[00:01:45] I noticed lots of photos uploaded via mobile web have no exif data, but do have a JPEG file comment of "*". Anyone have any idea what that's about, or what software is doing that? (Its not important, just curious) [00:03:04] Change merged: Yurik; [mediawiki/extensions/ZeroRatedMobileAccess] (master) - https://gerrit.wikimedia.org/r/70244 [00:48:29] awjr, still around? [00:48:49] d'oh, guess not >_> [16:26:43] MaxSem, i'm trying to figure out when which of the hooks in mobile fires and why do we need so many in zero [16:27:04] o_0 [16:27:51] there are 4: BeforePageDisplayMobile, GetMobileUrl, MinervaPreRender, GetMobileNotice [16:28:18] do you know anything about them? [16:28:47] dr0ptp4kt, ^? [16:30:05] MaxSem? [16:30:22] all look sane [16:30:36] what exactly do you have doubts about? [16:34:16] yurik, MaxSem: BeforePageDisplayMobile => for changing article HTML, GetMobileUrl => for defining FQDN stuff for MobileFrontend, MinvervaPreRender => for latest possible execution context on banners and pages, GetMobileNotice => for banner rewriting where we're fairly confident Central Notice won't be running (although not 100% confident, it's just that the $wg flag for Central Notice isn't checked). [16:34:27] MaxSem, i am not sure why zero needs to handle all of them - because it seems we know have a bit of a spagety :) [16:35:17] okay, which of them do you think we can ditch?:P [16:35:35] dr0ptp4kt, awesome, but do we need all 4??? [16:36:13] yurik, wfRunHooks( 'DoAllStuff' )?: [16:36:16] ) [16:36:29] I would have thought that we need to handle two cases: 1) a DOM object to simplfy all link rewrties, and 2) insert banner. Both could actually be done at once [16:36:52] MaxSem, you were thinking of exposing a DOM hook a at some point [16:37:18] currently btw the rewrite breaks HTML a bit [16:37:36] yep - will do once I finish my current refactorings [16:37:57] you'll still need GetMobileUrl [16:39:54] yurik, MaxSem: we can *probably* technically convert all of them to one hook, as long as that hook is called at the very last minute before rendering. but, the hook caller should pass in sufficient context to allow the hook implementation to trump stuff like banners. (as awjr says, globals are the devil) agreed with MaxSem that it would be easiest to retain GetMobileUrl (yurik, set a breakpoint on its implementation on it to see the call stac [16:39:54] it's not too horrible to figure out). you'll want to check with jdlrobson about any other side effects or abstraction or encapsulation no-nos. [16:41:15] yurik, MaxSem: remember when BeforePageDisplayMobile handled everything not in GetMobileUrl? ;) [16:41:47] yes. it was ewwww [16:53:41] MaxSem, dr0ptp4kt, what's really "eww" is the fact that currently both onMinervaPreRender and onGetMobileNotice call all 3 different banner renderers: renderUnknownCarrier, renderWarning, renderBanner [16:55:05] dr0ptp4kt, why do we need both of them again? I am not sure i understood the diff with sitenotice thing [16:56:00] [WikipediaMobile] jdlrobson pushed 2 new commits to master: http://git.io/dxFnRg [16:56:00] WikipediaMobile/master fa9b90d Hay Kranen: Added question about issues... [16:56:00] WikipediaMobile/master fb04734 Jon Robson: Merge pull request #324 from hay/patch-1... [16:59:15] yurik: MaxSem asked in Gerrit to use GetMobileNotice for the WAP context instead of onMinervaPreRender. jdlrobson asked to use onMinervaPreRender for other banners. it is "cheating" for knowledge of the caller, but the context in which GetMobileNotice is called is such that we are reasonably assured for WAP contexts that Central Notice won't be used, but for mobile web (non-WAP) contexts we also know that Central Notice calls will be invoked [16:59:16] later, so we need to override that with onMinervaPreRender. for onMinervaPreRender, on the other hand, we know the contexts from which that is called are bound late enough to assure us that we can override Central Notice banners. [17:00:04] yurik, later this afternoon i can show you a debug session, which will help, if you like. [17:00:41] sure, but this code really needs to be refactored to be just one function, even if it gets called from two hooks, not this mess [17:01:05] we can't have two places decideding which banner to show [17:08:57] Maryana: did you ping me after i left yesterday [17:08:58] ? [17:09:15] yeah, i was just gonna say that i created a 2013/2014 backlog [17:09:23] and started populating it w/stories from the etherpad [17:09:33] :D thanks for doing that [17:09:37] wanted to make sure i didn't blow up mingle in the process [17:09:42] i'll take a look [17:09:49] we should move stories from the q4 backlog too [17:09:53] yep [17:09:54] (if we want to keep any) [17:10:04] so, another question - is there a way to archive stories? [17:10:12] deleting feels antithetical to our culture :) [17:10:13] what do you mean? [17:10:31] well, it looks like there's only a delete button which totally wipes the story out of existence [17:10:38] oh i see what you mean [17:10:46] it would be nice if we could send it somewhere hidden away but be able to retrieve it [17:10:58] well, we can set things up however we like - there's no official concept of 'archive' but we could move stories to an 'archive' backlog or whatever [17:11:13] or just leave them in the existing backlog that we're not using anymore [17:11:26] they'd still be discoverable when we want, but not when we dont [17:11:49] yeah, the only downside is that we'd start accumulating release-related backlogs [17:11:52] there's already a couple [17:11:57] but maybe that's not so bad [17:12:25] the alternative would be to just create an 'archive' iteration not associated with any release or anything, and just move old cards to that [17:12:45] that kinda sounds like a better way to go for me, just so everything old could be in one place [17:12:59] either way is fine by me - the filters are pretty powerful in mingle so it would be easy to dig up old 'archived' cards either way, i think [17:13:02] ok let's do it [17:13:04] sweet [17:13:15] feel free to set it up if you like otehrwise i can do it later [17:13:28] if i have a min sometime today, i can do it :) [17:14:35] hmm it's supposed to be 45C/113F here tomorrow [17:14:58] im sure that's still not hot enough to impress YuviPanda, but i am not looking forward to it. [17:15:11] depends on with or without AC, awjr :) [17:15:20] it also rained here today, so the next few days aren't going to be that bad :) [17:15:35] ohhh man, i can't wait for the end-of-the-summer rains to start here; best time of year [17:15:45] it's only 84F right now! Pretty cool! [17:16:00] with only ~80% humidity! :D [17:16:05] except when im walking the dog, i'll be in the a/c, so i've got it easy. it just makes me not want to go outside. at all [17:16:11] hehe :) [17:16:18] hahaha humidity here is 14% [17:16:22] right now at elast [17:16:26] probably tomorrow too, maybe less [17:16:29] :) [17:17:22] is the rest of the team out because they were all partying in the castro last night? [17:17:42] that is highly possible, though jdlrobson did exhibit activity on Facebook 2 hours ago [17:17:47] that might also be him just going to sleep, however. [17:17:52] lol [17:17:59] I also get the feeling that I might've just missed the biggest party of the decade in SF... [17:18:33] YuviPanda: there are always biggest parties of the decade in SF :) [17:18:42] hopefully :) [17:19:11] * YuviPanda reads docs + writes PHP code on one monitor, reads drama on -commons on other... [17:23:20] dr0ptp4kt, ping [17:24:49] rfarrand: my friend is heading to the lodge this weekend so i can tag along w him [17:24:52] yurik, pong. about to drop while i change floors. holler at me if i don't get back on irc for some reason. [17:25:11] mhurd: good deal :) [17:25:16] thanks for letting me know! [17:25:41] rfarrand: np thanks for the offer - maybe next time [17:25:43] MaxSem, when do you think we could have the DOM in the hook? anything i could help with? [17:26:41] rfarrand: my climbing buddy only has to work evenings this weekend. yay [17:26:44] patience....:) [17:29:34] mhurd: awesome, what are you guys going to get after? [17:29:54] New patchset: MaxSem; "Create a thin base class for all our skins" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/69336 [17:33:21] rfarrand: not sure - anything on real rock will make me happy - it's been a few months! [17:35:29] dr0ptp4kt, so, what do you think about my point above? [17:35:59] yurik, didn't see message, will get back to you. please resend. on a video call with ottomata atm. [17:39:40] dr0ptp4kt, from above -- currently both onMinervaPreRender and onGetMobileNotice call all 3 different banner renderers: renderUnknownCarrier, renderWarning, renderBanner [17:43:16] dr0ptp4kt, why "jdlrobson asked to use onMinervaPreRender for other banners"? [17:47:39] speaking of which, where are jdlrobson and jgonera? [17:49:00] sorry awjr didn't realize i wasn't logged in to irc - open but not connected :) [17:49:07] ha no worries jdlrobson [18:04:29] awjr: i fixed up https://gerrit.wikimedia.org/r/#/c/70323/ with yours and MaxSem 's advice - this ready to go now? [18:04:43] will look shortly jdlrobson [18:08:30] LGTM [18:08:52] if there are no ojections from awjr, can be merged [18:10:41] New patchset: Jdlrobson; "Story 867: Update assets to reflect new article action bar (2/2)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70571 [18:29:15] New patchset: Jdlrobson; "Story 867: Update assets to reflect new article action bar (2/2)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70571 [18:29:34] New patchset: Jdlrobson; "Story 867: Update assets to reflect new article action bar (2/2)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70571 [18:38:38] New patchset: awjrichards; "Sandbox alpha and beta skins from stable" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70323 [18:39:02] jdlrobson, MaxSem i made a few minor changes - if they're ok, i think it's fine to merge ^ [18:39:15] looking [18:41:18] Change merged: jenkins-bot; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70323 [18:42:15] New review: MaxSem; "recheck" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70320 [18:42:47] awjr, can you review ^^^ please? [18:42:51] sure [18:42:54] looking now [18:44:05] MaxSem: so if someone requests mobileformat=wml they'll just get regular HTML + a warning? [18:44:09] or just the warning? [18:44:13] (i cant remember how this works in the api) [18:44:23] the former [18:44:46] isn't it weird to serve full HTML to someone requesting WML? [18:44:55] would it be better to just return an error to them? [18:46:36] there's not much difference actually because our WML is not real - just HTML5 with a few tags removed [18:46:52] so it's not going to cause a wap phone to explode or something? [18:46:59] or someone's data bill to exploe? [18:47:08] no wap phone would request this;) [18:47:23] that's why I'm removing it in the first place [18:47:25] oh right, it's the api req [18:47:27] yeah [18:48:17] MaxSem: that patchset causes api tests to fail [18:48:21] gotta update the test case [18:48:32] owchie [18:48:36] New review: awjrichards; "please update tests so they don't fail :)" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/70320 [18:49:20] oh dear [18:49:22] my hd is full [18:51:28] New patchset: MaxSem; "Remove support for api.php?action=parse&mobileformat=wml" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70320 [18:56:14] awjr, ^ [18:57:18] gah my full hd is causing me problems [19:01:29] Change merged: jenkins-bot; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70320 [19:51:15] MaxSem: did you ever hear back from antoine about https://gerrit.wikimedia.org/r/#/c/69585/? [19:52:59] yes, he has no idea [19:53:25] :-| [19:53:38] did he offer any suggestions about what to do? [19:53:59] you're not on tech talk? [19:54:05] tech talk? [19:54:21] oh crap i totally forgot about that [19:54:45] joining now :p [20:02:44] you didn't miss anything that makes sense awjr:P [20:02:58] ok, that was kinda the impression i've been getting :) [20:04:06] MaxSem: is that 'nothing he said makes any sense' or 'we just started'? [20:04:25] basically, he's pushing a total BS that has no practical use [20:06:45] hahaha [20:07:13] MaxSem: snake oil or 'I just read this from books'? [20:07:20] MaxSem: link? I need to join now :P [20:07:21] * YuviPanda looks [20:07:28] https://plus.google.com/hangouts/_/31cbae4c71809a05f02b71b302aca10eac3f87c3 [20:07:40] YuviPanda, https://xkcd.com/669/ [20:08:28] haha [20:08:35] am I too late to join the fun? [20:08:43] oh, this is recorded [20:17:10] * YuviPanda waves to brion [20:17:32] heyyyy [20:17:43] brion: any icon updates? [20:18:05] YuviPanda: so we'll stick with our beta one for now, and decide for sure what to do next in an iteration or so :) [20:18:13] brion: heh, okay [20:18:18] brion: I'll do the release on Monday then [20:18:27] sweet [20:18:35] brion: see if you can get to the Volley cache issue before that? [20:18:41] i'll peek at it [20:18:41] * YuviPanda doesn't want to context switch right now [20:18:43] okay [20:18:54] brion: it's a one line Change, in CommonsApplication.java, but for some reason doesn't work... [20:19:05] :( [20:24:13] New patchset: Jdlrobson; "Story 867: Update assets to reflect new article action bar (2/2)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70571 [20:26:20] MaxSem: I find whatever is going on now rather very informative, however :) [20:26:45] yep, Adam's talk is quite useful:) [20:27:24] also vim ftw :) [20:31:33] * brion stabs java [20:31:40] I have to cast the value of foo.clone() back to the right type? [20:31:41] ffs [20:32:58] brion: yeah, that is retarted, but not as retarded as checked exceptions [20:33:03] brion: also where are you using clone? [20:33:08] * YuviPanda has an inherent distrust of clone [20:33:20] maybe i should use an explicit ArrayList<> method instead [20:33:43] I'm abstracting the media metadata (description, categories, templates) into an object [20:33:53] and returning a fresh copy of an ArrayList<> from a getter method [20:34:01] so it's not mutable by mistake [20:35:19] could use categories.subList(0, subList.length) or something [20:36:11] why .size() here but .length() on strings? sighhh [20:36:39] lol [20:36:46] i still have to cast it back to ArrayList [20:36:54] since .subList returns List<> [20:37:13] brion: hmm, just from listening to that it sounds like a case of Javaitis... [20:37:21] not sure :P will review later but :) [20:37:23] hehehe [20:38:33] or i could just return an array [20:38:38] meh [20:43:25] brion: casts are bad, yes. [20:43:52] there ought be a way to specify 'return self type' on those methods [20:44:08] brion: Java has some concept of generics, but... they're only 'sortof' generics [20:44:15] yeah [20:44:46] brion: YuviPanda mhurd: reminder, Story prioritization meeting @ 2 [20:45:16] tfinc: okay, let me create board12 and invite everyone. [20:45:16] +1 [20:45:18] YuviPanda: todays is set but for future ones is 2PM here too late for you ? [20:45:35] tfinc: would be nice to move it up, yeah. [20:46:05] tfinc: I've been advised to try switching to 'normal' sleep patterns for health reasons, but you can see how well I'm following those :) [20:46:17] but yeah, in general - keeping everything before noon PST would be nice [20:47:29] YuviPanda: i support you in your normal sleeping habits that make you healthier [20:47:57] Problem is my productivity is highest at dark, and I clearly need to be productive to stay healthy... [20:48:03] :) [20:48:32] brion: tfinc Sprint 12 card created, everyone copied on it :) [20:49:25] whee [20:52:25] YuviPanda: thanks [20:52:40] np, lmk if I forgot to add someone [20:54:51] ooh, fancy. i'm finally starting to run some pageview numbers for the quarterly review. looks like there was a big spike in mobile pageviews after october 2012… aka, when we pushed out the new nav :) [20:54:54] dr0ptp4kt: nice talk :) [20:57:24] YuviPanda, thx, sorry if it was a bit dry. not feeling the hottest :( [20:57:49] dr0ptp4kt: I tuned in just as your talk was starting, and was very fun. Not what I'd call dry [20:59:03] YuviPanda, cool man. shouldn't you be sleeping or something? :) [20:59:12] dr0ptp4kt: what, it's only 2:30 AM! [20:59:22] dr0ptp4kt: my plan to 'beat the heat' is to stay up till 8AM and then sleep till 4PM :) [21:00:08] YuviPanda, smart plan. i remember the heat being a bit stifling when i was in mumbai and chennai a few years ago! [21:00:20] oh, you were in Chennai? [21:00:27] heh, true. Plus this roof is asbestos [21:11:19] YuviPanda, nah, born and raised in minnesota (north central us), but got to visit india for 15 days. the heat was pretty intense even in december, i can only imagine now. does the asbestos create a heat sink? [21:11:58] dr0ptp4kt: yeah, outside it'll be ~110F, inside feels like 150-160F, I think [21:12:13] dr0ptp4kt: consistently similar to how it feels to be right next to a Bus engine that's been running all day [21:12:34] YuviPanda, yikes. get here soon :) [21:12:42] heh, yeah, I've been filling forms now :) [21:12:50] dr0ptp4kt: feel better you too! :) [21:43:38] MaxSem: perhaps a mail to wikitech-l is in order RE https://gerrit.wikimedia.org/r/#/c/69585/ [21:53:32] YuviPanda: i added one more member to spring 12. bugello [21:53:37] YuviPanda: are you familiar with bugello ? [21:53:49] of course. I helped a tiny little bit set it up on toollabs :) [21:54:23] \o/ [21:54:33] bingle should also go to tool labs, at some point [21:54:43] whenever awjr has the time, that is [21:54:56] YuviPanda: icon update email, let me know what you think of may's latest revision based on my feedback [21:55:02] brion: looking! [21:55:08] tfinc: oh, we should add a card for 'respond to WLM' [21:55:14] YuviPanda: yeah, i am setting up bugello for the partner team and will probably move bingle over when i do that - likely tomorrow :) [21:55:16] tfinc: I saw that there's a thread on the WLM mailing lists already... [21:55:23] awjr: \o/ [21:55:24] YuviPanda: and i responded to it [21:55:42] tfinc: yeah, saw that too :) Good you responded, it was going towards the pitchfork store :) [21:55:50] tfinc: should we add a card for it? [21:55:53] no [21:55:55] awjr: toollabs <3 :) [21:55:58] tfinc: okay! [21:55:59] YuviPanda: i have it starred to respond to [21:56:07] but our quarterly review takes priority [21:56:29] :) okay [21:59:37] brion: how does it look on a screen? [21:59:55] brion: also, I'm still a bit unhappy that the public email thread got no response whatsoever. [22:00:01] YuviPanda: good on the demo screenshot jared showed me yesterday [22:00:05] yeah, we gotta be diplomatic and shit [22:00:15] move them gently towards open communication :) [22:00:21] email is a good medium for them though [22:00:36] jared in particular is in a lot of mediums so always-on irc isn't great to get hold of him [22:00:40] *meetings [22:00:48] that's fine, but this one isn't on any list, for example. [22:01:06] baby steps :) [22:01:27] :) [22:02:36] brion: do keep nudging them towards public channels of communication :) [22:03:13] i shall try :D [22:03:37] :) [22:03:44] brion: thoughts on a colored version of that? [22:04:03] i'm kinda digging the black version [22:04:12] if people really want color, give feedback to that effect :) [22:04:22] hmm, I'll look at it on a phone screen later [22:04:27] and then figure out :) [22:04:50] but yeah, a lot better than the previous one [22:07:47] tfinc: is the graph on slide 19 accurate? I don't know if mobile apps are that flat... [22:08:12] tfinc: http://stats.wikimedia.org/wikimedia/squids/SquidReportUserAgents.htm says otherwise [22:08:13] FrustratedPanda: ask Maryana, she has that section [22:08:28] FrustratedPanda: it looks low to me [22:08:28] Maryana: ^ [22:08:55] tfinc: ~5% mobile apps, not 0 [22:09:02] that's surely a charting issue, methinks [22:09:14] i wonder if she didn't pull the data yet [22:09:20] good catch though [22:09:45] :) [22:10:21] tfinc: Maryana I also suggest adding a deletion % in the metrics, since they do make a huge difference. [22:10:43] Android has about 2x the number of retained uploads the Web has, despite the fact that the web has 4-8x more uploads in total [22:10:50] (not even counting iOS) [22:11:04] will look at this over the weekend, I think :) [22:18:21] charge & lunch…. bbiab [22:33:10] yuvipanda: i was going off the raw csv data in http://reportcard.wmflabs.org/ [22:34:07] and this: http://stats.wikimedia.org/wikimedia/squids/SquidReportUserAgents.htm [22:34:54] Maryana: that slide was showing the share of apps' traffic to be 0 :) [22:35:23] squids says 11.94% web vs 4.75% apps [22:35:26] it's not 0, but it's less than 1% [22:35:37] of total traffic (desktop + mobile)? [22:35:41] where are you getting that number? [22:35:42] Maryana: yes [22:35:45] Maryana: http://stats.wikimedia.org/wikimedia/squids/SquidReportUserAgents.htm [22:35:49] you just gave me that link :) [22:35:51] we're looking at the same page [22:35:57] i was only counting wikimedia apps [22:36:13] sure, even that is 2.30% + 1.44% [22:36:29] that's what, 3.74% [22:36:31] not under 1% [22:37:02] Maryana: ^ [22:37:02] hmm, you may be right. i may have counted this up wrong [22:37:06] :) [22:37:19] these are % of total traffic, not % of mobile traffic alone [22:37:29] if you couned these as % of mobile traffic then yeah, it'll be <1% [22:37:33] but they aren't. [22:37:33] gotcha [22:37:43] these pages are confusing :( [22:37:44] :) [22:37:53] thanks for catching that [22:38:03] :) [22:50:20] New patchset: JGonera; "Story 831: Improve left nav" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70932 [22:50:47] yurik, sorry for the delay. agreed, code duplication is usually best avoided. feel free to deduplicate - you'll need to move a bit of of logic around. [22:57:50] New patchset: JGonera; "Story 831: Improve left nav" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70932 [22:59:03] New patchset: MaxSem; "Create a thin base class for all our skins" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/69336 [23:00:23] New patchset: JGonera; "Story 831: Improve left nav" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70932 [23:08:57] jdlrobson, the new nav commit, finally: https://gerrit.wikimedia.org/r/#/c/70932 [23:11:10] New review: Jdlrobson; "Making some adjustments" [mediawiki/extensions/MobileFrontend] (master) C: -2; - https://gerrit.wikimedia.org/r/70235 [23:15:14] New patchset: JGonera; "Remove useless HTML comments" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70935 [23:18:29] Change merged: jenkins-bot; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70935 [23:18:56] New patchset: JGonera; "Remove unused .full-screen-search CSS rules" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70937 [23:37:04] New patchset: JGonera; "Remove #mw-mf-page-left id" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/70942 [23:43:40] Maryana: i'm trying to dig up the tablet editing stats. do you remember which thread they were in? [23:43:49] hmmm [23:43:56] i'll poke around [23:44:28] if we have them they would help tablet discussion [23:44:38] it's called Edits by platform / device [23:48:57] Maryana: email or docs? i'm not seeing it [23:49:11] email. i can refwd it if you can't find it [23:50:14] Maryana: yes please [23:50:48] sent! [23:52:44] thanks [23:54:46] np [23:55:22] Maryana: am i reading that right that iPad added up is larger then Linux editors ? [23:58:02] yeah, either that or right about the same [23:58:18] … which is terrifying [23:59:09] Maryana: why is it terrifying to you? [23:59:35] just scary to imagine those two demographics facing off :)