[00:01] so when i built the image im rebooting to recovery, it installs my image and after that i need to flash armhf? [00:01] Yes [00:02] whats the command for flashing armhf [00:02] Do you have CWM or TWRP on your device? [00:03] i have ubuntu CWM [00:03] Then use it to flash the zip [00:04] i will try flashing this one http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/trusty-preinstalled-touch-armhf.tar.gz [00:05] Why not the .zip? [00:05] i just like tar.gz is there a difference? [00:06] den4ik: I think the .zip contains instructions for flashing for the recovery [00:06] i think its just file format. will see [00:35] if I were to install Cyanogen mod on my device (full install), then install ubuntu-touch dev preview, would my phone calls/sms/networking be working? === anonynimity is now known as Crimson_Rogue [00:37] Crimson_Rogue: no [00:37] awweeee :(((( [00:38] was worth asking anyway [00:39] no??? i thought it should work [00:42] den4ik Ubuntu Touch will only work as much as it is stated to work per port [00:45] 21 === Namidairo`bnc is now known as Namidairo [01:18] den4ik: if you install one OS, then wipe it out completely and install the second OS, nothing from the first OS will remain. [01:18] i know thanks [02:15] If there is anyone here from Canada, you might want to try this command out even with the latest version of jellybean. [02:15] https://plus.google.com/+JohnHolme/posts/UMskJydSsTw [02:16] Go to the dialpad/phone. Dial in the this number: * # * #4636# * # * (This will bring up "Phone info"). Tap WCDMA from under "Set preferred network type". Select "LTE/GSM/CDMA auto (PRL)". The 4G LTE network should now be activated. [02:16] that's for the mako users out there living in Canada and using windmobile. [02:17] Before doing it, I was getting 0.4Mbps download/0.4Mbps upload. After doing it, I got 2Mbps download/1Mbps upload. [02:18] That's pretty good for being on 4G LTE AWS BAND4 in Canada. [02:19] I used the speedtest.net app for these results. [02:20] Another word of caution. Don't listen to any windmobile reps at the store, they are not techies. The rep claimed not to have LTE, but the speedtest says otherwise. === duflu_ is now known as duflu [02:46] does somebody has working custom image? [02:48] !devices [02:48] You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [03:06] how long does it take for phablet-dev-bootstrap to download? [03:07] A few hours [03:07] Greattttt.... Thank you TechieElf [03:08] Well actually it depends on your download speed.. I have 2.4mbs/sec and it took less than an hour === chihchun_afk is now known as chihchun [03:11] 58.00 KiB/s [03:12] be back later then... thank you for your help TechieElf [03:13] no problem crimson_rogue === Namidairo is now known as Namidairo`bnc [03:44] hello semone know what function is utilised to get volume buttons value [03:44] ? === anders3408 is now known as anders3408|away === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === chriadam is now known as chriadam|away === iahmad is now known as iahmad|afk [08:41] fginther: good morning! are you still the right person to ask for CI to be set up for launchpad projects? [08:46] bregma, hi === mortal_ is now known as mortal [09:10] good morning [09:12] dholbach: good morning. :) [09:12] hi nhaines [09:13] dholbach: have a nice morning so far? :) [09:14] yep, doing well - I'm at dpm's place today - how are you? [09:14] Doing pretty well! Finally got the Ubuntu Touch emulator to work on trusty... in VirtualBox, haha. So that's a beginning. :) [09:15] an emulator in an emulator :) [09:16] If there's anything I enjoy, it's definitely meta. :) === iahmad|afk is now known as iahmad === b0bben_ is now known as b0bben [09:55] zsombi: what do you think of http://paste.ubuntu.com/6473048/ so i can use a C++ list model in there? [10:55] is ubuntu phone: http://www.ubuntu.com/phone same as ubuntu for android phone? http://www.ubuntu.com/phone/ubuntu-for-android [10:55] no [10:56] Ubuntu for Android = Android + Ubuntu desktop [10:56] Ubuntu Phone = Ubuntu Touch user interface with only enough Android kernel bits to access devices (GPS/sensors) [11:00] + Ubuntu desktop when the thing is done [11:00] true [11:04] Morning All [11:07] janimo, running: ARCHES=i386 CDIMAGE_NOPUBLISH=1 for-project ubuntu-touch cron.daily-preinstalled --live ... lets see what comes out ... [11:07] tsdgeos: that makes sense. I wanted to tight the model to ListModel to make sure we really have a model there, however forgot that it does not allow to use QAbstractListModels... [11:07] ogra_, \o/ [11:08] ogra_, last time I looked at live build IIRC you also had to export some more envvars but hopefully it is cleaner now [11:08] zsombi: oki, https://code.launchpad.net/~aacid/ubuntu-ui-toolkit/c++_models_tabbar/+merge/196518 you go [11:08] like PREINSTALLED and TOUCH [11:08] if it doesnt fall over (i suspect it will though) i'll just re-run it with the publishing enabled [11:08] tsdgeos: the only thing is that the model must have either the title or the tab role defined [11:08] yep [11:08] but well that was already the case [11:08] ogra_, even if it fails it's great to have this started, thanks [11:08] in ListModel [11:08] np [11:09] tsdgeos and in the latest (tab) one's case, the tab must have the title property available [11:09] ah, and there it returned [11:09] * ogra_ goes to look for logs [11:09] zsombi: i can just add a QIdentityModel to my existing model and add the title property [11:09] or just add the role [11:09] that is what i did :D [11:10] tsdgeos: thinking of unit tests... unfortunately the whole logic is in the style, and we still haven't committed to the Ambiance stability... [11:11] zsombi: well it'll break at the unity8 unittests then :D [11:11] and i'll have to fix it [11:12] janimo, http://paste.ubuntu.com/6473301/ [11:12] first hack already bites us [11:12] doing the phablet flash of the ubuntu phone. It says ubuntu update complete. but it seems to stop with a question Rom any flash recovery on boot. Fix? [11:12] tsdgeois: try to add some unit tests to guard it [11:13] tsdgeos: to the MP [11:13] ogra_, are there some dpkg hooks that are failing there? It's not clear to me what the error may be [11:14] tsdgeos: at least have few models and check the roles [11:14] janimo, the mutlimedia -touch packages are a giant set of diversions against the actual qt5multimedia packages [11:14] zsombi: tbh i don't know why i need to do that in my MP [11:14] its extremely ugly [11:14] ogra_, ouch, diversions [11:14] zsombi: the problem existed before my MP [11:15] tsdgeos: I know, but you were the lucky one to note it, and if you change the API then you have to guard it ;) [11:15] ogra_: oh god, I looked at the failure :P [11:15] yeah [11:15] zsombi: great way to make sure i'm not finding any flaws in your api in the future [11:16] tsdgeos: :D [11:16] hehe :) [11:16] tsdgeos: look, if you have changes for a component, you have to cover it with tests, right? [11:17] zsombi: i'll do the tests, don't worry [11:17] done [11:17] tsdgeos: I agree that the model wasn't covered properly before, but now you have the opportunity to cover it at least :) [11:17] tsdgeos: thx :) [11:17] zsombi: but you tell me what you want exactly, i'm not doign the tests you should have done just to get you to comment that this is not what you wanted [11:18] tsdgeos: right... [11:18] i did yes. phablet flash succeeded [11:18] janimo, for more details, the log is now synced at http://people.canonical.com/~ubuntu-archive/livefs-build-logs/trusty/ubuntu-touch/20131125/livecd-20131125-armhf.out [11:18] * janimo looks [11:18] roughly 20 mins to finish entire recipe as described: http://www.ubuntu.com/phone/install [11:19] tsdgeos: so as previously only ListItems would have been there, our side would have been to test the availability of either tab (+tab.title) or title roles. Now you have to check whether an arbitrary model given to the Tabs has these roles or not. [11:19] janimo, err, no ... thats the armhf one obviously [11:19] janimo, i guess syncing will still take a moment [11:19] ogra_, ok, I was wondering why there's no error :) [11:19] heh [11:20] sorry [11:20] tsdgeos: the only problem I see is to test the QAbstractListModel from QML test.. you may need to have a C++ test for that [11:21] zsombi: i'm sorry but not sure what kind of test you want really, you want me just to feed it a ListModel to TabsBar and then what? check that it did not give any warning? or? [11:21] tsdgeos: ListModel is just one model, right? [11:21] tsdgeos: array of JSON objects is another one [11:21] tsdgeos which is similar to ListModel [11:22] yes, i know there are multiple models [11:22] it's what i added [11:22] tsdgeos: then a QAbstractListModel derived on eis the third [11:22] i'm just not sure what the test is supposed to test [11:23] tsdgeos: I was trying to find a way to catch whether the given model will throw error on the TabBar. [11:24] janimo, http://paste.ubuntu.com/6473354/ [11:25] tsdgeos: so like in TabBar onModelChanged you could check whether the model has the roles required or not and throw error... [11:25] ok [11:25] i can do that [11:25] tsdgeos: then you could test whether the given model produces error on the TabBar or not === chihchun is now known as chihchun_afk [11:31] ogra_, what is with ubuntu-touch-generic-initrd ? Renamed to something else? [11:32] no [11:32] that should just bve fine i suppose [11:32] (on i386 at least) [11:32] we'll see once we get that far [11:32] ogra_, it should be added to the i386 list it's currently only in armhf in ubuntu-touch [11:33] first that error needs to be fixed [11:33] well, you dont really need it inside the image [11:33] ogra_, I was just looking at ubuntu-touch and noticed that, while thinking the diverts are in that package [11:33] as long as the binary initrd lands in the android build [11:33] ogra_, ok [11:34] having it in the rootfs is only hellpful for re-generating the initrd on the running system ... not overly important for a start [11:34] (surely something we need to fix at some point, but not as urgent as getting the media stack installable) === Namidairo`bnc is now known as Namidairo [11:35] ogra_, I see nothing in particular in qtmultimedia-opensource-src-touch that is armhf specific apart from a symbol file [11:36] well, the installation paths differ ... [11:37] i guess the diversion logic uses them ... or se are missing Replaces in debian/control or whatever [11:38] s/se/we/ [11:38] ogra_, are there more detailed errors from live-build? [11:39] zsombi: error out via console.error? [11:39] arent these detailed enough ? [11:39] shows pretty clear whats broken .. [11:40] ogra_, not clear to me. Errors encountered, error code 1 [11:40] hardly detailed [11:40] err [11:40] or is that a summary only [11:40] read the log, loook for the dpkg error above [11:41] ogra_, ah ok, I saw it now above [11:41] that just means "hey, some subprocess didnt finish properly" [11:42] janimo, the package has .presinst files that create the diversions [11:43] and it does not have any Replaces lines in debian/control [11:48] janimo, can you confirm bug 1254686 [11:48] bug 1254686 in qtmultimedia-opensource-src-touch (Ubuntu) "package is uninstallable on i386 due to file conflicts" [High,New] https://launchpad.net/bugs/1254686 [11:49] rsalveti, ^^^ [11:52] ogra_: that's for sergio, the fun with diverts [11:53] that's why we need the mediaserver [11:53] zsombi: do you have a way to check there's been an error printed in console.log for the tests? [11:53] console.error i mean [11:53] rsalveti, well, for now we need an installable package frst :) [11:54] i'll assign it to sergio then [11:54] ogra_: yup, he created the diverts for that package, might be adding something specific for armhf then [11:54] ogra_: janimo: but yeah, there's nothing really specific to armhf in that one [11:54] well, we want it to function the saame on x86 [11:54] rsalveti, indeed that's what I see too === _salem is now known as salem_ === MacSlow is now known as MacSlow|lunch [12:06] ogra_, I wonder what the differences between i386 and i686 are and whether the tools consistently pick the same [12:10] janimo, i386 is the arhcitecture name for i686 [12:11] i wonder what _ARCH_ expands to ... looking at the prinst [12:11] *preinst [12:11] ogasawara, maybe DEB_TARGET_GNU_TYPE should be used [12:11] ogra_, placeholder, sedded by rules [12:11] zsombi: actually, scrap that, i don't know how to do the test you suggested since i can't access the role names of the models [12:12] while i386 is a valid cpu-target in dpkg, in ubuntu it is set to target i686 instead. see DEB_HOST_GNU_CPU=i686 [12:12] zsombi: so i would have to instantiate the model myself and check one of the delegates for the properties, but what if the model is empty? [12:12] (GNU_TYPE is also i686-linux-gnu, the install location for multiarch however is compatible with debian arch name i.e. i386-linux-gnu) [12:14] right, i think that replacement is wrong on i386 ... while it seems to work fine on armhf [12:15] sed "s/__ARCH__/$(DEB_HOST_GNU_TYPE)/g" -i \ [12:15] debian/libqgsttools-p1-touch.preinst [12:15] thats the code in question ... [12:16] Adding 'diversion of /usr/lib/i686-linux-gnu/libqgsttools_p.so.1 to /var/lib/libqgsttools-p1-touch-diverted/libqgsttools_p.so.1.diverted-libqgsttools-p1-touch by libqgsttools-p1-touch' [12:17] i assume we would want that dir to be i386 here === chihchun_afk is now known as chihchun [12:17] * ogra_ has no plain i386 installs to verify this === chihchun is now known as chihchun_afk [12:18] on my amd64 installs it is always i386 ... not i686 though [12:22] janimo, i think DEB_TARGET_GNU_TYPE only expands to i386-linux ... missing the trailing -gnu [12:39] salem_, ping [12:39] om26er, pong [12:40] salem_, my phone network only appears when wifi is turned on and connected to a wireless network (on mako). Do you think its something you can look into ? [12:40] image 28 [12:51] * Mirv dropped qtsensors5-private-dev, no remaining reverse dependencies [12:51] so following Debian === b0bben_ is now known as b0bben [13:16] boiko, salem_: do we support the SIM phone book in the UI? ofono supports it (and so does phonesim) [13:16] boiko, salem_: I can't find it anywhere in the UI, but want to make sure [13:16] om26er, sorry the delay. you mean the indicator only appears? or the network itself? [13:17] pitti, we dont support it yet as far as I know. [13:17] pitti: nope, we don't. [13:17] salem_, boiko: OK, thanks for confirming; ofono only supports importing them, but not changing them, FTR [13:17] pitti: we discussed that long ago, but it ended up not being implemented. I think the idea is just to import the addressbook from the sim [13:18] pitti: I guess that would be enough [13:18] yes, makes sense [13:18] salem_, the network indicator only shows wifi icon and calling my Ubuntu phone in this case does not work neither can I call. Only once I connect to wifi a second icon for sim signals appear in the top panel [13:18] and I can make calls/sms [13:18] boiko, salem_: sims are too limited for modern rich contacts anyway, but import sounds nice [13:18] if I disconnect wifi again, I don't receive any calls even though the caller on the other end seems to hear the ringing tone [13:18] boiko, salem_: BTW, do you know a rough estimate for the PIN unlock UI? [13:19] pitti, yes, I think I did a quick research some time ago around this topic. It is possible to import them, but I am not sure if ofono-ril currently implements the addressbook interface. === alan_g is now known as alan_g|lunch [13:19] pitti: hmm, that's not something we are working on, it was decided this was going to be implemented outside the dialer and messaging apps [13:20] pitti: not sure who is working on that [13:20] salem_: no, it doesn't; it only works with phonesim, not with ril [13:20] boiko: yes I know, it doesn't make sense in the dialer; it needs to be asked at boot [13:21] pitti: I guess awe would be the right person to ask about that, but he is not online right now [13:21] om26er, ouch, it seems this problem lives in the backend. probably not related to apps or telephony-service. I would suggest asking tony about it. [13:21] rsalveti: do you know anything about PIN unlocking UI? [13:21] salem_, ok. will ask him [13:22] om26er, ok, thanks. [13:22] boiko: ack, thanks [13:22] boiko: hm, nops [13:23] yeah, wait for awe [13:24] om26er, hey, boiko reminded me about a setting that might be related to this issue. can you run this command as phablet? "dconf read /org/gnome/empathy/use-conn" [13:28] rsalveti: thanks [13:30] salem_, returns false === MacSlow|lunch is now known as MacSlow [13:32] janimo, so i just did a test build here ... using DEB_TARGET_GNU_TYPE gets me an empty replacement :( === dandrader is now known as dandrader|afk [13:36] aha, because /usr/share/dpkg/architecture.mk doesnt export "TARGET" stuff [13:36] om26er, ah, it is correct. so the problem is probably somewhere else. [13:38] * ogra_ adds the appropriate snippet to debian/rules and re-tries === dandrader|afk is now known as dandrader [13:48] where can I find trusty-proposed changelogs? [13:49] randomcpp, http://people.canonical.com/~ogra/touch-image-stats/ [13:49] thanks ogra_ :) [13:50] the proper mapping of the versions can be found by looking at http://ci.ubuntu.com/smokeng/trusty/touch/ [13:50] (first is the system image number, second is the rootfs (which is the one you want)and the last one is the android build) [13:52] thanks again :) [13:56] pete-woods, yes, I can still setup projects under CI for you [13:58] mhall119: in general, impossible (most .debs have all kinds of assumptions such as paths that won't work in the extremely constrained environment of click packages); in some specific cases it can be made to work and sergiusens was doing something like that for core apps for a while. alien has a MUCH easier time of it because it doesn't have to rearrange the file structure on disk. [14:00] mhall119: basically anything more complex than a single self-contained executable will probably be intractably difficult [14:01] xnox: would be nice, before uploading the change, to also close the MR (and possibly merge it as well): https://code.launchpad.net/~rsalveti/ubuntu/trusty/qtbase-opensource-src/android-emulator-workaround/+merge/195622 [14:01] fginther: I forgot that you were on a different time zone! basically it's a newish launchpad project (https://launchpad.net/unity-voice) [14:01] the fix was uploaded but the mr is still open [14:03] pete-woods, I should have it added today, will send email when it's done. [14:03] fginther: thanks very much! [14:06] hello semeone kno what function control the press button volume if is a gpio or what else? [14:11] rsalveti: those branches are annoying. So i've merged and pushed the branch where qt packaging is actually maintained: lp:~kubuntu-packagers/kubuntu-packaging/qtbase-opensource-src_5.0.2 [14:11] xnox: right, but it's useful to have a working udd branch as well [14:12] xnox: and closing the mr is also a good thing :-) [14:12] rsalveti: no, it's useful to use one branch, be it udd one or the specific packaging one. [14:12] xnox: having a working udd doesn't mean we're using it as the packaging branch [14:12] Mirv: why are you pushing merge proposals for both lp:ubuntu/qt* and lp:~kubuntu-packagers/kubuntu-packaging/qt* ? [14:13] having a working udd is good so we can have a packaging merge proposal/code review when needed [14:13] evne if the original packaging is maintained in a separated branch [14:13] Mirv: if you want sponsorship, you can simply request merge proposal review by "~ubuntu-core-dev" on the kubuntu-packaging branches. [14:14] rsalveti: all core-devs have ~kubuntu-packagers ( i think) so really those are the only one that should be used. [14:14] rsalveti: or in otherwards anyone who can upload, have (or should have) the ~kubuntu-packagers access. [14:15] xnox: right, but we might want to do a packaging review using the udd branch, for example [14:15] not saying for core-devs [14:15] rsalveti: similar with all the ~ubuntu-desktop packaging branches, and e.g. seeds branches for each derivative. [14:15] for such packages, we shouldn't be doing uploads without review and such [14:15] rsalveti: no, we should not use udd branches for any qt* packages, cause that's not where they are maintained. [14:16] rsalveti: otherwise that's not really working with the rest of the team who maintain qt* packages, but against them. [14:16] xnox: I know, all I'm saying is that we shouldn't block people from using the udd branchs for simple code review, if needed [14:16] ok. we shouldn't block. but eg. there is no reason for you, Mirv, and I to not use the ~kubuntu-packagers branches, since we know better =) [14:17] sure [14:17] same e.g. with ~ubuntu-installer branches all of them are not using UDD branches, but share history with git imports of debian's d-i, and therefore do proper VCS mergers (since all of them are native packages). [14:17] xnox: I closed the mr btw [14:17] thanks. [14:17] so we don't have another forever-opened mr around :-) [14:18] (imho it's unfortunate that one cannot force ignore udd branch.... *sigh*) [14:18] maybe push --override "README => use this branch instead" =))) [14:19] well, udd might still be useful, if you want to upload something without necessarily going via the original packaging branch [14:19] such as before the release [14:19] if you want review, of course :-) [14:19] which we should be doing anyway (at least for packages touching the touch seeds) [14:24] thanks cjwatson === alan_g|lunch is now known as alan_g [14:36] janimo, patch added to teh bug and verified on three arches [14:37] ogra_, great! [14:38] now i have to dig through that CI crap to actually create an MP ... (sigh, uploading would take me 5 min) [14:38] ogra_, I can sympathize :/ [14:39] ogra_: well, code review at least is always a good thing :-) [14:39] rsalveti, sure sure :P [14:39] (its not like the patch i added to the bug would be hard to review :P ) [14:40] right, janimo can review it easily I guess [14:40] ogra_, if you have a problem creating an MP, the problem is you :P === rickspencer3_ is now known as rickspencer3 [14:42] ogra_: at least once we have a cli way to automatically create the mp ticket and submit it to the "landing spreadsheet" it'll be easier [14:43] looking forward to have such tool in hands [14:44] rsalveti, you mean having your approved MP added automatically to the "landing spreadsheet"? [14:44] Ursinha: well, shouldn't it be added there as soon the mp is in place? [14:44] with a 'needs review' status? [14:44] rsalveti, https://code.launchpad.net/~ogra/phablet-extras/qtmultimedia-touch/+merge/196541 [14:45] ogra_: oh, so it was indeed the i386 x i686 problem? [14:46] Ursinha, i have the package already local, have it built several times fo different arches ... now instead of just uploading the verified fix i have to first find the upstream branch, add my change there, push to LP, find the right name for my ~ogra push, fill an MP form etc etc ... its like 20min extra work where i could just have uploaded and have started my next image build for janimo in the next hour or two [14:46] rsalveti, yeah [14:47] Ursinha, it adds pointless paperwork delay to my work ... [14:47] ogra_, why you say it's pointless? [14:48] (honest question, I don't know) [14:48] Ursinha, because my patch was tested and can be reviwed n the bug [14:48] MPs add an additional buerocracy layer on top of that [14:48] ogra_, and how would that be? you generate a patch, attach to the bug, have it reviewed by someone then apply yourself the patch to the trunk? [14:49] Ursinha, well, what i did for the last 9 years was apt-get source the package ... add my fix, d testbuilds in different chroots for the different target arches i want to test, submit a debdiff to the bug ... and then upload after someone told me "looks good" [14:50] if i follow the habit i did (and most of the other ubuntu-devs did too) for the last 9 years i have to now add additional paperwork [14:50] ogra_, okay, creating a branch, making the fix, submitting a MP isn't exactly the same? [14:50] Ursinha, if i'm a new dev who only works with branches, that might indeed be different ... [14:50] ogra_: right, but just because we're now requiring code review and somebody else to test your change [14:51] so we avoid landing possible broken code in the archive [14:51] as sometimes fixing it after it's already in the archive causes us more problems [14:51] rsalveti, well, it woul be good if we could automate the crossover here ... so you dont have to do the paperwork twice [14:51] right, we should probably optimize the path indeed [14:52] ogra_, you don't need to do the debdiff linking part, if I understood it well [14:52] i'm not objecting to have two ways of doing stuff ... not at all ... but forcing twice the paperwork on ourselves doesnt feel right [14:53] ogra_, agreed, I'm trying to understand why you still need to do the previous paperwork as well when using the MP [14:54] i dont *need* to ... but it helps i.e. debian to find the fix based on the bug description later [14:54] right, got it [14:54] its just good behavior to note down your patch in a bug [14:56] ogra_, if you're using an MP the patch is there, but I think I understand your point [14:57] Ursinha, in a perfect world, i would just upload and $something would hold my upload from entering the archive and generate an MP automatically ... [14:58] ogra_, for that to work you would need to submit to another system than the official one, as there's no way to prevent something to be uploaded if it's uploaded by a coredev [14:58] ogra_, I'll think about it [14:58] well, there is already proposed intercepting ... such a system would need another such layer in the way [14:58] -pre-proposed :) [15:02] ogra_: DONE :-) reviewed, merged, uploaded [15:03] * ogra_ higs rsalveti [15:03] *hugs too [15:03] :D [15:03] * rsalveti hugs ogra_ [15:03] balloons: ping [15:04] iBelieve, pong [15:06] balloons: hi, I submitted a bug fix Saturday for File Manager and the Autopilot tests passed locally. But jenkins failed on two of the tests. Maybe you can help figuring out how to get jenkins to approve my branch. [15:06] balloons: here's the merge request: https://code.launchpad.net/~mdspencer/ubuntu-filemanager-app/fixed-home-icon/+merge/196424 [15:06] iBelieve, perfect. Yes, I'll have a look right now [15:07] iBelieve, what really needs to happen is to migrate the tests to using a new temporary folder under home [15:07] rather than patching home or using home itself [15:07] make sense? [15:07] balloons: isn't that what's happening now? [15:07] since we can't patch home on the actual phone devices, we run into problems [15:08] on the device right now, it reuses home [15:08] iBelieve, ohh, this is a code fix, not for tests :-) [15:08] gotcha, I misread [15:09] balloons: I can change how the tests work if that will help the tests to pass on jenkins, since all three of my branches failed [15:10] iBelieve, yes, what I said above is what needs to happen in order for stuff to run on the device [15:10] but they shouldn't be failing in the lab [15:10] let me see what it says [15:11] iBelieve, looking at the lab video is interesting [15:13] balloons: yeah, for the cut test, it looks like it never clicks the folder. That test works just fine locally. [15:14] so iBelieve look at this log: http://91.189.93.70:8080/job/generic-mediumtests-trusty/220/testReport/junit/ubuntu_filemanager_app.tests.test_filemanager/TestFolderListPage/test_file_context_menu_shows_with_mouse_/ [15:14] I put in extra debug lines at one point because I saw this happening in the lab. Notice in the video the folder never appears? [15:15] ye, check this out: 15:51:56.984 DEBUG test_filemanager:54 - Created /home/ubuntu/tmptewpch/tmpfmDv76e4, a file in HOME [15:15] 15:51:56.984 DEBUG test_filemanager:64 - Directory Listing for HOME [15:15] ['tmpfmDv76e4', '.local', '.Xauthority', '.config', '.cache'] [15:15] the directory exists; it's properly created and can be seen by the filesystem [15:15] balloons: very strange. [15:16] yes. I *think* it's a lab thing and doesn't affect the devices [15:17] zsombi, ping [15:17] balloons: so is there a way to force jenkins to approve the merge, since the tests are working fine locally? === chihchun_afk is now known as chihchun [15:18] cwayne: pong [15:19] hm, my indicator is kind of drunk with latest image [15:19] incoming is actually displaying battery [15:19] Ursinha: is that the issue you had over the weekend? [15:20] rsalveti, I filed that bug, was on r27 [15:20] since r27 [15:20] Ursinha: have the bug number? [15:20] rsalveti, heh, see #ubuntu-ci-eng ... [15:20] * ogra_ just said the same there [15:20] oh, cool [15:20] zsombi, hey, so i was having a bit of trouble dropping my custom theme into XDG_DATA_DIRS [15:20] Ursinha, got a number so we can confirm ? [15:21] it's already confirmed [15:21] cwayne: what was that? [15:21] popey checked that once I filed that, I'm grabbing the bug #, a moment please [15:21] zsombi, so in XDG_DATA_DIRS we have /custom/xdg/data [15:21] if i create a dir for a theme /custom/xdg/data/UbuntuCustom/Themes/Custom it will work [15:22] but if i made it like, /custom/xdg/data/themes/Ubuntu/Custom it won't work [15:22] ogra_, rsalveti, bug 1253804 [15:22] bug 1253804 in unity (Ubuntu) "Indicator icons don't match the settings they display" [Undecided,Confirmed] https://launchpad.net/bugs/1253804 [15:22] thanks ! [15:22] didrocks, ^^^ [15:22] thanks ogra_, /me tracks people now :) [15:23] cwayne: because then you would have to have to have the theme name as "themes.Ubuntu.Custom..." [15:23] ogra_, I'm filing a shitload of bugs, I'm glad you like it lol [15:23] Ursinha: FYI, there is an Unity8 source package, Unity is for Unity7 [15:23] zsombi, yeah, i changed that to reflect it, it still didn't work [15:23] (fixed) [15:23] * ogra_ hugs Ursinha [15:23] didrocks, I haven't linked to that bug, I filed against Ubuntu UX [15:23] iBelieve, well let's debug a bit [15:23] ah, interesting, someone opened it :) [15:23] iBelieve, however, I'll try and make sure your merge isn't held up [15:23] Saviq: mind having a look at bug #1253804? (maybe not you ;)) [15:23] bug 1253804 in unity8 (Ubuntu) "[regression] Indicator icons don't match the settings they display" [Critical,Confirmed] https://launchpad.net/bugs/1253804 [15:24] didrocks, that was popey, you can see it in the comments [15:24] "bad bad popey" [15:24] told that twice in less than 20 minutes! \o/ [15:24] didrocks, yeah I saw that today [15:24] cwayne: have you had everything in place? also in the qmldir in the theme file? (though that is needed only if you have types exported there) [15:24] Saviq: unity8 is a potential right victim? [15:24] cwayne qmldir in the theme folder I mean [15:25] balloons: okay, thanks. [15:25] tmoenicke, btw i can get pinyin to work in today's build, but i still need to switch back and forth first [15:25] didrocks, victim, yes, since it wasn't released for two weeks or something [15:25] didrocks, there is https://code.launchpad.net/~larsu/unity8/per-profile-indicator-positions/+merge/190472 in trunk, though [15:25] zsombi, ah! i forgot about qmldir [15:25] i bet that's it! [15:25] Saviq: ok, Unity8 will be released soon I guess [15:25] zsombi, otherwise, everything else was in place [15:25] Saviq: just after mir [15:25] cwayne: i thought that was fixed, letme check [15:26]