| [00:00] <aquarius> might need a little tiny bit of poking in the qml to read out the folder that the qml file is installed in and pass it to the html | |
| [00:00] <aquarius> assuming that a qml file can know its own path, otherwise (horrors) it'll need a *C++* extension, which would be rubbish., | |
| [00:03] <daker> hmm wired it's using the ubuntu app launcher but in the browser-app | |
| [00:04] <daker> it works using* | |
| [00:04] <aquarius> yeah, it won't work in browser because you're not loading from a file url then | |
| [00:04] <aquarius> and you can't CORS file: URLs either | |
| [00:04] <aquarius> chrome has the --let-me-load-file-urls-with-xhr switch (it's not called that, but it's called something like that) precisely for this reason :) | |
| [00:05] <daker> but it's same engine :) | |
| [00:05] <aquarius> yeah, it's a config switch | |
| [00:05] <aquarius> but letting unprivileged web pages load file URLs with xhr would be a hole in the security policy the size of the Moon ;) | |
| [00:06] <daker> maybe i should continue to look at the i18n stuff | |
| [00:06] <aquarius> why would you be loading file URLs with the browser app anyway? | |
| [00:06] <aquarius> if you've got files on the device, then you're a click package, at which point you can use the ubuntu-html5-app-launcher :) | |
| [00:07] <daker> ya | |
| [00:07] <daker> aquarius: i trying to implement an i18n mechanism | |
| [00:07] <daker> using <link rel="prefetch" type="application/i18n" href="locales.json" /> | |
| [00:08] <aquarius> daker, yeah, there are a few of those around. Are you doing it from scratch, or using some existing js port of gettext, or does it back-end onto the qsTr stuff that Qt alrady provides? | |
| [00:09] <daker> scratch but really simple thing | |
| [00:09] <daker> locales.json will have something like : | |
| [00:09] <daker> { | |
| [00:09] <daker> "fr":"locales/fr.json", | |
| [00:09] <daker> "en":"locales/en.json" | |
| [00:09] <daker> } | |
| [00:10] <daker> then depending on the navigator.language select the correct one | |
| [00:10] <aquarius> ah, man, you're inventing your own storage format for translated strings? | |
| [00:11] <aquarius> dpm will cry a million million tears if you tell him that :) | |
| [00:11] <aquarius> there are gettext ports already -- then people get the tools to create translations :) | |
| [00:11] <daker> https://developers.google.com/chrome/web-store/docs/i18n#details | |
| [00:12] <daker> https://github.com/mozilla-b2g/gaia/tree/master/locales | |
| [00:12] <aquarius> I stand corrected | |
| [00:12] <daker> com/mozilla-b2g/gaia/tree/master/shared/locales | |
| [00:13] <daker> https://github.com/mozilla-b2g/gaia/tree/master/shared/locales | |
| [00:13] <aquarius> however, that Google stuff describes how to localise your app's metadata, not your actual app, unless I'm missing something? | |
| [00:13] * aquarius looks at the boot2gecko stuff :) | |
| [00:13] <daker> https://github.com/mozilla-b2g/gaia/blob/master/shared/js/l10n.js | |
| [00:15] <aquarius> omg. I have to ship HTML that looks like <span data-l10n-id="lots-of-work">lots of work</span>? | |
| [00:15] <aquarius> I suppose there isn't really a better way to do it | |
| [00:15] <aquarius> but, ouch. | |
| [00:15] <daker> https://developer.chrome.com/extensions/i18n | |
| [00:15] <aquarius> I apologise for the accusation, though. This looks like the way everyone else is doing it too, yeah | |
| [00:16] <daker> we were just brainstorming (me & Alex), so basically those were one of the method to do it | |
| [00:17] <aquarius> I am a bit annoyed that whoever invented this invented a whole new thing which has *no* tooling rather than using gettext which has loads of tooling | |
| [00:17] <aquarius> but, well, it's invented now. | |
| [00:17] <daker> so i am just testing it, if you have a better idea we can discuss it :) | |
| [00:18] <daker> i am not really sure how gettext will work | |
| [00:19] <aquarius> This stuff totally does not document how to translate your HTML. Obviously the b2g stuff does it by looking for *[data-l10n-id] but nothing actually *says* that anywhere. | |
| [00:19] <aquarius> If you're creating all your strings from JavaScript then you are already doing it wrong. :) | |
| [00:20] <aquarius> need to work out a way to roundtrip the lang.json files into and out of rosetta :) | |
| [00:22] <daker> aquarius: the b2g is only for b2g not for apps | |
| [00:22] <aquarius> really? | |
| [00:22] <aquarius> bah. | |
| [00:23] <GabrieleG> it works really well now! :D | |
| [00:24] <daker> aquarius: https://developer.mozilla.org/en-US/Apps/Build | |
| [00:24] <daker> the localisation section has nothing on it | |
| [00:25] <daker> so there is no official way to do i18n in FF OS | |
| [00:27] <daker> aquarius: http://jsgettext.berlios.de/doc/html/po2json.html hhh :) | |
| [00:27] <aquarius> ya | |
| [00:27] <aquarius> not the same json format, though, although it's not far off :) | |
| [00:28] <aquarius> I was just checking the appdev FAQ for ffos to see if it even mentions translations | |
| [00:28] <aquarius> which it does not ;) | |
| [00:29] <daker> here https://developer.mozilla.org/en-US/Apps/Build | |
| [00:29] <daker> Location | |
| [00:29] <daker> Advice for creating location-based applications. | |
| [00:29] <daker> but nothing in it | |
| [00:29] <aquarius> I imagine that'll be about geoloc rather than languages, though | |
| [00:30] <aquarius> https://github.com/recurser/jquery-i18n is pretty common. jquery, of course, sigh, but the underlying idea and formats might be useful to look at | |
| [00:31] <aquarius> (why, why, why is this a jquery plugin, why) | |
| [00:34] <daker> the issue is how to do translation for the HTML | |
| [00:36] <aquarius> yeah | |
| [00:36] <aquarius> the b2g way seems about the most sensible way | |
| [00:36] <aquarius> I hate the idea of putting data-l10n-id on everything | |
| [00:37] <aquarius> but I hate all the alternatives more :) | |
| [00:37] <aquarius> I'm not at all sure about having to think up unique ids for all your translateable strings, though | |
| [00:37] <aquarius> gettext's way of using the string *as* its ID is better, I think. | |
| [00:38] <aquarius> so you'd just do <span data-l10-translatable>Hello, world</span> | |
| [00:38] <aquarius> and then "Hello, world" would be the ID | |
| [00:38] <aquarius> I do not know how the b2g stuff handles nested elements, though. | |
| [00:38] <aquarius> <p>Hello, <em>world</em></p> | |
| [00:39] <aquarius> Where you really want to translate "Hello, world" as a string, not as two separate independent strings. | |
| [00:39] <aquarius> I figured that someone somewhere would have already solved all these problems, like gettext has for desktop apps :) | |
| [01:37] <vthompson> all, anyone around that wants to create a construct/prototype for the "bottom edge" implementation seen today? From the presentation, blog post, and what I've seen from Shuttleworth on G+, it seems like Canonical may be trying to source the implementation from the community. Maybe we can get a jump start on this for the SDK team? | |
| [01:40] <kerjava> yes we see u | |
| [01:41] <kerjava> your welcome | |
| [01:41] <kerjava> you have a lot of instruments ^^ | |
| [01:41] <sarnold> kerjava: are you sure you're in the right place? I see your half of the conversation but no one elses.. | |
| [01:42] <kerjava> agreed ... | |
| [01:42] <aquarius> kerjava, I think you're watching Jono's Q&A session? The video isn't live -- that was broadcast earlier today | |
| [01:42] <kerjava> i see | |
| [01:42] <kerjava> the timing was uncanny lol | |
| [01:43] <kerjava> latter all ^^ | |
| [02:32] <snapplejacks> the tabbed interface seems kind of clunky, is there a trick i'm missing or does it always require multiple gestures? | |
| === coderz is now known as Guest43300 | |
| [07:09] <Guest43300> Hello, I have a problem when compiling my program. I am trying to (#include <array>), but I reciecve this error: | |
| [07:09] <Guest43300> #error This file requires compiler and library support for the ISO C++ 2011 standard. This support is currently experimental, and must be enabled with the -std=c++11 or -std=gnu++11 compiler options. #error This file requires compiler and library support for the \ ^ | |
| [07:11] <sarnold> make sense so far.. | |
| [07:13] <DanChapman> good morning | |
| [12:08] <sergico> hi | |
| [12:09] <sergico> looking for help on running qt apps on ubuntu touch phone | |
| [12:51] <funky1> hi there, i'm trying to compile a package but having some trouble telling in configure option the right path to a library, can someone have a look and help me out? http://pastebin.com/niVaPY0D | |
| [18:06] <PaoloRotolo> nik90, ping | |
| [18:40] <nik90> PaoloRotolo: pong | |
| [19:37] <FulLinuxonNexus7> Hi All! | |
| [19:38] <FulLinuxonNexus7> Trying to get a version of Linux on my Nexus 7 2013 flo Tablet | |
| [19:39] <FulLinuxonNexus7> Ubuntu touch is ok.. but I'm trying to find something more functional | |
| [19:39] <FulLinuxonNexus7> seems there are solutions for the earlier nexus 7 model but none seem to be working on mine | |
| [19:40] <FulLinuxonNexus7> Any thoughts or suggestions would be glady appreciated! | |
| [19:40] <FulLinuxonNexus7> ShareITStuff@gmail.com | |
| [19:56] <FulLinuxonNexus7> Found something... http://www.mattfischer.com/blog/?p=502 | |
| [19:56] <FulLinuxonNexus7> Kali linux on Nexus 7 2013 | |
| [19:57] <FulLinuxonNexus7> Incase anyone actually reads the chatrooms | |