[00:03:48] (03PS13) 10Dr0ptp4kt: WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [00:04:55] (03CR) 10jenkins-bot: [V: 04-1] WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [00:07:01] ^jdlrobson, oops. had this working before trying to rebase. in rebase broke something. will figure this out tomorrow morning [00:46:05] (03PS2) 10Jdlrobson: WIP: Use MediaWiki UI buttons in alpha [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102581 [00:46:11] (03CR) 10jenkins-bot: [V: 04-1] WIP: Use MediaWiki UI buttons in alpha [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102581 (owner: 10Jdlrobson) [00:48:07] (03CR) 10Jdlrobson: "I have attempted to move mobile towards using the standard mediawiki.ui buttons but am hitting difficulties that will need to be resolved " [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102581 (owner: 10Jdlrobson) [01:07:31] jdlrobson: https://wikitech.wikimedia.org/wiki/Help:Addresses#Request_a_Public_IP_address [01:54:28] (03PS5) 10Robmoen: Added handling for captcha support on saveError [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 [01:55:35] (03PS6) 10Robmoen: Added handling for captcha support on saveError [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 [01:56:22] (03CR) 10Robmoen: "So this is working great for me now. It's rebased and I've marked it dependent on a change in the commit message." [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 (owner: 10Robmoen) [02:18:09] AndyRussG: You around? [02:18:20] Hey kaldari [02:18:24] Yep here I am [02:18:52] AndyRussG: Jon mentioned that the Education Extension uses some sort of HTML templating system. [02:19:05] Ah [02:19:07] Is that accurate and do you know anything about it? [02:19:30] I'm trying to inventory every templating system used in every MediaWiki extension [02:19:41] so that we can decide on a standard and put it in core :) [02:20:19] It's not the Education Extension. It's a bit of code that I showed during the interviews, but it's from the experimental Article Creation Help extension that I made prior to working on the EP extension [02:20:41] I think that's a great goal you're working on, though [02:21:48] https://github.com/AndrewGreen/mediawiki-extensions-ArticleCreationHelp/blob/3586632d4593e780bc003c9f9cf71897695e80b7/modules/ext.articlecreationhelp.js#L475-532 [02:22:39] In a nutshell, since I didn't know of any standard ones in core, I just made a small ad-hoc one, with the idea of changing it to whatever got into core eventually [02:24:00] It's basically similar to dust in the basic approach. Compiles stuff into an eval. https://github.com/AndrewGreen/mediawiki-extensions-ArticleCreationHelp/blob/3586632d4593e780bc003c9f9cf71897695e80b7/modules/ext.articlecreationhelp.js#L338-340 [02:25:57] Ah cool, I'll list it as "custom" then [02:26:06] Yeah, custom, exactly [02:26:34] Which solution is in the lead so far? [02:28:58] probably Hogan.js or Mustache.js since they ahve both PHP and JS implementations [02:29:19] Hmm interesting [02:29:21] although personally, I like Underscore.js [02:29:39] but it's only client-side [02:30:16] Yeah, I agree it's nice [02:30:39] gotta run! [02:30:43] K [02:30:45] Thanks for the info! [02:30:46] Have fun! [02:30:52] U bet [02:31:00] Here's the RfC so far.... [02:31:12] https://www.mediawiki.org/wiki/Requests_for_comment/HTML_templating_library [02:31:41] Feel free to add to it if you have any thoughts/ideas [02:41:58] (03PS2) 10Krinkle: [Bug 37278] Set X-Analytics header on XHR reqs [apps/win8/wikipedia] - 10https://gerrit.wikimedia.org/r/84135 (owner: 10Brion VIBBER) [06:58:08] (03CR) 10Siebrand: [C: 04-1] "The Manifests special page needs an entry in the special page aliases file." [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [13:54:35] (03PS14) 10Dr0ptp4kt: WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [13:55:10] (03CR) 10Dr0ptp4kt: "Added to aliases file." [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [15:07:22] (03PS15) 10Dr0ptp4kt: WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [15:14:12] (03PS16) 10Dr0ptp4kt: WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [15:19:15] (03PS1) 10Yuvipanda: Ensure we close opened files [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102693 [17:25:30] yuvipanda: Any idea what causes grrrit-wm to login as local-lolrrit instead of lolrrit? [17:25:51] marktraceur: hmm, not sure. [17:26:04] I gotta figure someone's restarting it in a weird way [17:26:11] that's possible, yeah [17:26:39] And it's not like people would LOG WHEN THEY RESTART [17:26:50] Grumble grumble grumble [17:27:04] legoktm: How do you restart grrrit-wm? [17:27:15] (as the only person who's not me who's in the logs) [17:27:26] um [17:27:31] ./src/kick.bash ? [17:27:37] Damn, that checks out. [17:27:39] https://wikitech.wikimedia.org/wiki/Grrrit-wm [17:27:43] Yeah, I know [17:27:45] just copy what's there :P [17:27:49] I meant, how do you personally do it [17:27:59] Anyway [17:28:14] The mystery continues [17:28:30] [09:26:49 AM] yuvipanda: Any idea what causes grrrit-wm to login as local-lolrrit instead of lolrrit? <-- isn't that just the ident thing that tool labs has? [17:28:43] legoktm: yeah but you can override it i think [17:28:48] all my IRC bots run from labs sign on as local-lego[bot] [17:28:57] when you connect [17:29:06] I just logged in normally and did ./src/kick.bash and it's logged in as lolrrit@ [17:29:28] :| [17:29:35] that's what I normally do. [17:30:28] legoktm: marktraceur maybe it disconnected at some point and reconnected (network?) and that caused it to lose lolrrit@? [17:30:28] idk [17:30:40] Ah, might be [17:32:02] marktraceur: btw, we postponed plans for media viewing in the wikipedia app, at least for a while [17:32:10] marktraceur: doing anon editing and logged in editing instead :) [17:32:27] marktraceur: do you have any idea of when viewer will get online everywhere? [17:36:36] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [17:37:12] yuvipanda: Uh...you mean out of beta? [17:40:32] (03CR) 10Siebrand: [C: 04-1] "i18n/L10n reviewed." (031 comment) [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [17:41:11] marktraceur: yeah? [17:42:29] yuvipanda: Probably not 'til March-ish [17:42:43] marktraceur: alright! [17:54:25] rmoen: you there [17:55:00] (03PS1) 10Cmcmahon: load the page into cache before trying to use it [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102719 [17:55:24] (03PS7) 10Jdlrobson: Added handling for captcha support on saveError [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 (owner: 10Robmoen) [17:56:00] (03CR) 10Dr0ptp4kt: "Set it to 'MobileWebApp' => array( 'MobileWebApp' )" (031 comment) [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [17:56:11] (03PS17) 10Dr0ptp4kt: WIP: Experiment: Special:MobileWebApp and manifests for offline FF & Chrome, FFOS + CWS. [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [17:58:08] rmoen: jdlrobson: will one of you be able to do a brief demo of VE on mobile during today's planning? [17:58:15] (03PS2) 10Cmcmahon: load the page into cache before trying to use it [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102719 [17:58:21] awjr: sure [17:58:30] sweet! [18:00:19] jdlrobson: try $wgCaptchaTriggers['edit'] = true; [18:03:42] (03CR) 10Jdlrobson: [C: 032] Introduce EditorOverlayBase [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 (owner: 10Robmoen) [18:03:59] (03CR) 10Jdlrobson: [C: 04-1] "Please rebase!" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99776 (owner: 10Robmoen) [18:04:01] yurik, able to make the standup? [18:04:19] (03Merged) 10jenkins-bot: Introduce EditorOverlayBase [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/99773 (owner: 10Robmoen) [18:04:31] http://etherpad.wikimedia.org/p/MobileWeb-Q3-Planning [18:04:34] brion: hey! [18:04:37] kenanwang: http://etherpad.wikimedia.org/p/MobileWeb-Q3-Planning [18:05:03] dr0ptp4kt, joining [18:05:05] yo yo [18:05:53] brion: saved page images - it's getting a bit hairy, but should be done soon [18:05:59] sweet [18:06:10] brion: tried out 4.3 vs 4.4 for the marshaller error? [18:06:51] awjr: you're still in SF yes? [18:06:56] chrismcmahon: yup! [18:07:03] chrismcmahon: aye [18:07:15] yuvipanda: not yet [18:07:20] booking my xmas flights [18:07:28] cause i really don't want to drive to 8-10 hours each way [18:07:41] and the high-speed rail will never be built in my lifetime *grumble* [18:07:51] brion: haha! :D [18:08:50] awjr: I'm pretty sure I know why Mobile browser tests always got 503s for Chrome not Firefox, and I'm pretty sure I have a workaround at least that should prevent the 503s from causing test failures, if we can't prevent the 503s themselves. [18:09:28] sounds cool chrismcmahon - im in quarterly planning meetings now but will be free to chat more later today [18:28:58] (03CR) 10Siebrand: [C: 031] "i18n/L10n reviewed." [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/100919 (owner: 10Jdlrobson) [18:32:18] brion: btw, awesome thing I just found today [18:32:23] brion: adb lolcat works :D [18:32:33] waaaaaaaat [18:32:45] haha [18:32:52] brion: there's also adb hell [18:35:21] (03CR) 10Jhall: [C: 032] load the page into cache before trying to use it [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102719 (owner: 10Cmcmahon) [18:35:55] (03Merged) 10jenkins-bot: load the page into cache before trying to use it [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102719 (owner: 10Cmcmahon) [19:10:57] (03CR) 10Brion VIBBER: Ensure we close opened files (031 comment) [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102693 (owner: 10Yuvipanda) [19:11:42] (03CR) 10Yuvipanda: Ensure we close opened files (031 comment) [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102693 (owner: 10Yuvipanda) [19:27:55] awjr, ceiling cat is watching you [19:28:09] ?!!?!?! [19:28:15] MaxSem: cieling cat? [19:28:24] your emails [19:28:24] you're freaking me out man [19:28:26] ... [19:28:40] my emails?? [19:28:58] awjr: MaxSem knows all [19:29:07] !!!! [19:29:16] awjr: NEEDS MORE EXCLAMATION MARKS! [19:29:27] In non Soviet America, MaxSem watch you [19:29:31] !!!!!!!!!!!! [19:35:48] mhurd: brion yuvipanda : is https://trello.com/b/YVy108mS/mobile-app-sprint-20-saved-pages up to date? i'm still seeing two stories in the to do column [19:36:01] which goes against our thinking that we would go through all of these [19:36:10] tfinc: well, look at the checklists on each of them [19:36:23] tfinc: they each have one item remaining, and I should be able to finish one today and one tomorrow [19:36:31] actually my bigger confusion is that nothing is in the doing column [19:36:39] yuvipanda: mhurd --^ [19:37:01] tfinc: ah, right. I had moved one to CR because one of the checklist items was in CR, and then it got merged, now moved it back to 'Doing' [19:37:05] mhurd: how is https://trello.com/c/dotX2UCi/8-bug-56967-uploads-broken-in-version-1-0-10-on-ios7 going ? [19:37:06] I think that's confusing, so I'll stop doing that [19:37:24] ok [19:37:39] the CR column is mostly useless now at the current size of stories, we should consider killing it a couple of iterations down the line [19:37:50] tfinc: been refactoring / bug fixing / menu experimenting. will take look at commons upload bug today [19:37:58] brion: grettings [19:38:08] hey hey [19:38:12] mhurd: you can always see if brion can pick it up [19:38:38] brion: hey brion! will have that menu commit for you soon! [19:38:48] spiffy [19:40:11] crap, I keep hitting F5 in the hangout window [19:41:38] brion: think you could get hatjitsu up and running for all of our teams ? [19:47:21] tfinc: http://brionv.com:5000/ [19:47:54] awjr: --^ [19:48:09] oh sweet! [19:48:25] tfinc: is it stable enough for us to use right now? [19:48:41] brion: ? [19:48:47] it shouldn't crash :) [19:48:47] brion: gah, so much threading. grr [19:48:52] the software is fine. just a matter of the server handling it [19:48:58] let's try it :) [19:48:59] i've got it running in a screen session so it won't die if my wifi dies [19:49:22] awjr: lets try it http://brionv.com:5000/30939 [19:49:31] brion: awjr : jump into the one i just linked [19:49:45] tfinc - yeah, i want to wait for this convo to die [19:49:57] awjr: just a test with the three of us first [19:50:02] lets vote! [19:50:12] one more vote .. [19:50:16] seems like it works :) [19:50:18] \o/ [19:50:48] hhaaattt--jit-sssuuu! [19:50:48] jdlrobson: jgonera rmoen MaxSem http://brionv.com:5000/59248 [19:51:04] kaldari: http://brionv.com:5000/59248 [19:53:03] moizsyed: http://brionv.com:5000/59248 [19:53:28] make sure to mark yourself a non-voter :) [19:53:29] kaity: ^ [19:53:33] tfinc: ^ [19:54:35] moizsyed: kaity tfinc one of you appears to be a voter - please mark yourself as a nonvoter [19:54:46] i'm not on anymore [19:54:59] but i wanted to vote [19:55:03] i smell a what is my name patch to hatjitsu [19:55:07] democracyyyyyy! [19:55:14] moizsyed: only devs vote. awjr can fill you in [19:57:33] (03PS1) 10Yuvipanda: Download all the images in an article while saving page [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 [19:57:36] brion: ^ [19:57:43] \o/ [19:57:53] brion: it's a little bit hairy because of the multithreading going on [19:58:00] brion: I also added okhttp as a dependency [19:58:02] will tune it laster [19:58:02] heh [19:58:03] ok [19:58:03] *later [19:58:11] brion: it also doesn't actually *use* the images anywhere yet [19:58:34] brion: and the reason I'm not using md5 to save image names is that I'll have to re-implement that in JS too, so preferred something simpler. [19:58:43] ah [19:58:58] well js md5 implementations ar easy enough to drop in if we need em [19:59:04] brion: yeah, BUT PERFORMANCEEE!!!! [19:59:07] haha [20:00:06] yuvipanda: so monte's experimenting with using libxml2 to parse the html outside of the web view for this, which has the advantage that we can fetch the image list during background refreshes without having to instantiate an offscreen WebView [20:00:18] any thoughts on that approach vs the in-webview lookup? [20:00:31] brion: makes sense if you want to do a background refresh :P [20:00:36] :) [20:00:53] brion: I'd want to stay off libXML as late as possible [20:01:11] hehe [20:01:13] brion: since, if we parse it into a DOM [20:01:17] brion: that uses up memory again [20:01:26] brion: while we *already* have a DOM [20:01:28] in the webview [20:01:36] brion: so yes, it makes sense for background operations [20:01:42] but we can implement when we get to it :D [20:01:44] well you can do a straight event-based parse too [20:01:46] yep [20:01:48] we have options :D [20:01:51] brion: yeah, and even then i want to do a pull parse [20:01:54] than a DOM [20:02:54] brion: but pull parses are complicated, so I'd rather not do it until we absolutely need to [20:03:03] *nod* [20:03:04] YAGNI [20:03:25] brion: we'll have to develop some form of abstraction there too, since what we need is an XPath like thing but without a DOM [20:03:39] brion: but yeah, not now :) [20:03:48] hmm, wonder if there's a CSS selector library that uses pull parser [20:04:06] could return you just DOM fragments you want, or callback [20:04:34] brion: i dunno if you can efficiently implement 'full' CSS selectors without a DOM [20:04:54] brion: something as simple as "a" needs to look at *all* elements, for example [20:05:11] brion: so needs to be a subset [20:05:41] brion: and yeah, we should return DOM fragments, and cache those too... [20:05:48] brion: it's a rather fun problem to solve, yea [20:06:26] brion: anyway, review on the current patch? [20:06:31] looking it over :) [20:06:40] (03CR) 10Brion VIBBER: Download all the images in an article while saving page (031 comment) [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 (owner: 10Yuvipanda) [20:07:27] brion: yeah, I think I'll just md5 it. [20:07:57] heh [20:08:13] there's always another corner case [20:08:50] brion: yeah [20:11:16] (03CR) 10Brion VIBBER: Download all the images in an article while saving page (031 comment) [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 (owner: 10Yuvipanda) [20:12:05] brion: what do we do when some of the images fail? [20:12:10] (03CR) 10Yuvipanda: Download all the images in an article while saving page (031 comment) [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 (owner: 10Yuvipanda) [20:12:16] brion: right now any failure just... blows everything up :P [20:12:43] hah [20:15:10] brion: so... do we say 'save failed'? [20:15:16] brion: if one of them goes down? [20:15:26] brion: okhttp already tries some retries and stuff... [20:15:54] probably either say save failed, or at least continue the process instead of waiting forever :D [20:16:08] or will it have just thrown an exception from the other thread? [20:16:36] brion: right now it just throws an exception. [20:16:52] brion: I think 'save failed' sounds appropriate [20:16:57] ok, at least that'll be a visible error (app closes with crash) [20:17:05] brion: yeah, that's the case right now [20:17:07] yeah catching and showing an error is probably better than just crash [20:17:08] ok [20:17:17] brion: onCatch's default behavior is to wrap it in a RuntimeException and throw it [20:17:25] brion: so if we don't have an onCatch handler, by default it just throws up [20:17:28] ah bingo [20:17:59] brion: one of the reasons it's called *Sane*AsyncTask :P [20:18:05] hehe yeah [20:18:21] as opposed to "InsaneRegularJavaThingThatFailsSilently" [20:20:58] time to feed [20:20:59] bbiab [20:45:58] brion: are you planning to keep your hatjitsu instance running foreva? [20:46:31] maybe we should get it up and running a labs instance - i think the web team would like to start using it regularly [20:52:14] (03PS2) 10Jdlrobson: Fix fatal [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102463 (owner: 10MaxSem) [20:54:35] (03CR) 10Jdlrobson: [C: 04-1] "Something wrong is going on - when I go to Special:Contributions/Namethatdontexist I still see the header bar and it queries the entire da" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102463 (owner: 10MaxSem) [20:58:55] (03PS3) 10Jdlrobson: Fix fatal for non-existent users on SpecialMobileContributions [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102463 (owner: 10MaxSem) [20:59:30] (03CR) 10Jdlrobson: [C: 031] "Max thoughts on my slight rewrite?" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102463 (owner: 10MaxSem) [21:06:22] brion: I found a md5 js library [21:06:24] brion: is 6KB :( [21:25:03] (03PS1) 10Yuvipanda: Clear bridge listeners when they are no longer used [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102818 [21:30:39] MaxSem: better? [21:30:46] aha [21:38:58] brion: around? [21:41:02] yuvipanda: yo [21:41:20] awjr: yeah we should stick one on labs [21:41:32] this one'll die next time i reboot the machine for maintenance :) [21:41:50] brion: md5, the library I found was like 6KB :( Don't want to add something that heavy to every page for some feature that'll not be used all the time... [21:42:19] yuvipanda: do we need to do the md5'ing in js? [21:42:28] we could bounce an array of urls to the java side [21:42:41] brion: yeah, that's what I'll probably do. [21:42:46] k [21:42:47] feels a little ugly, tho. [21:42:53] 6kb ain't too bad really [21:43:00] but every byte counts on slower phones :) [21:43:06] * brion stabs 2.3 [21:43:30] brion: slippery slope. Today we'll be including 6KB of JS into every page, and tomorrow it will be legal for men to marry macbook pros! [21:43:42] haha [21:44:07] brion how hard is it to get running? [21:44:26] awjr: iirc it's pretty much 'check out, run npm something, run 'node server.js'' [21:44:41] unless you want to puppetize it :D [21:44:56] all non-puppetized, non-toollabs things will disappear when labs migrates to eqiad in a few months [21:45:02] https://github.com/richarcher/Hatjitsu.git [21:45:08] bah [21:45:31] and of course node server.js will not survive restarts :P [21:45:49] init scripts are left as an exercise for the student :) [21:46:26] brion: heh, node community uses something called 'forever', which is like an init script but not really... [21:46:33] you could also use init scripts, of course [21:46:39] back in my day, we added things to /etc/rc.local by hand [21:47:02] we had to pipe files uphill through the snow both ways [21:47:31] we had to know PERL! [21:47:32] brion: heh, now we have supercharged sockets that deal with actual strings! no need to make us mess around with buffers manually even :P [21:47:49] * brion overflows$&(*#HR*&FH^GW*&HDSUIJ [21:48:25] brion: heh [21:48:35] brion: anything other than the md5 in that patch? [21:48:48] yuvipanda: i think we're good [21:48:54] can improve the error handling later if not already in there [21:49:05] brion: ok [21:49:54] (03CR) 10Brion VIBBER: [C: 032 V: 032] "edge case here can be cleaned up later, merging" [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102693 (owner: 10Yuvipanda) [21:50:12] yuvipanda: you making a change to https://gerrit.wikimedia.org/r/#/c/102743/ with the md5 or another commit? [21:50:30] brion: another commit, I'd think. [21:50:36] ok i'll merge it up [21:50:38] ok [21:51:24] (03CR) 10Brion VIBBER: [C: 032 V: 032] "merging, but the url key will have to be changed to a full url hash later for those long-file corner cases. this'll come in another commit" [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 (owner: 10Yuvipanda) [21:52:32] brion: I'm thinking 'hash of url minus protocol + extension' [21:52:34] brion: sounds ok? [21:52:56] eh i'd just use full url. why crop it? [21:53:13] brion: minus protocol? [21:53:29] brion: i mean, it'll be "hash.extension", and hash is computed after stripping protocol [21:53:38] ah i see what you mean [21:53:39] brion: not stripping extension off URL :) [21:54:11] brion: yeah, cache fragmentation and all that [21:54:14] yeah may as well strip the protocol in case we have to jump around between http and https [21:54:22] like visiting china and then returning to https land :) [21:54:49] and adding the extension makes sure the file loads as right type when used direct [21:54:52] is good :D [21:54:58] brion: :D [21:56:15] brion: I think stripping protocol is not necessary [21:56:22] brion: because we use protocol relative URIs everywhere [21:56:38] well [21:56:49] if you fetch as img.src it may expand it actually [21:57:02] might need img.getAttribute('src') [21:57:04] brion: ah, yes [21:57:10] brion: I just checked that, and you are correct [21:57:14] i'm not 100% sure though, i think a.href does the same annoying magic [21:57:15] stripping away! [21:57:25] woo [21:57:25] brion: I checked, it does base it off the same protocol as tag [21:57:26] so https [21:57:31] hah [21:57:41] html dom, they name is wtf [21:57:46] *thy [21:59:08] anyone here knows rails? [21:59:17] MaxSem, yep, why? [21:59:26] pm [21:59:33] jgonera: there's an RFC for rewriting mw in rails, I think [21:59:39] haha [21:59:51] brion: remember to tackle ^ in the arch summit :P [21:59:57] I would not use Rails for mw [21:59:58] yuvipanda: hm, was there a resource for the toast messages? i'm testing and i realize it's not actually building [21:59:59] heh [22:00:09] brion: hmm? Yes there was [22:00:16] i think it got forgotten :D [22:00:24] brion: oh, yes. [22:00:32] brion: intelliJ kinda... put them in a different file [22:00:33] stupid [22:00:48] (03CR) 10Brion VIBBER: "toast resource ids are missing, needs fixing. whoops!" [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102743 (owner: 10Yuvipanda) [22:01:02] unintellij [22:01:50] brion: heh [22:01:55] (03PS1) 10Yuvipanda: Add missing toast strings [apps/android/wikipedia] - 10https://gerrit.wikimedia.org/r/102823 [22:01:56] brion: that's also because I had used R.id instead of R.string [22:01:58] brion: & [22:02:00] gah [22:02:04] ^ [22:02:10] hah [22:02:15] how did that build for you?? [22:02:19] or did neither of us test it [22:02:24] * brion <- CR FAIL [22:02:29] brion: no, it built for me [22:02:34] brion: because it created an file [22:02:40] ahhhhh [22:02:41] brion: and internally they all compile down to the same thing [22:02:43] weird [22:02:45] brion: since they are all ints [22:02:54] brion: if I had added the ids.xml file it would've compiled for you too [22:03:15] brion: we should setup jenkins to build :P [22:03:20] we should yeah [22:03:22] heh [22:09:11] (03CR) 10JGonera: "So .toc will be removed elsewhere now?" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102364 (owner: 10MaxSem) [22:10:04] brion: thumbnails will always have an extension, right? [22:10:24] yes [22:11:58] brion: ok [22:11:59] (03CR) 10MaxSem: "http://git.wikimedia.org/blob/mediawiki%2Fextensions%2FMobileFrontend.git/af604045f030c16b1f7305787c642a6a17a499bd/includes%2FMobileFronte" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102364 (owner: 10MaxSem) [22:12:07] (03CR) 10JGonera: [C: 032] Avoid using router for Issues overlay when we are in Editing overlay [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102286 (owner: 10Kaldari) [22:13:13] (03Merged) 10jenkins-bot: Avoid using router for Issues overlay when we are in Editing overlay [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102286 (owner: 10Kaldari) [22:14:22] jdlrobson, kaldari, MaxSem, awjr_away I'm thinking about going to bed at 3pm PDT. do you have anything for me? I'd also prefer to work more EU hours tomorrow and wrap up shortly after stand up tomorrow [22:14:44] have sleep jgonera [22:14:46] nothing pressing [22:15:07] jgonera: Although I would love to get your input on https://www.mediawiki.org/wiki/Requests_for_comment/HTML_templating_library tomorrow [22:15:10] I should have the overlay manager patch ready tomorrow [22:15:20] sure [22:17:20] kaldari, so far so good. do you need help writing the implementation section? [22:18:09] (03CR) 10Jdlrobson: "After applying this patch I still see the confirmation "Do you really want to abandon your edit?" when clicking in preview this page has i" [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102286 (owner: 10Kaldari) [22:18:09] yes [22:18:22] OK, I'll add something tomorrow [22:20:11] brion: I'm going to ditch extensions. I don't think browsers need 'em anymore [22:22:10] (03CR) 10JGonera: [C: 04-1] Reset the issues array before we populate it (032 comments) [extensions/MobileFrontend] - 10https://gerrit.wikimedia.org/r/102317 (owner: 10Kaldari) [22:22:19] kaldari: jgonera that change broke editing in stable [22:22:22] jgonera: [22:22:25] Error: Module not found: OverlayNew Error {} [22:22:33] jdlrobson, which change? [22:22:41] https://gerrit.wikimedia.org/r/102286 [22:22:56]