[15:33:42] MaxSem: someone, then, should update the linked documentation about it: http://meta.wikimedia.org/wiki/Mobile_Projects/Mobile_Gateway#How_do_I_format_a_mobile_homepage.3F [15:34:11] cough cough http://meta.wikimedia.org/wiki/Mobile_Projects/Mobile_Gateway#Global_Prefix [15:46:21] Cough cough: http://meta.wikimedia.org/w/index.php?title=Mobile_Projects%2FMobile_Gateway&diff=3931470&oldid=3857634 [15:47:13] (ps: no one will look down to the global prefix section if they can find something above it which might work) [16:57:25] http://mobile-geo.wmflabs.org/w/index.php/Special:MobileOptions/BetaOptIn < linSmith awjr MaxSem [17:01:27] * MaxSem looks around [17:01:35] MaxSem, jdlrobson: https://plus.google.com/hangouts/_/738f600c611d6f618c45f60359d0d9b442b6e091 [17:01:52] jdlrobson, doesn't seem to opt in for me [17:02:31] MaxSem: WFM... [17:02:38] opting in is weird [17:03:06] jdlrobson: standup! https://plus.google.com/hangouts/_/738f600c611d6f618c45f60359d0d9b442b6e091 [17:03:12] i'm going :D [17:10:09] MaxSem: i hear asher's point about getting things running in labs, but i am concerned about us getting hardware in time if something needs to be ordered/set up [17:10:49] MaxSem i can respond to the RT ticket in a bit and let him know [17:11:03] cool, thanks [17:13:46] aharoni was asking yesterday if we had any plans to get our own OSM tile server [17:13:50] do we? [17:14:09] http://mobile-geo.wmflabs.org/w/index.php/Special:MobileOptions/BetaOptIn < philinje [17:14:32] There is an OSM tile server being run on toolserver, and there was discussion about it being moved to other boxen. [17:14:45] Aude: who was running the toolserver instance? [17:17:39] bbl [17:24:24] jdlrobson: looks good, thanks [20:43:21] MaxSem: observation - http://en.wikipedia.org/w/index.php?title=Special:ApiSandbox&debug=true#action=tokens&format=json should accept values for token that match up with other parts of the api that need them. For example to login I need an lgtoken - I should be able to do http://en.wikipedia.org/w/api.php?action=tokens&format=json&type=lgtoken [20:49:03] jdlrobson, action=login is inconsistent with others, which don't prefix their token parameters [20:49:22] also emailuser,rollback,upload,userrights,filerevert,login [20:49:31] e.g. action=edit&token=.... [20:51:58] I know how it is I'm just commenting It's not consistent and thus confusing [20:54:57] and I'm trying to explain is that it's action=login which is unlogical, not action=tokens [20:59:23] jdlrobson, does https://gerrit.wikimedia.org/r/#/c/16187/ mean that this callback will be called twice? [21:00:25] MaxSem: I don't follow [21:00:49] In Chrome a pop state event is called on page load. In Firefox it doesn't [21:01:08] Thus currently Api Sandbox is throwing an exception for me in chrome as they are undefined on startup [21:01:10] ah, got it [21:01:21] cool :) [21:01:26] I'm explicitly calling it later [21:02:00] so not creating these controls for the first time doesn't break anything [21:08:52] MaxSem: thoughts on http://www.mediawiki.org/wiki/Thread:Extension_talk:MobileFrontend/How_Do_we_make_to_Mobile_view_stick%3F#How_Do_we_make_to_Mobile_view_stick.3F_16841 ? [21:11:36] the setcookie change he proposes will break WMF, I think [21:11:59] or fix it, since sticky cookies are failish right now:P [21:12:30] ^ awjr [21:12:57] eh? [21:13:01] that link doesn't seem to work [21:13:05] "No such thread" [21:13:42] http://www.mediawiki.org/wiki/Thread:Extension_talk:MobileFrontend/How_Do_we_make_to_Mobile_view_stick%3F#How_Do_we_make_to_Mobile_view_stick.3F_16841 [21:14:33] or http://is.gd/dgCrKu [21:15:27] reading quickly the change he offered will not work as desired on the cluster [21:17:13] i think we just need to set the cache to vary on stopMobileRedirect cookie, which i thought it had been doing at one point [21:17:19] i need to dig through old config [21:20:43] part of the problem is that the way that view toggling needs to work on the cluster is very different from how it should work out-of-the-box because we have kind of a funky set up to ensure optimal caching for our huge amount of traffic [21:21:16] it looks like the modifications that guy made were to get it to work locally [21:21:29] once i get thorugh reviewing a few more changesets i'll take a closer look. [21:23:00] gawd, JS gives me teh creeps [21:24:06] tehe [21:37:40] awjr: there is an open bug for that [21:37:42] let me dig it out [21:37:57] awjr: https://bugzilla.wikimedia.org/show_bug.cgi?id=36894 [21:40:15] MaxSem: > https://gerrit.wikimedia.org/r/16191 [21:40:19] this was what I was trying to do [21:40:27] but failed due to inconsistent api [21:41:56] thanks jdlrobson [21:42:09] hi Amgine [21:42:10] i think the real fix there is integrating MF and device detection with core [21:42:19] but who knows when that will actually happen [21:47:05] jdlrobson, already looking [21:47:46] awjr, right! and serve everything from the same domain [21:47:56] yah [21:50:21] @preilly do you know if you have any breakdown on usage by day of the week? [21:50:31] i remember we used to get a ton of usage on friday/sat night EST/PST [21:50:35] wondering if that still happens [22:31:29] jdlrobson, it's so great to have one more person interested in APISB! [22:32:56] :) [22:33:02] I was thinking about a wizard for newbies [22:33:08] handling basic stuff [22:33:14] e.g. Special:ApiSandboxWizard [22:33:42] might be a bit ambitious but would be good if it could generate code too :) [22:37:16] yeah, code generation is what brion proposed during the initial review [22:38:03] might have a look at that now [22:40:57] MaxSem: where do the descriptions come from btw? [22:42:01] from action=paraminfo [22:48:25] so they are code driven rather than wiki pages? [23:24:28] MaxSem: applyObject is one scary function [23:24:52] hehe [23:29:05] does it have to be recursive? [23:30:43] yes [23:38:28] MaxSem: just submitting something for review.. unless I'm misunderstanding something this does the same thing? [23:39:16] eh? [23:39:44] wait for it.. [23:39:49] https://gerrit.wikimedia.org/r/16199 [23:39:54] doesn't that do exactly the same thing? [23:41:36] is this code more readable? [23:41:50] recursion vs. code chaining [23:42:15] I think so [23:42:19] recursion is dangerousd [23:42:39] if something doesn't need to be recursive I don't think it should be [23:47:31] fixmed [23:47:58] jdlrobson, I unset fields for a reason [23:48:10] what was the reason? [23:49:39] to avoid setting them again [23:49:39] does it matter if they get set again MaxSem ? [23:49:50] are there any side effecs? [23:49:59] there was a bug about it [23:50:04] you could always blacklist those fields [23:50:28] anyway… i can always bring back the deleting