[00:09:37] preilly: peter geheres con fired the near by bug that i was seeing [00:09:57] whats your time like to take a look vs wp zero ? [00:16:38] [WikipediaMobile] brion pushed 1 new commit to master: http://git.io/RNXhjA [00:16:38] [WikipediaMobile/master] Followup 3ba59ff52b10f30f5d977df9274ab2e47fdc83e4: commit updated icon files! - Brion Vibber [00:16:51] Project WikipediaMobile - Nightly builds build #44: SUCCESS in 6.5 sec: http://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/44/ [00:16:51] brion: Followup 3ba59ff52b10f30f5d977df9274ab2e47fdc83e4: commit updated icon files! [00:20:55] brion: yuvipanda: philinje : what else do we want to fix before RC4 ? [00:21:27] it's probably too late to stop calling them RCs, though I guess we could fix that for next version [00:22:14] tfinc: currently investigating https://bugzilla.wikimedia.org/show_bug.cgi?id=33289, so I'd like that to go in [00:22:49] though it needs better phrasing. [00:23:11] tfinc: brion also can we decide on position:fixed? [00:24:00] yes, unless Tomasz disagrees, we should remove position fixed [00:24:24] Brion is trying one idea soon - put some default text in the search field and make it selected [00:24:52] that would help the case of hardware search not causing obvious focus on the search field [00:25:00] philinje: unless i'm mistaken, removing position:fixed should kill our focus problems, no? [00:25:13] yes, except the hardware search [00:25:46] yuvipanda: maybe it makes sense for you to move over to language issues? if nearby issues are settled [00:26:14] philinje: yes, looks like that's what I'm doing. [00:26:43] philinje: we've only 6 bugs blocking 1.0 release. Seems too less :P And I don't see the ICS bugs there... [00:27:05] what is ICS? [00:27:59] philinje: Ice Cream Sandwich? [00:28:19] philinje: https://bugzilla.wikimedia.org/show_bug.cgi?id=33218 [00:28:22] philinje: https://bugzilla.wikimedia.org/show_bug.cgi?id=33219 [00:33:36] yuvipanda: thanks. Not sure if we should block our 1.0 for ICS [00:34:03] philinje: from what I gather, they completely make the app unusable on ICS. Perhaps get brion's device and check it out? [00:34:26] yes, but it's possible ICS is a moving target [00:34:46] yeah it's unusable right now [00:35:16] some things already fixed in phonegap 1.3.0, other things not yet [00:35:22] some things may require more workarounds on our part [00:37:17] also heard some chatter about specific releases of ICS recently, makes it sound like it's not fully baked [00:46:46] philinje: are you suggesting we don't target ICS right now? [01:00:54] philinje: brion tfinc hitting the bed. gnite. [01:00:59] will be checking out lang bugs tomorrow [01:01:08] also I think there are 2 pull reqs needing merge love :) [01:01:17] bye everyone! :) [01:13:32] [WikipediaMobile] brion pushed 2 new commits to master: http://git.io/7lvuXw [01:13:32] [WikipediaMobile/master] add Globalization plugin for android - Brion Vibber [01:13:32] [WikipediaMobile/master] Use Globalization plugin to get language on Android; can now change when the activity gets reinitialized after changing languages but not fully shutting down the app - Brion Vibber [01:14:18] Project WikipediaMobile - Nightly builds build #45: SUCCESS in 7.6 sec: http://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/45/ [01:14:18] * brion: add Globalization plugin for android [01:14:19] * brion: Use Globalization plugin to get language on Android; can now change when the activity gets reinitialized after changing languages but not fully shutting down the app [01:23:06] yuvi: sorry to miss your good bye - another late one, eh? Thanks a lot! [01:28:17] [WikipediaMobile] preillyme pushed 1 new commit to master: http://git.io/dlE5bA [01:28:17] [WikipediaMobile/master] Merge pull request #90 from yuvipanda/better-language-switch - Patrick Reilly [01:28:25] [WikipediaMobile] preillyme pushed 1 new commit to master: http://git.io/nW6ipA [01:28:25] [WikipediaMobile/master] Merge pull request #91 from yuvipanda/actual-everything-https - Patrick Reilly [01:28:29] Project WikipediaMobile - Nightly builds build #46: SUCCESS in 6.1 sec: http://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/46/ [01:28:30] * yuvipanda: Go to language main page on language switch [01:28:30] * yuvipanda: Moved everything to https [01:51:16] preilly: how much of your time would the iOS refactoring take vs wp zero ? [01:51:41] tfinc: I'd need like a half day tomorrow to finish it [01:52:36] and whats your workload look like for zero ? [01:53:14] tfinc: not too bad [01:53:20] tfinc: I think I can fit both in nicely [01:53:37] ok. lets discuss tomorrow morning at the stand up then. [01:53:42] but I'm thinking its worth it [01:53:47] tfinc: okay [01:54:29] the other question is wether we should hold off on more testers till we get this fixed [01:54:35] i have lots more people ready to go [01:55:00] and a newly signed IPA [01:55:35] but .. peter geheres ran into the problem right away [01:56:16] tfinc: I'd say hold off [02:02:52] jorm: what are you thoughts about removing the position fixed placement of search on the android app ? [02:03:11] dew eeet. [16:06:55] [WikipediaMobile] Nikerabbit pushed 1 new commit to master: http://git.io/zR5wUA [16:06:55] [WikipediaMobile/master] Translation updates from translatewiki.net - Niklas Laxström [16:07:11] Project WikipediaMobile - Nightly builds build #47: SUCCESS in 8.6 sec: http://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/47/ [16:07:11] niklas.laxstrom: Translation updates from translatewiki.net [17:10:42] does the app from master work for anyone? I'm getting a weird js error.. [17:17:51] seems to for me [17:17:55] without the last i18n update [17:56:52] greetings all [17:57:12] MaxSem: yuvipanda_ ready for the standup ? [17:57:38] yup [17:57:42] tfinc: yes [18:00:39] MaxSem: if you can hear us .. we can't hear you [18:02:19] tfinc, my mic disappeared from devices list:P [18:02:34] yay windows [18:06:44] philinje: are we doing this for 1.0? https://bugzilla.wikimedia.org/show_bug.cgi?id=32906 [18:11:59] OpenStack is epic, as always [18:12:03] "Failed to allocate new public IP address." [18:13:30] so .. fixed position [18:13:52] MaxSem: I think you need to use the one allocated in the pool [18:14:25] my opinion is that we can kill fixed position if we can bring back the search button with one tap (like the kindle app) . if a user has to scroll to the very top of the screen to get it back then i think were failing [18:14:51] tfinc: there's the dedicated 'search' key [18:15:32] we won't get that on iOS and can potentially lose it on future android devices [18:15:58] technically can we do the tap to see the search input box ? [18:15:59] both tap and search key will have the focus bugs [18:16:19] yuvipanda_: 32906 is one Brion was working on. Brion? [18:16:19] same exact issues ? [18:16:23] yes [18:16:52] Android doesn't like something over an iframe [18:17:22] so what would happen if i was at the bottom of an article and i tapped the hardware search button ? [18:17:46] the keyboard appears but the search field does not seem to be in focus [18:17:47] i haven't done any work on https://bugzilla.wikimedia.org/show_bug.cgi?id=32906 - just made some comments on the subject when we discussed it yesterday [18:17:49] would it just shop up to the position I've scrolled. would it scroll all the way to the top and show me the search box ? [18:18:08] i mean if we didn't have fixed position [18:18:11] tfinc, scrolls tot op yes [18:18:27] thats a pretty jarring experience [18:18:34] Brion is trying an idea I saw in another app - use some default text in the search field and make it selected [18:18:34] brion: can't we just put it actually outside and above the iframe [18:18:38] and hide/show it as needed? [18:19:01] hmm, is the iframe scrolling or is the app scrolling? [18:19:11] yuvipanda_, nope! we have to let the app window scroll so the iframe is very tall [18:19:22] and when scrolled, there's no space at the top of the screen that's not on top of the iframe [18:19:34] philinje: as in "Type your search here..." ? [18:19:36] sigh [18:19:41] sigh indeed :P [18:19:42] tfinc: yes [18:20:05] if the default text is selected, it's very obvious the search field is in focus [18:20:10] philinje: thats the text we use on full screen search :) [18:23:56] preilly, howww? [18:24:15] yuvipanda_: Brion is dealing with 32906, may not make it into 1.0 [18:24:31] lets assign bugs so that we don't have confusion [18:24:36] okay. [18:25:46] browser share data for december, 2012: http://arstechnica.com/business/news/2012/01/state-of-the-browser-chrome-closes-on-firefox-ie6-dying-out.ars [18:26:07] as usual, mobile safari utterly and completely dominates the FUCK out of the mobile space. [18:26:12] added a couple notes on https://bugzilla.wikimedia.org/show_bug.cgi?id=32906#c10 [18:26:58] http://static.arstechnica.com/01-04-2012/browser-stats/firefox-adoption.png <- this makes me ecstatic [18:27:06] jorm: that safari traffic matches up nice with with breakdown that nimish had [18:27:11] the new firefox version rollouts are mostly working pretty well [18:27:14] 3.6 ain't dead yet though :D [18:27:21] brion: you going to the meetup? [18:27:35] 53%. [18:27:36] think i'll skip this one [18:27:55] http://static.arstechnica.com/01-04-2012/browser-stats/internet-explorer-adoption.png <- hate hate [18:27:58] i love how mobile ie is less than 1% [18:28:15] ie 6 is less than 1% in the usa now. [18:28:21] so that's something to celebrate. [18:28:23] jorm: i'm surprised its even on the graph [18:29:23] MaxSem: not sure why it's failing [18:29:38] you can see on those graphs when and where the chrome and firefox builds got "automated upgrades" [18:29:44] preilly: do we need someone from ops to help [18:30:13] for chrome, version 11. for firefox, version 4. [18:30:32] well. it was there in 3.5, but broken in 3.6 [18:30:34] tfinc: probably yes [18:30:44] tfinc: sounds like Ryan is our only option as well [18:31:40] nobody on the whole ops team knows how to manage the labs infrastructure besides ryan ? [18:33:18] tfinc: well, at least the IP related stuff [18:51:42] yuvipanda_, ready to discuss the geolocation stuff? [18:51:56] MaxSem: anytime :) [18:52:15] MaxSem: IRC or do you want to skype? [18:53:50] IRC, I think [18:54:33] MaxSem: so basically we're looking to skip geonames and directly query an API ourselves [18:54:38] it's all that simple: you do a request to api.php?action=query&list=geosearch&format=json&gscoord=25.10|37.68&gsradius=1000 [18:54:46] awesome [18:55:07] MaxSem: when can you get it live in labs someplace? [18:55:53] the result will be like that: http://dpaste.org/GrF3o/ [18:56:17] first, I need to restore my access to labs instances [18:56:40] then, I'll start a dump import [18:56:52] MaxSem: all of enwiki? :D [18:57:02] why not? :) [18:57:12] haha [18:57:16] :D [18:57:28] why sub select when you can get it all [18:57:35] i'm guessing eventually labs should give us access to a slave... [18:57:40] during that time I'll work on finalizing all the basic stuff [18:58:10] FYI, Ryan didn't want to get into the habit of putting much mysql into VMs [18:58:13] slave access is not enough: i"ll need to create a table and join it with other wiki's tables [18:59:21] Reedy, then he should come up with some magic pixie dust that does what I want [18:59:28] heh [18:59:37] He was going to set up some non virtualised DB machines AFAIK [19:00:00] luckily, I don't need old revisions;) [19:00:59] also, I'll need a DB of approximately real size to tune performance [19:01:16] yuvipanda_: amir will be filing the locale bugs very soon [19:01:37] tfinc: cool! I just fixed the default locale one, testing before push. [19:01:56] yuvipanda_, what's the app's workflow on retrieving the coordinates: get a few, then get more or get a lot at once? [19:02:12] MaxSem: batches of 'few'. [19:02:18] meh [19:02:30] MaxSem: not 'gimme 200 articles around these!' [19:02:44] that means that I'll have to introduce paging [19:02:57] MaxSem: oh, wait. [19:03:05] MaxSem: it's usually 'get X articles around point P' [19:03:13] MaxSem: then 'get X articles around point P2' [19:03:17] which I wanted to avoid because with sorting by distance from the needed point paging is slow [19:03:29] currently there's no way to say 'get X+10 articles around point P' [19:03:43] keep in mind that tablets will use this eventually and the'll surface more results [19:03:53] tfinc: true, but you don't need 'paging' [19:04:05] tfinc: you just need to have a larger single request. [19:04:34] that is, I'm fine with api.php?....&limit=5000 [19:04:52] yeah i don't think we'll need more then 5000 per tablet location :D [19:05:38] but if you try to do 100 times api.php?....&limit=50&offset=N it will retrieve and sort full data set for every request [19:06:42] hehe paging. [19:07:05] MaxSem: so yeah i think if I can do requests for X places, where X is reasonably large, I don't think I'll need paging [19:08:55] also, I'm working on a prop=coordinates module [19:09:12] MaxSem: also, would the api allow me to tell what 'nearby' means? [19:09:20] MaxSem: like, '100km' or '1km', etc? [19:09:29] MaxSem: would be useful for doing requests based on zoomlevel [19:09:34] yes, via the radius parameter [19:09:45] cool :) [19:09:49] currently capped at 10km [19:10:22] MaxSem: is that a hard limit or a soft limit? [19:10:40] also, by default it is 'sorted' by 'nearness'? [19:10:55] Configurable in LocalSettings [19:11:01] ah, cool. [19:11:08] yes, nearest coordinates first [19:11:49] MaxSem: what kindofa timeline before this gets on Labs? [19:12:00] a week? a month? a couple of weeks? tomorrow? :D [19:12:10] mhm [19:12:45] as I said dump import time and schema finalization [19:12:53] running concurrently [19:13:13] (cause I don't need GeoData to import) [19:13:33] then add {{#coordinates}} to {{coord}} [19:13:48] and runJobJueue till it's empty [19:14:00] *runJobQueue [19:15:03] * MaxSem starts putty [19:15:15] yuvipanda_: https://bugzilla.wikimedia.org/show_bug.cgi?id=33518 [19:15:50] uh oh. i'll need to go looking for hebrew keyboard... [19:15:59] philinje: --^ pretty serious bug with hebrew input [19:16:15] tfinc: is amir here? [19:16:19] maxsem@bastion1:~$ ssh mobile-feeds.pmtpa.wmflabs [19:16:19] Permission denied (publickey). [19:16:25] :evil: [19:16:28] oh nevermind, I think I can just respond to that bug and have him test something out [19:16:47] he's on gchat [19:16:52] just log into your @wikimedia account [19:17:31] I am, don't see anyone except you and tfinc [19:17:32] err [19:17:34] you and philinje [19:17:34] wait .. i'm lying [19:17:43] :D [19:17:45] let me pm you his personal gmail [19:17:46] invited [19:18:14] Amir's personal email is in bugzilla [19:18:19] tfinc: invited [19:19:30] yuvipanda_: are you jumping on the Hebrew bug? [19:19:48] philinje: yes, once i'm done testing the one i'm currently on [19:19:58] ok, will assign it to you [19:25:54] philinje: just an update: amir isn't getting his phone until friday, which, uh, two days from now. He can't test until then, and I'll see if I can load hebrew up into my emulator. [19:26:46] ok thanks [19:30:19] Hallo. [19:31:31] aharoni: ohai :) [19:31:37] Is the URL in the channel topic correct? It may be good for cloning, but in a browser it's 404. Maybe it should be https://github.com/wikimedia/WikipediaMobile ? [19:31:47] ah :) [19:33:34] aharoni: nightly builds in /topic now [19:34:07] oh cool [19:42:53] Another thing i noticed when i tested is that in the map view "Redo search in this area" is not translated. [19:45:12] aharoni: known issue; that view isn't web-based and can't (yet) easily hook in to our localization [19:46:28] tfinc & yuvipanda_, also: what kind of points you want? just those in the page head (primary) or any of those mentioned in articles? [19:46:59] aharoni: the maps also wouldn't be... [19:47:09] also, i've never tested this, but does 'near me' show up per-language stuff? [19:47:49] brion: tfinc philinje clicking points in 'near me' always takes you to the english language version... [19:48:04] hmmm that ain't good [19:48:08] it might be hardcoded :P [19:48:19] stick that in bugzilla [19:48:20] because our current geolocation service supports only enwiki? [19:48:36] you can specify a lang parameter to geonames [19:48:37] yes, exactly. [19:48:39] oh [19:48:39] yuvipanda_: --^ [19:48:41] cool [19:48:42] then [19:49:04] countryCode the ISO country code of the artic [19:49:06] from http://www.geonames.org/export/wikipedia-webservice.html [19:49:13] yup am there [19:49:15] will file a bug [19:49:25] excellent [19:49:31] good catch [19:50:00] the keyboard issue isn't just localized to hebrew. i see the same thing in russian [19:50:33] tfinc: even in english, sometimes the searchbar swallows the first few keypresses. [19:50:54] philinje: yuvipanda_ lets make sure to move bugs to assigned when we assign them to people [19:51:06] you'd think bugzilla could do it on its own [19:51:08] but its lazy [19:51:09] tfinc: can you try brion's 'no position:fixed' build with russian and see if that happens? [19:51:15] wah, i thought bugzilla did that :| [19:51:34] nah: https://bugzilla.wikimedia.org/show_activity.cgi?id=33518 [19:51:38] i had to set it [19:51:42] tfinc: http://www.missionmission.org/2012/01/03/day-drinking-just-got-real/ [19:52:27] whatever binasher .. its no juice box [19:52:49] so true. you should have brought some back for the office [19:53:08] ohh .. stocking stuffers [19:53:11] yes i should [19:53:23] tfinc: will do. [19:53:58] in what I call funny, gcal event invitation offers to show me a map for 'IRC' [19:54:17] ROFL [19:54:33] and ofcourse, thousands of locations named 'IRC' turn up [19:54:40] though not sure why they're showing me Europe [19:56:43] tfinc: sorry, was ignoring the assigned setting, now I know [19:59:51] ok .. brion yuvipanda_ ether pad to help us chat about the meet up http://etherpad.wikimedia.org/WikipediaPhoneGapAndroidMeetup [19:59:54] i just started it [20:00:10] so were scheduled to do a meet up at adobe hq 1/23 [20:00:21] on Wikipedia, Phonegap, and Android [20:00:23] it should be fun [20:00:27] as it'll be right after the hackathon [20:00:33] hurrah roller coaster of awesome [20:00:38] :D [20:00:49] 15 jam packed days for me, hope I don't get lost [20:01:06] i'd love to get you guys involved in it of course [20:01:11] okay, i'm going to wait a bit for either of you people to fill in some stuff 'coz i was late to the party. [20:01:16] especially you yuvipanda_ as you'll be far more technical then i can be [20:01:30] as your deeper in the code now then i am [20:01:33] so [20:01:39] lets discuss about what points we want to hit [20:01:42] for the talk [20:01:49] tfinc: sure! [20:02:12] tfinc: can you fill in on some history? the 'why' part? [20:02:39] sure, yuvipanda_ name yourself in the pad or we'll do it collaboratively for you [20:02:40] ;) [20:02:47] and you don't want to see what preilly comes up with [20:02:51] ;) [20:02:55] there we go. [20:06:43] yuvipanda_: brion : do note the "= Pitfalls of Doom =" section [20:06:45] i'm sure we have some [20:06:55] :D [20:07:01] MediaWiki sucks? [20:07:03] Is that in there? [20:07:06] And PHP, ofc [20:07:31] wah, my highlight is pink. [20:07:42] yuvipanda_: so change it [20:08:10] TIL [20:08:28] there you go :) [20:09:19] this is coming along nicely [20:09:46] yuvipanda_: how would you feel about presenting on the pitfalls section ? i'll create a skeleton talk and then we can iterate on it [20:10:18] tfinc: sure can do, unless brion wants to do it. [20:10:30] brion: also, WP7? [20:11:01] ah wp7 will be fun [20:11:21] wp7 mobile so you mean <1% of internet traffic ? [20:11:23] :D [20:11:24] brion: C# is an amazing language :) [20:11:33] zinggggg [20:12:18] yuvipanda_: lets keep it as the two of us for now. IF you can't make it to the states then brion can fill in (if his schedule permits) [20:12:30] tfinc: cool. [20:12:57] are there any big high level points that you guys want to talk about that aren't there yet ? [20:13:32] oh .. we should mention that someone already forked it and added basic photo upload [20:13:33] done [20:14:24] brion: yuvipanda_ : any other high level points? [20:14:29] i'm sure we'll have more low level ones [20:14:58] i'll talk with philinje about what we should mention front the mobile roadmap [20:15:01] tfinc: Nothing I could think of, other than android's font issues. [20:15:47] oh … what about a .. phone gap made this easier then iOS/java section ? [20:16:41] DEFINITELY [20:16:59] sorry, I meant to say D-E-F-I-N-I-T-E-L-Y [20:17:10] ok i added a "= Rays of Phonegap/Callback/Unicorn/Cordova sunshine =" [20:17:13] section [20:17:22] ha ha [20:17:27] unicorns farting rainbows, right on cue [20:17:42] brion speaks what was on my mind [20:17:58] :D [20:18:20] committing RobotUnicorn extension now [20:18:30] just don't call the fart method [20:18:31] queue the rainbows [20:18:48] unless you're using reflection [20:19:10] we should write a mongodb driver for it [20:19:15] so, y'know, it becomes webscale [20:19:31] lol [20:20:14] ok .. yuvipanda_ brion .. jot down some notes for the pros of phone gap section and then i can take this an create a first rev presentation [20:20:28] brion: we should mention our upstream changes too [20:20:43] to showcase us as good open source citizens [20:21:26] brion: also, the bugs you reported - is someone working on them upstream? [20:24:22] possiblyt not yet -- the main initial ones were already on track [20:24:44] main remaining one (the about:blank loads in separate browser on ics) has been confirmed by one or two other folks but i've seen no work on it [20:24:49] haven't had a chance to track folks down yet [20:25:09] brion: does adobe have people working on phonegap? [20:25:10] tfinc: ^ [20:25:25] yuvipanda_: adobe bought the team that created phone gap [20:25:26] so yes [20:25:46] well, was it an 'aquihire' where they buy the team, open source the project and let the team work on something else? [20:26:19] i'm a bit vague on the plan [20:26:24] its corporate and mysterious so who really knows. [20:26:27] heh [20:26:36] i bet its a play to get mobile HTML dev talent [20:27:04] adobe loves to make tools [20:27:12] so i can see them making phone gap a central piece of it [20:27:31] tfinc: i'm hoping we could actually talk to some people at adobe about this when we're there? [20:27:38] or ignoring it and trying to do a better proprietary version [20:27:42] or is that a 'community' meetup and just 'happens' to be at adobe office? [20:27:49] philinje: +1 [20:27:52] ok. the ether pad is looking pretty good. let me take couple of days to make a presentation out of it [20:27:57] yup, we'll be doing that [20:28:12] the only other open source 'effort' from adobe i remember is..flex [20:28:14] yuvipanda_: some of the former Nitobi folks are involved with the meetup [20:28:20] philinje: cool! [20:28:37] yeah steve the event host is their community manager [20:28:45] separately, we are trying to talk with the folks at Adobe about their plans for PhoneGap [20:28:56] tfinc: no reply from Brian? [20:28:57] MaxSem: what instance needs an IP? [20:29:04] philinje: sadly none [20:29:08] philinje: let me mail steve [20:29:09] mobile-feeds [20:29:21] Successfully associated 208.80.153.216 with instance ID i-000000c1. [20:29:26] whee [20:29:33] thank you [20:30:25] ouch, it redirects to the canonical server, which is localhost:8080 :P [20:30:59] thanks brion yuvipanda_ for helping draft this [20:31:24] whee [20:31:29] MaxSem: it now has the hostname http://mobile-feeds.wmflabs.org/ [20:32:03] MaxSem: http://mobile-feeds.wmflabs.org/w/index.php/Main_Page works for me [20:32:47] preilly, if you can log in into it, could you change $wgServer? [20:33:07] MaxSem: can you change the canonical server name [20:33:15] food .. bbl [20:33:20] MaxSem: you can't login right now? [20:33:32] I can't even log into Bastion ATM:) [20:35:26] MaxSem: I can't login either [20:35:31] MaxSem: Ryan_Lane is fixing it right now [20:35:35] grrrrr [20:36:42] MaxSem: you can now [20:38:02] Ryan_Lane: ssh mobile-feeds.pmtpa.wmflabs [20:38:02] Permission denied (publickey). [20:38:22] MaxSem: did you forward your agent? [20:38:24] MaxSem: did you forward your key? [20:38:36] let's talk about this in #wikimedia-labs [20:38:36]