobson/38896 - Brion Vibber [23:45:01] jdlrobson: can you give me an example? [23:45:03] Project WLMMobile - Nightly builds build #70: SUCCESS in 7.1 sec: https://integration.mediawiki.org/ci/job/WLMMobile%20-%20Nightly%20builds/70/ [23:45:03] want me to discuss on skype? [23:45:03] jrobson: fix search for android 4.0.4 (bug 38896) [23:45:09] jdlrobson: yes [23:45:12] jdlrobson: give me a minute [23:45:40] preilly: let me know if you need a device to test the new open path build [23:46:38] jdlrobson: ok ready [23:47:12] philinje, do we need https://mingle.corp.wikimedia.org/projects/wlm_android_app/cards/172 in this iteration? [23:47:36] hang on awjr [23:47:45] jdlrobson: no prob, ready when you are [23:47:55] MaxSem: no [23:52:18] New patchset: preilly; "fix m zero vary issue with cache" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20841 [23:52:47] tfinc: I was using the c3 [23:53:05] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/20841 [23:53:06] preilly: k [23:53:23] awjr: MaxSem: how often do we sync with ts for wlm data? [23:53:36] tfinc, daily [23:53:54] the same frequency it gets updated [23:54:02] MaxSem: and how do we sync? feel free to send me to a wiki page if its documented [23:54:32] MaxSem: 172 was moved to the backlog, low priority compared to other stuff [23:55:13] tfinc, currently, ops haven't merged the cron for it, so the script gets ran manually [23:55:26] what details do you want to know? [23:56:20] MaxSem: just how it works [23:56:26] i've gotten questions about it [23:56:30] and haven't been able to answer them [23:57:19] toolserver geneartes a tab-separated dump, we download it and import with mysqlimport [23:58:28] tfinc, you can direct them to me