=== chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === marcusto_ is now known as mr_t_ === mr_t_ is now known as marcustomlinson [05:32] hello [05:33] I noticed that Ubuntu phone's web browser has Android user agent, this creates harsh situation at times where sites try forcefeed Android app instead of showing the site [05:33] a good example is Google Music which I cannot reach at the moment due to this [05:34] is there known way to change user agent without creating some web app klik package or some other workaround? [07:04] good morning === jjj is now known as Guest7150 === chriadam is now known as chriadam|away === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === davidcalle_ is now known as davidcalle === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [08:37] popey, mzanetti: Ping [08:37] sturmflut2: pongs [08:37] popey: Have you seen https://uappexplorer.com/app/neverball.lb and https://uappexplorer.com/app/neverputt.lb? [08:38] I am shaing on twitter/g+ as I type :) [08:38] it runs flawlessly [08:40] popey: I was a bit jealous at first that they got it working before me ;) but I didn't have the necessary time anyways [08:40] sturmflut2, what's with that? [08:41] mzanetti: Two libSDL based games in the store! [08:41] oh! [08:42] mzanetti: With Touch support and tilt sensors and everything! And blazingly fast on the bq! [08:43] that's nice [08:43] yeah, it's really impressive [08:44] https://github.com/pseuudonym404 [08:44] code on github === chihchun is now known as chihchun_afk [08:45] they start _instantly_ too [08:45] popey: Yeah! [08:45] how do they handle the app lifecycle mgmt [08:46] did you check that ? [08:46] yes [08:46] it works as expected [08:46] cool [08:46] i have both games "running" [08:46] http://people.canonical.com/~alan/screenshots/device-2015-05-28-093841.png [08:46] yay [08:47] and a GPS to tell you where to put the ball ! [08:47] apparmor profile just says "audio" and "sensors" [08:47] haha [08:47] ogra_: Do they have to handle the App Lifecycle somehow? Doesn't Unity8 just tell the kernel to stop the app process? [08:47] sturmflut2: there was a problem with OpenAL not getting stopped when playing audio [08:47] guess this app does audio differently [08:47] sturmflut2, well, if they get killed due to OOM and restarted ... [08:47] not tested that [08:48] ogra_: Ah, yes, that would be a problem. [08:48] * sturmflut2 didn't think of all edge cases [08:48] for that case the app needs to do something and save its state [08:48] or you come back to it and it only starts newly, losing your progress [08:48] I'm just impressed how well an SDL game works on Mir [08:48] also, it ships with very few libraries bundled in [08:49] it is gigantic ... still downloading here :) [08:49] 123M in total 118M of which is level data [08:49] app itself is 1.2M and the res is libs [08:49] well, i started with neverball, thats only 52B [08:49] 123 unpacked [08:49] oh [08:50] guess the data compresses well [08:50] next step - make a multi-arch version for the intel tablets :) [08:51] lol [08:53] Okay, they use the same approach as I did in the beginning: the .desktop file points to a shell script, which sets the LD_LIBRARY_PATH to the included libSDL2-2.0.so.0 and then runs the binary. That could be simplified, as I have learned from mzanetti's Network Scanner [08:54] sturmflut2, it probably has to for multiarch [08:55] unless the script fiddles with that... [08:55] but the lib folder in click packages exists for every arch and is automatically set up for the proper arch [08:55] wow, really nice on the arale :) [08:56] mzanetti: Yeah, I learned that trick with the lib folder by spying on you [08:57] sturmflut2, you shouldnt need to set LD_LIBRARY_PATH at all [08:57] (as long as you ship your libs in the right location in the click) [08:58] ubuntu-app-launch exports it when it runs the .desktop file [08:59] https://developer.ubuntu.com/en/start/platform/guides/app-confinement/ (see "Runtime Environment") [09:02] so who packages dota2 now ? [09:02] :) [09:02] heh [09:02] so what is the "right" location in the click? [09:02] libs/triple ? [09:02] see the page above [09:02] k [09:03] ./lib/arm-linux-gnueabihf/ [09:03] for arm [09:03] yay [09:04] tested that and it works [09:04] i386-linux-gnu for i386 and x86_64-linux-gnu for amd64 [09:04] will file an issue on github, because that way we can get an x86 build in [09:04] although the main binary is arch specific so that will need work too [09:04] yep [09:04] :) [09:04] well, see PATH in the page [09:05] you can just ship two binaries [09:05] same setup :) [09:05] nice [09:05] (just with /bin) [09:10] fgimenez, brendand: do you think we should go ahead and upload the selenium change to wily already? [09:16] dholbach, not sure, i have this https://code.launchpad.net/~fgimenez/+junk/python-selenium that builds fine with a recipe, but still ships the binaries [09:17] dholbach, zigo proposed bundling the binaries in a separate package and upload that to non-free in debian [09:18] hum [09:19] so that'd be a separate source package or separate binary package? [09:20] Good morning all; happy Amnesty International Day! 😃 === chihchun_afk is now known as chihchun [09:39] dholbach, i guess a separate source package but without compilation, just including the binaries that come in the selenium source [09:40] dholbach, if the debian upload doesn't go ahead we can try also removing the binaries, in this case we should check that this doesn't break anything with previous versions of ff [09:41] dholbach, don't know if shipping the binaries is acceptable in terms of licenses, etc [09:43] as the separate source package will have to go through NEW anyway, it might make sense to upload your patch already :) [09:44] and make the selenium-modules-non-free package (or whatever it's going to be called) conflicts/replaces: python-selenium << 2.45.0~ [09:47] dholbach, yes, for uploading the patch into debian it must be a NMU, we are waiting for a possible response from the original maintainer [09:49] ok [09:49] and uploading it to wily already? [09:51] dholbach, i'm not sure, should it include the binaries? should we split the package in free/non-free (not sure how is this done in ubuntu)? [09:53] fgimenez, I prepared a build in a vm and these seem to be the difference in terms of contents of the binary packages: http://pastebin.ubuntu.com/11408827/ [09:53] do we know if users rely on the precompiled firefox webdriver files? [09:57] ... and if the update will break things [09:57] ? [10:00] dholbach, no idea, we should check that of course, i think elopio mentioned that he was going to try with old ff versions [10:02] fgimenez, ok... I was just asking about wily because it might help to get it out for testing and backport it, so more folks can make use of it for html5 apps [10:03] (it's part of the default html5 app template in the sdk already) [10:03] maybe you can discuss it with elopio [10:03] let me know if you need somebody to upload it to wily [10:06] dholbach, ok thx, i'll ping you back [10:06] rock and roll === MacSlow is now known as MacSlow|lunch === chihchun is now known as chihchun_afk [11:21] question, if I'm on 15.10 on the phone, can I somehow change the channel to vivid+stable? [11:21] and then just magically receive stable updates :p [11:21] that sounded bad when writing it out loud [11:45] jgdx: The "system-image-cli" command on the phone has a "--switch" parameter [11:46] jgdx: I am not sure how it handles this case though [11:48] yeah, might as well just flash it === alan_g is now known as alan_g|lunch === mzanetti is now known as mzanetti|run === MacSlow|lunch is now known as MacSlow === greyback__ is now known as greyback === mzanetti|run is now known as mzanetti [13:10] is the next OTA wily? [13:11] vivid [13:11] no, wily will not come to phones [13:15] kenvandine, jgdx, btw I plan to look at migrating u-s-s to newer Ubuntu Component version (1.2) and replace deprecated listitems by the new listitem widget [13:15] ogra_: out of curiosity, any idea if there'll be a new framework release based on the vivid overlay at somepoint? with wily being skipped it'd mean a year before any new APIs could be introduced otherwise [13:16] we use outdated versions and have bugs and runtime warnings due to that [13:16] seb128, yeah, that would be awesome [13:17] kenvandine, jgdx: I'm probably going to do it in chunks, not all panels together, easier to review/test and trying to avoid conflicts in things that have pending refactoring [13:17] jgdx, you haven't seen any issues with libqofono in silo 36 have you? [13:17] seb128, that'd be sweet! [13:18] kenvandine, other than it not building? or running uss? [13:18] libqofono built fine :) [13:19] oh ok === alan_g|lunch is now known as alan_g [13:19] kenvandine, no, but haven't look specifically [13:19] I'll do that now [13:19] jgdx, did you get to test the reset context stuff? [13:19] kenvandine, yes [13:20] basically i want to change the version revision from 3.1 to 4, because that's the version i want to land in wily [13:20] but if i bump it, and we change the patch at all... it gets annoying [13:23] jgdx, ok, i'm going to upload 0.70-0ubuntu4 to the ppa for wily then [13:23] kenvandine, but [13:23] kenvandine, it was never tested with the change to nuntium or networkmanager [13:24] jgdx, oh... maybe we should actually create a new silo for wily... [13:24] though it is unlikely that a change in libqofono will be needed if that process fails... [13:24] and keep this silo for the overlay/vivid backport [13:24] since we have a proper libqofono for vivid overlay in this silo [13:25] jgdx, in fact, libqofono could just go to wily now right? [13:26] i could just upload it to wily and keep the silo simple for you [13:27] kenvandine, sure [13:27] thanks [13:27] 3.1 seems to work very well from a lot of phonesim testing [13:27] is the ofono reset context changes in wily? [13:28] yes it si [13:28] is [13:28] cool [13:28] kenvandine, yep according to 1338758 [13:33] jgdx, ok, settings is building in silo 33 === JackH1 is now known as JackH [14:16] Elleo, oh, no idea, thats a question for the SDK team i guess [14:17] ogra_: okay, will poke give them a poke and see what the plan is === chihchun_afk is now known as chihchun [15:07] abeato, awe: Can user_data passed into g_ril_send be freed in callback. [15:09] Idea is passing single instance of struct and pass it into multiple calls of g_ril_send, keeping ref count, and when callback is called, check ref count and free it if ref count is 0 [15:09] s/passing/creating/ [15:11] peat-psuwit, if you don't pass in a destroy func, the user_data won't be cleaned up when the callback is processed [15:11] s/destroy func/notify func/ [15:12] well, actually it's type is GDestroyNotify [15:12] and the param name is 'notify' [15:12] so if you don't provide a func, then user_data won't be touched when the response for a sent request is handled [15:13] awe: So it's OK if user_data is freed when the callback return, right? [15:14] if you want to free it in your callback, you can [15:14] but again, to do so, you'd pass NULL as the last arg in g_ril_send () [15:15] the GDestroyNotify parameter is just a convenience mechanism [15:17] awe: Thanks. [15:17] peat-psuwit, np [15:27] Test dc === dandrader is now known as dandrader|lunch [16:41] kenvandine, jgdx, was there any fix you were still trying to land for ota4 u-s-s? === alan_g is now known as alan_g|EOD === CarlosNeyPastor_ is now known as CarlosNeyPastor === dandrader|lunch is now known as dandrader [17:28] hum [17:29] when a device doesn't boot anymore [17:29] (stucked on the ubuntu logo) [17:29] what's the right way to restore it? [17:29] I tried to boot in fastboot mode but and do a u-d-f boostrap but the device says that recovery doesn't support it [17:31] you need a recovery with adb in it [17:32] seb128, nothing else for settings for ota4 [17:32] ogra_, and if I don't have that, what can I do? [17:33] kenvandine, k, you are not trying to land the network and battery graph init fixes anymore? [17:33] seb128, http://askubuntu.com/questions/602035/how-do-i-use-ubuntu-device-flash-with-the-bq-aquaris-e4-5 [17:34] seb128, there is a download link and instructions [17:35] seb128, i don't think they are on the ota4 list [17:35] so waiting [17:35] ogra_, the device in an arale not a bq [17:36] seb128, oh === chihchun is now known as chihchun_afk === oSoMoN_ is now known as oSoMoN === boiko_ is now known as boiko [18:52] With Android M, apps will no longer request permissions while installing it but rather when you start using a specific feature similar to UT. Looks like Google is learning ;) [18:57] nik90, indeed :) [19:11] will telegram go smoother if I clear history? === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [20:43] Hi all [20:51] seb128, did you see the mako thread? [20:51] jgdx, no, about? [20:51] seb128, mako not booting, stuck on logo [20:51] jgdx, no, but I don't own a mako [20:51] I also fixed my arale issue [20:51] if that's what you were replying to [20:53] (wifi died) [20:53] seb128, right, never mind! [20:54] I had to nuke the cache to fix my prob [20:54] jgdx, thanks for replying! [20:54] :) [20:55] Can someone help me to debug recovery segmentation fault? === salem_ is now known as _salem [21:29] if i switch Ubuntu image channel (bq device), will the apps and data be wiped? [21:36] mariogrip: not unless you explicitly use the --bootstrap argument. [21:36] mariogrip: as in --bootstrap will wipe all data... [21:36] nik90: Okay, thanks. i will switch to rc-proposed then. === mibofra is now known as Guest65273 [22:27] Name yogmog [22:39] Hello ! Is there someone knowing how to install ubuntu touch on a meizu mx4 ? === danielg4 is now known as DonkeyHotei