[00:01:51] i think we should indeed get rid of the toggle transitions. not having them on ICS is actually really nice - it's just cleaner and faster. on 2.3.6 they work fine, but the toggle transition is just a smidge slow and choppy - and it seems like the text doesn't fully render until the transition is done, making it seem even more choppy [00:01:59] i agree [00:06:43] jdlrobson will you be around during PST hours at all tomrorow? [00:06:52] probably not.. [00:07:08] i've noticed sometimes you overlap in the morning [00:07:11] what about weds? [00:07:21] i'd rather push the opera fix when you're around [00:07:42] well tomorrow.. im out in the evening [00:07:42] b [00:07:46] but i can be back this sort of time [00:07:51] for about an hour [00:08:13] could you do a bit earlier, like 3pm pst? [00:08:51] * jdlrobson does timezone math [00:08:52] New patchset: Jdlrobson; "turn off transitions by default" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6895 [00:08:57] oh wait that's no good for me [00:09:05] wtf [00:09:11] wtf? [00:09:14] tfinc: is our 1:1 at 330 tomorrow? [00:09:18] just scheduling confusion :) [00:09:31] ok [00:10:10] awjr: I can be here anywhere between 11pm BST and 1.30am BST - that's 3pm PST - 5.30 [00:10:18] we can push the above as well [00:10:38] ok - jdlrobson, let's shoot for 3-4pm PST (11pm - 12am BST) and i'll clear up my scheduling with tomasz when he's free [00:10:44] k np [00:10:48] if that needs to change i'll shoot you an email tonight. [00:10:51] i'm going to catch some z's now [00:10:56] cool sleep well! [00:30:00] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6895 [00:30:04] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6895 [00:32:24] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6892 [00:32:26] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6892 [00:32:47] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6893 [00:32:49] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6893 [00:55:50] awjr: i'm still in the meeting but move it to whatever works [01:23:24] tfinc: ok i moved the meeting to 5pm tomorrow so i can work with jon to get the opera bug fixes out. also, i added status updates to http://www.mediawiki.org/wiki/Mobile_support_in_MediaWiki_core/status - not sure if that's quite what you were looking for, but i wasn't really sure how else to show progress since most things that have happened so far have been very preliminary and not apparently for end-users. [02:02:59] still some folks around? [02:04:35] preilly? [02:06:09] awjr? [02:06:19] a bit [02:06:20] what's up? [02:06:40] you guys use jquery for search on mobile, right? [02:07:03] i think on devices that support jquery [02:07:15] so iphone/ android [02:07:26] do you know if you have something similar in your code base: [02:07:27] $.ajaxSetup({ scriptCharset: "utf-8" , contentType: "application/json; charset=utf-8"}); [02:08:12] not off the top of my head, but you should feel free to take a look - you can check the MobileFrontned code out from our git repository [02:08:20] most important aspect is the contentType [02:08:30] okay, what file(s) are most suspect? [02:08:42] wait i can grep it [02:08:46] nevermind [02:08:53] just clone it from github? [02:10:01] you can find details here: http://www.mediawiki.org/wiki/Extension:MobileFrontend [02:10:08] thx [02:10:12] np [08:19:27] New patchset: Jdlrobson; "revert bad rule introduced in dbaa7218" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6915 [09:22:36] jdlrobson, how do you like those usertesting videos? :) [09:22:46] yeh they are brilliant [09:22:49] just watching them no [09:33:54] https://gerrit.wikimedia.org/r/6915 is big problem [09:34:01] noticed lots of people commenting about it on twitter :( [09:45:50] sigh, why doesn't everyone have a 4-inch displays and fastt WiFi? [09:46:31] sections collapsed a minute after page seemed to have loaded just killed me [09:47:50] * MaxSem goes to try it on mom's S60 [09:53:51] sigh [09:55:39] MaxSem: > https://bugzilla.wikimedia.org/show_bug.cgi?id=36633 [10:05:15] * jdlrobson goes to run errands [10:43:35] MaxSem: the table issue is really worrying me [10:43:53] is there any way we can fix it in production before this evening [10:44:30] let's ask in #wikimedia-dev [10:45:11] New review: MaxSem; "(no comment)" [mediawiki/extensions/MobileFrontend] (master) C: 1; - https://gerrit.wikimedia.org/r/6915 [10:46:39] now that we're 100% on Git, we should get back to giving ne a deployment banhammer [10:48:10] can't believe this one slipped through the test net [10:49:04] we direly need QA [10:49:16] why does http://bits.wikimedia.org/zh.wikipedia.org/load.php?debug=true&lang=zh&modules=site&only=styles&skin=SkinMobile&* load Vector's css? [10:49:31] I mean MediaWiki:Vector.css [10:51:14] liangent, because Mobile isn't 100% registered, so on this request system assumes default skin [10:51:21] * MaxSem goes to bz [10:55:50] liangent, i've submitted the bug, might poke at it later [10:56:16] MaxSem: maybe someone in another room might be able to help? is there a list of people who are able to deploy on a wiki somewhere? [10:56:36] http://meta.wikimedia.org/wiki/System_administrators [10:56:44] #wikimedia-tech and #wikimedia-operations [10:56:56] not sure that's updated [10:56:57] but they should all be in -dev already [10:57:44] Katie_WMDE, it's dreadfully outdated [10:57:49] yeah [10:58:42] * Katie_WMDE imagines Jimbo doing sysadmin stuff [10:59:32] he used to do that 10 years ago, so why not? [11:04:50] demon is around [11:07:13] could mail him? [11:07:32] yea [11:07:33] s [11:22:49] * jdlrobson has a laptop battery that's about to die [11:35:26] no choice i think but wait for SF people [11:59:20] New review: Dzahn; "per demon and Katie. just remove "display: none;" from table CSS to fix mobile site layout." [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6915 [11:59:22] Change merged: Dzahn; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6915 [12:05:42] hi all [12:06:06] Katie_WMDE: anything us non-SF folks could help with? [12:06:08] mutante: heya [12:06:09] demon just told us (ops) about mobile site breakage [12:06:19] YuviPanda: yeah, make sure the fix is good [12:06:22] he couldnt merge the fix, is in final exam or something [12:06:28] so i merged that just now [12:06:34] * Katie_WMDE does not want to be responsible for breaking things [12:06:37] okay, I just came on so I've no idea what you guys are talking about [12:06:43] but i don't know about the deployment part now, after merging it in gerrit [12:06:51] MaxSem ^ [12:06:58] * YuviPanda goes to check logs [12:07:07] YuviPanda: https://gerrit.wikimedia.org/r/#/c/6915/ [12:07:14] there was "display: none" in CSS on table [12:07:18] now it is removed again [12:07:33] it said that was a clear mistake that it was ever in there [12:08:04] Katie_WMDE just showed us this example: http://en.m.wikipedia.org/wiki/Motorola_6847 [12:10:27] ah, right. [12:10:43] https://bugzilla.wikimedia.org/show_bug.cgi?id=36627 [12:11:03] <-- looks like this is not actually resolved [12:34:41] mutante, the only thing I know is http://wikitech.wikimedia.org/view/How_to_deploy_code [12:42:34] MaxSem: yeah..hm,,,bugzilla says a deployment is planned for today anyways, i would leave it at that [12:43:11] that made somebody call the ticket resolved already, well reopened it, i don't know when today though [12:43:49] heh, if you haven't delpoyed code before, now it's probably not quite an emergency for a crash course:P [12:43:56] so, it's patrick and arthur who do it? [12:44:06] i am guessing just a couple hours more [12:44:19] dont expect them to do it late in the afternoon in US time zone [12:45:03] * Katie_WMDE reads http://www.mediawiki.org/wiki/Extension:MobileFrontend/Deployments [12:45:14] MaxSem: no i haven't. i just happened to be in this because demon asked us to answer questions as he wasnt able to get to a keyboard / on the server right now [12:46:32] ^demon will be back in a few hours also [12:46:49] yeah. pretty sure he will be back asap [12:48:26] Katie_WMDE: or Roan [12:48:37] yes [13:11:12] jdlrobson: one look at https://github.com/yuvipanda/WikipediaMobile/commit/97e4116af30b86607ccf48738e16781796f4d19c for CR? [13:11:15] single commit only [13:11:32] sure [13:12:13] what a wonderful thing credit card fraud is [13:12:14] xhr is just a string ? [13:12:14] sigh [13:12:20] you've had credit card fraud ? [13:12:33] jdlrobson: it seems to be just a string when there is no network [13:12:43] jdlrobson: and an xhr object when there is 'other' error [13:12:47] jdlrobson: yup [13:12:53] jdlrobson: on my GSoC 2011 card, none the less [13:13:02] about 350$ of telephone usage in Mexico [13:13:14] so I now need to get a new card shipped to Google HQ and have them ship it to me [13:14:12] :( [13:14:17] ahh man [13:14:31] ok yuvipanda [13:14:37] CR = looks good [13:14:52] so about 30 mins spent on the phone with Citi today - at least it was free via skype. International calling would've killed me [13:15:00] jdlrobson: cool, I'll just go ahead and merge that pull req then :) [13:15:38] jdlrobson: done. Thanks :) [13:15:44] YuviPanda: thanks [13:15:46] [WikipediaMobile] yuvipanda pushed 5 new commits to master: http://git.io/hZ2DFg [13:15:46] [WikipediaMobile/master] Show error page when retrieving page data - YuviPanda [13:15:46] [WikipediaMobile/master] Handle network failures during searching properly - YuviPanda [13:15:46] [WikipediaMobile/master] Removed extranous console.logs - YuviPanda [13:15:55] jdlrobson: I guess table: none explains why you weren't seeing infoboxes in the app yesterday :) [13:16:01] yehh [13:16:06] i'm kicking myself for not making the connection [13:16:10] horrible horrible horrible [13:16:48] a fairy inside me dies everytime someone tweets something along the lines of why is wikipedia not showing me pictures [13:16:54] hoping someone can deploy the fix soon [13:16:56] soon enough [13:17:05] in a way it was useful [13:17:13] as it made the user testing less easy [13:17:21] and we got some useful information from it [13:17:28] (usually people find the answers in the info box) [13:17:32] YuviPanda, did you use your card online often? [13:18:51] YuviPanda: you might like localphone.com [13:19:13] MaxSem: for a year [13:19:16] MaxSem: and locally too [13:19:31] MaxSem: a day before the fraud happened, I used it to pay for my servers at prgmr [13:20:22] mutante: I wasn't using it for telephony services. I looked into my card today and found it had only $17 in it, rather than the ~$300 I expected [13:20:30] mutante: (This is a Google Summer of Code prepaid card) [13:20:49] YuviPanda: we should start getting some tests into the app after this next release - lack of tests is killing us imo [13:22:26] jdlrobson: qunits? [13:22:32] yup [13:23:11] jdlrobson: agreed. [13:23:47] jdlrobson: Unsure how to go about it though - I'll be jumping into WLM app right after this. [13:24:02] YuviPanda: just saying when i saw you saying "international calls would've killed". it gives you a local "alias" number for an international number you can call from phone without a laptop. saved me when i had to call airline in foreign country to get a ticket rebooked [13:24:14] At least that one I can start from ground up with tests + proper jslinting. [13:24:18] mutante: aaah, thanks! [13:24:58] YuviPanda: they have different products, but this is "local numbers" .. see the 1 2 3 on http://www.localphone.com/services/local_numbers [13:25:13] mutante: yup, was looking at it [13:25:20] rates lower than skype..ack [13:25:27] mutante: I'll use this next time I need to make a call that is not toll free [13:25:34] mutante: citi's number was, thankfully, toll free :) [13:27:04] yea, i had this ticket from Malaysia airlines and their number was toll free from "within Malaysia" but i was never in the country, they were just my airline [13:27:11] haha [13:27:23] I guess this might come in useful for German/EU things now [13:27:28] mutante: thanks! :) [13:27:46] yes, i thought so. but gotta check which country combinations work [13:29:37] New patchset: Jdlrobson; "add support for browsers based on ie8" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6924 [13:29:38] New patchset: Jdlrobson; "clearly define variables" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6925 [13:40:29] hi YuviPanda, do you know whether you guys have something like this [13:40:30] $.ajaxSetup({ scriptCharset: "utf-8" , contentType: "application/json; charset=utf-8"}); [13:40:46] in the frontend when doing an api / open search call? [13:42:23] i did a quick grep but couldn't find it myself [13:42:57] the reason i am asking is that it I am finding "application/json; charset=utf-8" in the server logs and that extra space is causing a lot of headache [13:43:04] New patchset: Jdlrobson; "move padding from absolute to relative element (bug 36618)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6926 [13:43:05] New patchset: Jdlrobson; "hide back links in reference overlay" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6927 [13:43:12] YuviPanda: > https://gerrit.wikimedia.org/r/6927 [13:43:12] drdee_: IIRC no, we don't [13:43:45] jdlrobson: thanks :) [13:44:04] YuviPanda: I wonder what piece of code is responsible for doing this [13:44:09] drdee_: perhaps webkit is doing that by default? [13:44:35] drdee_: search.js in the app, beta_opensearch.js on MF [13:44:38] jdlrobson: ^ [13:46:26] drdee: able to share a dump of the logs? [13:47:10] we don't use ajaxSetup anywhere in MobileFrontend extension [13:48:00] it's not about ajaxSetup, it's about setting the contentType to application/json; charset=utf-8 [13:48:24] I don't think we access json anywhere in the MobileFrontend extension [13:48:26] maybe jQuery does that by default? [13:48:33] * jdlrobson googles [13:48:33] that's also possible [13:48:44] jdlrobson: the opensearch API call is not json? [13:48:53] nope.. using xml at moment [13:49:03] but it also happens to text/xml [13:49:07] has the same problem [13:49:13] jquery ajax default ajax call is Default: 'application/x-www-form-urlencoded' [13:51:26] it doesn't seem to me that the nginx / varnish servers modify this, it seems to come from the frontend, or not? [13:54:14] I'm not sure what could be causing this [13:56:25] who might know more about this? [13:59:14] hi jdlrobson [13:59:20] hi Katie_WMDE [13:59:35] drdee_: maybe MaxSem ? [13:59:37] ^demon can help with deploying if it's urgent enough [13:59:47] there's a deploy this afternoon SF time i understand [14:00:09] <^demon> I love being volunteered for things ;-) [14:00:24] well basically ^demon the mobile site is not showing **any** tables [14:00:37] <^demon> Yeah I gathered that much. [14:00:38] thanks jdlrobson! i'll ask MaxSem (is he PDT?) [14:00:54] MaxSem should be around now [14:01:14] drdee_, are talking about request or response headers? [14:01:25] ^demon: have you ever done a MobileFrontend deployment? [14:01:57] request headers because it shows up in the server log [14:01:59] jdlrobson, it shouldn't be any different from any other deployments [14:02:20] <^demon> jdlrobson: Unless there's something magic and different from a normal extension's deployment, I've done my share :) [14:02:38] drdee_, I'm a backend guy [14:02:38] well it should be just a case of applying the patch to the css [14:02:54] and then refreshing the caches [14:03:18] no refresh needed, it's RL [14:03:20] the patch is just the removal of a line > https://gerrit.wikimedia.org/r/#/c/6915/ [14:03:27] ok well then it's just applying that patch then :) [14:03:41] * MaxSem looks at backscroll [14:03:52] * Katie_WMDE wonders how it got broken [14:04:10] Katie_WMDE: how tables disappeared? [14:04:21] <^demon> Checking out 1.20wmf2 so I can update it. [14:04:22] yeah [14:04:32] that would be me.. :-/ i was doing some debugging and some how committed that line - it was a big diff and i must have missed it [14:04:37] no cookies for me this week [14:04:52] what kind of testing is there for MF? [14:04:56] (although to be fair wasn't noticed in code review :)) [14:05:11] just unit tests at the moment [14:05:13] MaxSem: as a backend guy, why do a lot of the mobile server requests have as mime type a charset append to it? like "application/json; charset=utf8" for example [14:05:16] i'm not sure how best to catch css stuff like this [14:05:29] yeh me neither -selenium would cover it somehow - but we'd need real content [14:05:29] * Katie_WMDE is learnign testing for wikidata [14:05:49] ie. test: go to this page check table visible [14:05:56] yeah [14:06:44] drdee_, at a first glance, MobileFrontend's JS doesn't set any content headers (tested with Firebug) [14:06:54] i'm hoping when I move to the SF office I'll be able to grab someone and get something setup - e.g. a dev labs instance running selenium tests [14:07:09] MaxSem: yeh that's what I thought [14:07:15] drdee_: could it be coming from a 3rd party app? [14:07:18] * MaxSem wonders if squids, varnishes of other insane stuff we use add one [14:07:43] jdlrobson: no it's coming from WikipediaMobile [14:07:54] that sounds like the app [14:07:57] drdee_, do you have an example request logged, or there's no complete information? [14:08:06] got plenty of examples [14:08:26] indeed it does [14:08:31] * drdee_ preparing anonymized examples [14:08:48] drdee_: thanks [14:09:44] ^demon: let me know if you need any help [14:10:33] <^demon> Can you do a sanity check on gerrit changes 6928 and 6929 for me before I merge? [14:15:20] <^demon> jdlrobson: ^? [14:15:29] sure thing [14:16:28] [WikipediaMobile] yuvipanda pushed 3 new commits to master: http://git.io/EtXbSw [14:16:28] [WikipediaMobile/master] Updated to latest MobileFrontend - YuviPanda [14:16:28] [WikipediaMobile/master] Removed hacks.css temperoarily - YuviPanda [14:16:28] [WikipediaMobile/master] Updated CHANGELOG and version numbers - YuviPanda [14:16:46] looks ok to me - MaxSem can you triple check ? ^ [14:17:12] <^demon> Jenkins is claiming merge conflict. Wonder if I missed something. [14:18:12] jdlrobson, MaxSem: check pastebin.com/W2ggQD8g (first 2 hits have charset set, third one not) [14:18:24] ^demon, LGTM [14:18:40] <^demon> Hmm, I think jenkins is just being stupid. [14:19:43] ^demon, if you're updating to head you'll need to increment $wgMobileResourceVersion like in old good times:) [14:20:06] <^demon> Hmm. It seems to be stuck in merge pending. [14:20:48] iPhone app 3.1.... [14:21:19] * ^demon whacks gerrit with a 2x4 [14:25:35] hmm, that is from the iPhone app [14:26:09] there is no code that explicitly sets that, however [14:26:17] drdee_ YuviPanda what version of jquery was that using? [14:26:38] drdee_: 1.7.1 [14:26:39] <^demon> jdlrobson: Well, nobody's deploying this revision. Something's up with gerrit. [14:26:52] ^demon: :( [14:27:12] jdlrobson: ^ [14:27:35] thanks for trying ^demon.. i guess it will have to wait till later today [14:27:47] <^demon> I'm asking #gerrit for support ;-) [14:28:33] YuviPanda: so it this new behaviour with query 1.7.1? [14:29:16] drdee_: it is possible. I'm unsure. [14:29:22] drdee_: I can check in a bit. [14:29:33] no idea drdee_ .. but looks like jQuery might be doing something to do that ... either way.. if people don't upgrade their app that's going to continue regardless of what we do [14:29:50] * YuviPanda greps through http://code.jquery.com/jquery-1.7.1.js [14:30:48] I see nothing unusual [14:31:29] f ( s.data && s.hasContent && s.contentType !== false || options.contentType ) { [14:31:30] maybe? [14:31:50] oh no that's to do with ajaxSetup [14:32:13] I don't see contentType being set anywhere other than the default [14:32:35] and no mention of the string 'utf' at all [14:33:10] so weird [14:34:16] yeah, we seem to be using that phrase very often in this channel [14:34:16] sigh [14:34:21] from the examples, it only happens with http traffic and not https traffic [14:34:46] drdee_: http traffic is *only* from the app on iOS 4.2, and nowhere else [14:34:49] (if that helps [14:34:51] ) [14:36:34] so is this caused by iOS 4.2 (because if you look at the pastebin then you see it 2 GET requests, one API and one regular article and they both have charset appended [14:38:36] drdee_: possible [14:39:14] drdee_: I don't have access to a 4.2 device or sim now, so can't test that with wireshark :( [14:40:15] Thehelpfulone: new beta! http://dumps.wikimedia.org/iOS/Wikipedia-v3.2-beta2.ipa [14:40:19] i'll be mailing it out shortly [14:40:54] ok :) [14:41:00] I can try it on 5.1.1 [14:41:12] Thehelpfulone: wonderful :) [14:41:29] Thehelpfulone: testing saved pages + scrolling would be awesome :) [14:41:35] ok [14:41:41] i'm going to grab dinner, brb [14:43:23] nah, it also happens on Android [14:44:28] * jdlrobson must go - doctors appointment - be back soon [14:58:08] hi guys! [14:58:25] i'm working with diederik to figure out why mobile requests are including the charset in the content type field of access logs [14:58:44] wanted to ask you some Qs about how you guys develop [14:58:45] hey ottomata [14:58:55] particularly about how code locally and test [14:59:00] hiya Yuvi! [14:59:08] ottomata: MobileFrontend (web) or the apps? [14:59:08] do you have a standard setup you all use? [14:59:12] ummmm [14:59:16] i think web [14:59:21] or either maybe? [14:59:22] MaxSem: ^ [14:59:35] do you guys make up your own apache and whatever else configs locally? [14:59:43] ottomata: Extension:MobileFrontend is pretty easy to setup, and IIRC you can use useformat=mobile locally to test [14:59:44] or is there some standardization that puts your configs closer to what might be in prod? [14:59:45] yes [14:59:59] ottomata, what do you want to test? [15:00:13] i want to find out why the content-type header looks like this in the logs [15:00:19] Content-Type: text/html; charset=UTF-8 [15:00:27] and i'm wondering [15:00:36] if you guys use the same apache configs [15:00:38] that production uses [15:00:40] while you develop [15:00:47] * MaxSem doesn't [15:00:48] if so, then I can look at that and we can test locally [15:01:01] * YuviPanda doesn't either [15:01:05] ha [15:01:12] so you make up your own setup, right? [15:01:16] install apache, [15:01:18] modify sites [15:01:19] etc. [15:01:20] ? [15:01:31] ottomata, production has extremely tricky setup [15:01:42] including a Squid extension [15:01:55] so nobody has this setup exactly [15:02:04] yeah, wouldn't need exact setup, but a version that was close to prod would be good, no? [15:02:12] including all the proxies/caches [15:02:21] anyyyway, ok, i got my answer :) [15:02:26] will go bug ops now [15:02:31] thank youuuuuuuuuuu! [15:09:19] Thehelpfulone: able to test? [15:09:47] yep just grabbing that cable to download and sync [15:10:08] awesome [15:15:18] MaxSem: so, action=mobileview on image pages shows all the image info and license info [15:15:23] but doesn't seem to be showing the actual image [15:15:41] MaxSem: just checked with File:BradPittBAR08.jpg on en.wikipedia.org [15:16:25] was it needed? [15:16:47] sigh, working without specs has its own cons [15:17:05] I just assumed that it was obviously needed. Sorry :| [15:17:17] butt... what use would a image info page be without the image?! [15:18:01] btw, I asked you to look at it last week;) [15:18:10] I could try to smuggle it into today's deployment [15:18:17] I looked at the html, it seemed fine :P [15:18:21] I didn't look hard enough. [15:18:39] but what width should the image have? [15:19:47] MaxSem: hmmm, I guess is just the default image... [15:19:57] default image size thumbnai, I mean [15:19:59] wassdat? [15:20:20] which is 180-220 pixels... [15:20:28] paste fail [15:20:29] i meant [15:20:37] http://en.wikipedia.org/wiki/File:BradPittBAR08.jpg and http://en.m.wikipedia.org/wiki/File:BradPittBAR08.jpg [15:21:11] this one is too small itself [15:21:15] how about http://en.wikipedia.org/wiki/File:Nikolic_Tomislav.jpeg [15:21:31] it's huge for mobil3e screens and bandwidths [15:22:27] hmm current mobile site doesn't do anything special either [15:22:50] MaxSem: I suppose for a start we just push out the default, and then open a thread on the mailing list about how to handle this? [15:22:56] MaxSem: I'd suppose it would have to be different per image [15:23:08] should it be an app setting? [15:23:45] MaxSem: about? [15:23:50] image size [15:23:51] size of images to display/ [15:23:51] ? [15:23:53] doubt it [15:24:35] It's either per-image, or perhaps we request for different sizes based on the screen's size [15:24:45] sigh, moar copypasta from core [15:25:11] can't refactor? [15:25:17] ah, backwards compat cabal? [15:27:31] cat CHANGELOG [15:27:33] grr [15:37:33] time to transport self elsewhere [15:37:35] brb [15:40:01] YuviPanda: I'd added one item to my saved page list [15:40:07] I click on the X in the top right [15:40:11] it opens the confirmation up twice [16:03:52] ^demon: any luck with gerrit? [16:04:01] <^demon> Nope. [16:04:10] :( [16:05:11] gerrit is so gerrit [16:06:13] srikanthlogic: are you able to replicate https://bugzilla.wikimedia.org/show_bug.cgi?id=36407 ? [16:06:34] * srikanthlogic checks [16:06:36] hmm, no aharoni [16:06:39] * YuviPanda checks #mediawiki [16:11:19] jdlrobson: while I do that, another feedback [16:12:06] It might just be my phone, but when I clicked section expand on today FA, it took a long time, when I expanded second section I got a memory full error [16:13:09] jdlrobson: and latest Opera Mobile is 12, 9.8 was around a year back IIRC [16:13:31] jdlrobson: on a shorter article, I cant reproduce it [16:14:38] :/ [16:15:53] jdlrobson: suggest you deliver old skin for old phones :) [16:16:40] S60 / S40 will have trouble loading big pages with js on their small memory [16:19:48] yeh... we need dynamic sections [16:20:20] YuviPanda: Android should have a "Share to Bugzilla" option in the camera. [16:20:34] Instant screenshots of bugs! [16:20:37] hehe, maybe I should build that :D [16:21:43] * YuviPanda is tempted to go look up the BZ API for uploads [16:21:54] actually every app should have such feature [16:22:09] srikanthlogic: with Android, such features need not be in the app [16:22:20] srikanthlogic: I just need to build a bugzilla app that exposes this to other apps [16:23:42] YuviPanda: ah yes, even in S60 for that matter. I wish I had gravity supporting bugzilla along with twitpic, yfrog, etc :P [16:25:08]