[00:08] stgraber: awesome === chriadam|away is now known as chriadam [00:36] *timezone specific greeting to you all* I'm trying to flash my HTC One (which currently has CyanogenMod 10.1 on it) to test Ubuntu Touch. When I flash the .zip from cdimages.ubuntu.com nothing seems to have changed. Can someone help? [00:43] vthompson: the .zip files on cdimages.u.c are only for the Nexus line of devices [00:44] vthompson: you'll have to find somebody (probably an XDA developer) who has images for your specific phone [00:44] so the generic *phablet-armhf.zip is Nexus only? [00:46] I'll have to continue working towards getting the image to compile then... It's been a long road thus far. [00:47] (on a side note, the mentioning of "cdimages.u.c" didn't lead me to assume you meant the server. I thought the wiki meant a C source file..." === jeremy_ is now known as Guest54016 [00:49] hello? [00:51] can anyone answer questions? [00:51] Guest54016: hello :) irc tends to work best for asking questions if you just ask the question... [00:51] cool. i want to do the dev prev on my polaroid tablet [00:52] but obviously its not on the list, is it possible to do it anyone and have some or full functionality? [00:52] *anways not anyone [00:54] there is a custom rom for my device, but it doesnt include the camera driver, so i havent tried that either as, Im just a beginner and i havent a clue of where to install or which driver to load [00:58] i really want ubuntu on my tablet, i have ubuntu on my dual boot laptop, and i like to use qbittorrent w ubuntu as for some weird reason, comcast can see if i use qbittorrent on windows 7 but not if i use it on ubuntu yet its not hidden by fox fi etc [00:59] hmmmmmmmm [01:00] i guess its helpful to answer my own questions>? [01:00] can't hurt :) [01:00] lol [01:01] well i know how to install pkgs to my ubuntu laptop [01:01] and I'd fully expect your ISP to determine if you're running a bittorrent client of any sort on any OS. bittorrent kind of stands out, in particular. [01:01] but, how/ where would i find the drivers if i can't get the camera to work for example? [01:03] i havent figured out the comcast deal yet. its my dual boot laptop. if i download/ share w qbittorent on windows 7, comcast sends me cease and desist emails, if i use ubuntu on the same laptop and qbittorrent they never send anything (comcast) [01:09] well, in my opinion, i think if i want helpful advice this isnt the place [01:10] Guest54016: or, pick different timezone.. [01:10] Guest54016: europe is asleep, americas are done for the day, except for nutters like me. and I can't help. :) [01:10] something im a third shifter though === daker_ is now known as daker [01:15] Guest54016: you'll really just have to install the ROM you're provided and ask the developer who provided it for help [01:15] No one here knows what camera driver you'd need (in particular). [01:16] sounds to be so [01:16] i read all the info and forums on the rom in question, the last post was a month ago and the alledged developer doesn't go online often [01:17] There seems to have been a lull in the porting effort recently (if you ask me) [01:18] I'm attempting to do the same and it's mostly crickets on the xda forums [01:18] yeah [01:18] Guest54016: when you're running the image with the working camera, do you notice anything obvious in lsmod output? or lsusb? or lspci? [01:18] I think people were under the assumption they'd be given a "beta" for multiple devices. [01:18] Guest54016: when the camera application is running, does lsof or fuser show the camera application using any specific device files? [01:19] right now, the image is just a stock rooted rom [01:20] the dev made a custom rom that was for a model 705x, and i have the model 705. someone tried the x copy for the standard model, works great, but there was dev made it for a different model so the camera wouldnt work [01:20] according to the posts in the forum [01:23] You can do what I'm doing; ensure that your device (705x) is supported by cyanomodgen and then follow the porting guide to compile an image [01:24] It may be hit or miss with your image, however. It's been a bit of "miss" for me on the HTC One (m7spr) [01:27] robert_ancell, heyo. lp:~mterry/unity8/split has a rebased version of my split branch. It will build a unity8-greeter package. I haven't finished testing it to make sure it works still, but it should... === h is now known as Guest76094 [01:30] Hello all [01:30] hello [01:31] yeah my device isnt supported by cynagen mod [01:31] lol if it was easy i wouldnt be here [01:46] I've always been confused on this (mostly because I didn't care), but will there always be a Touch dependence on CyanogenMod? If so, I think there really needs to be more synergy between xda and ubuntu on the OS level. Having "app coding" sessions seems rather silly with a community such as theirs. [01:48] By OS, I mean rom/image level. === _salem is now known as salem_ [01:53] ogra_: ubuntu-touch-generic-initrd 0.9 fixes your problem somewhat more elegantly (I couldn't quite get to my preferred mechanism for the reasons documented in the changelog, but this should be good enough). [01:56] ogra_: I'm not entirely sure why it passed for you even given that you weren't using sbuild. My best guess is that you were running dpkg-buildpackage as root. [02:19] mterry_, Does this sound familiar? [02:19] In file included from /home/bob/bzr/unity8/split/plugins/ListViewWithPageHeader/listviewwithpageheader.cpp:92:0: [02:19] /home/bob/bzr/unity8/split/plugins/ListViewWithPageHeader/listviewwithpageheader.h:20:48: fatal error: private/qquickitemchangelistener_p.h: No such file or directory [02:19] #include [02:23] robert_ancell, no [02:23] mterry_, I just branched and ran ./build [02:23] robert_ancell, do you have qtdeclarative5-private-dev installed? [02:23] robert_ancell, maybe do ./build -s first? it installs some packages [02:24] mterry_, that did it [02:25] robert_ancell, but really if you want a package, do a full package build [03:00] vthompson: I don't think a CyanogenMod dependency is part of the long-term plan, we're already flipping the device images to be an Ubuntu base with Android in a container === NotAww is now known as Aww [03:01] vthompson: that said, XDA has a *lot* of app developers too, and they've shown interest in developing for Ubuntu as well [03:01] ahhh! you can put ubuntu on my android phone?! [03:01] Whennn?! [03:01] LT1stNomad: depends on the phone [03:02] !devices | LT1stNomad [03:02] LT1stNomad: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [03:02] No galaxy s4 yet :( [03:37] mhall119, aside from doing it because you want to and its so darn cool... what advantages are there to having ubuntu on your phone over the android? [03:47] tannji: having the same apps on your phone and your desktop [03:47] or, heck, having the same device be your phone and desktop [03:47] having all the goodness of a full GNU/Linux distro on your phone [03:48] having software that you can contribute to, and that you can have a say in it's development [03:48] different reasons for different people [03:49] also, the edge-gestures are addictive, I keep trying to switch apps by swiping on my DroidX [03:50] multi-tasking on Ubuntu Touch is unlike any other mobile OS I've used (admittedly though I haven't used many) === salem_ is now known as _salem [03:57] good points. The demo I saw recently looked nice. How painful will this end up being? Especially if they end up going a deifferent route than Cyanogen? [04:24] tannji: I don't know what you mean by difficult, it'll just be another form-factor target for Ubuntu [04:25] Cyanogen won't come into the equation [04:25] for the user, it's already a very nice experience [04:36] can a multiple ubuntu-terminal-apps be launched? [04:52] Answer: Yes, by creating a new .desktop and modifying the Name= property === chihchun is now known as chihchun_afk [06:22] why ubuntu-sdk new project not have ubuntu project? [06:22] who faced with this problem? [06:22] any one help me? [06:28] Design standard for "About" and "Options" | http://askubuntu.com/q/315729 [06:40] Clickable emails in Label | http://askubuntu.com/q/315733 [06:44] good morning [07:01] hi everybody, do i need to install some additional package in order to use QtDeclarative in my app? i already have qtdeclarative5-dev installed, but "#include " still gives me an error [07:03] ah, never mind, i should use QtQml / QtQuick... === nik90_ is now known as nik90 === jacky is now known as jalcine === Namidairo is now known as zz_Namidairo [07:26] get the ubuntu sdk preview | http://askubuntu.com/q/315751 [07:37] does ubuntu touch support S2 [07:38] fat: https://wiki.ubuntu.com/Touch/Devices lists the devices an image is available for [07:38] what's s2 again? [07:39] phix: galaxy s2 [07:39] ah === popey changed the topic of #ubuntu-touch to: Ubuntu Touch Support & Discussion | Installation https://wiki.ubuntu.com/Touch | Devices https://wiki.ubuntu.com/Touch/Devices | Release notes https://wiki.ubuntu.com/Touch/ReleaseNotes | Porting guide (advanced) https://wiki.ubuntu.com/Touch/Porting | File bugs at https://bugs.launchpad.net/touch-preview-images/+filebug | Under active development! ☻ === schwuk_away is now known as schwuk === chihchun_afk is now known as chihchun [08:26] cjwatson, well, i didnt run dpkg-buildpackage as root but with -rfakeroot option [08:26] slangasek: linaro published cross-toolchains are amd64 to androidarmel. But point taken. Good that we have multiarch and can just install :amd64 packages anywhere. [08:30] cjwatson, and thanks so much, that would have taken me days [08:39] Good morning all, happy Compliment Your Mirror Day! :-D === yofel_ is now known as yofel === degville is now known as G-Morrison [09:02] Set background for Page{} element in ubuntu touch | http://askubuntu.com/q/315789 [09:07] Anyone here comfortable with autopilot qt? [09:08] I am having trouble accessing QML ListView properties in autopilot [09:08] I have a listView as shown in http://paste.ubuntu.com/5839545/ [09:08] I need to know how to acess the label in the listview using autopilot [09:14] * popey flashes 20130703 to devices [09:17] * ogra_ too ... === udi is now known as Guest1608 [09:23] ogra_: can you see http://popey.mooo.com/mirror/ ? === vrruiz_ is now known as rvr === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [09:52] popey, yep (sorry was afk) [09:54] ta === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [10:39] nik90: the only way I've found to get entries of a ListView was to get the childrens of the first children of the ListView [10:41] nik90: self.main_window.get_object('QQuickListView', 'worldList').get_children()[0].get_children() === chihchun is now known as chihchun_afk === oSoMoN_ is now known as oSoMoN === MacSlow is now known as MacSlow|lunch === alan_g is now known as alan_g|afk [11:23] m-b-o: hi :) [11:23] m-b-o: I currently have resultsList = self.main_window.get_object('QQuickListView', 'worldList') firstResult = resultsList.get_children()[0].get_children()[0].get_children()[0] [11:24] m-b-o: btw I got the idea to use children already from ur autopilot MP :-) [11:25] m-b-o: however the problem is that in my ListView (if you look at the paste.ubuntu) is that I have a Label inside the ListItem.standard. So for some reason I am not able to access the Label. === alan_g|afk is now known as alan_g === schwuk_away is now known as schwuk_away_away === chriadam is now known as chriadam|away [11:37] nik90: I see. Tricky... [11:39] nik90: can you use the text property of the Standard? === docsporpcl is now known as casimiro [11:50] t1mp: The reason I used a Label inside a Standard was because I needed to change the font size, horizontal placement that I got from the visual designs [11:51] t1mp: Initially I used the text property, however I had to change that to accomodate the design request [11:56] nik90: ok, then perhaps it is better to use ListItem.Base which is similar to Standard but without the label [11:56] t1mp: ah ok. I wasnt aware of that. [11:57] t1mp: But would that make accessing it from autopilot easier though? [11:58] hi [11:59] Is the gms/call/sms really work on "Ubuntu touch" ? 'cause I flashed with the manual way on my Nexus 4, and it don't. [12:01] nik90: no I don't think it'll change something for autopilot [12:04] nik90: did you try running autopilot vis to check how you can access the elements with autopilot? [12:04] netcurli: I haven't heard of autopilot vis until now [12:05] netcurli: Until now I have been just getting ideas from other people code [12:05] netcurli: where can i find more info about it? [12:05] mom [12:06] mom? [12:06] just a moment, I am looking [12:07] eylith: seems dialling is broken, will file a bug [12:07] http://developer.ubuntu.com/resources/tutorials/quality/how-to-write-autopilot-tests/ [12:07] netcurli: will take a look. Thanks [12:08] eylith: https://bugs.launchpad.net/phone-app/+bug/1197378 [12:08] Launchpad bug 1197378 in phone-app "Cannot make calls on mako 20130703" [Undecided,New] [12:09] popey: thank you [12:09] rsalveti: popey Hello [12:09] popey: I'm pretty new so i'm not sure this is actually relevent but [12:09] popey: the network section on the property window is empty [12:09] do you guys know this device http://oppostyle.com/ [12:09] oppo find 5 [12:09] ? [12:09] i do [12:09] looks beautiful [12:09] I'm not sure the phone is actually connected to the gsm network [12:09] i'm thinking to get one [12:10] eylith: yeah, mine can't get on the gsm network [12:11] franco Kernel probably is going to write Kernels [12:11] really ? [12:11] Francisco Franco btw === MacSlow|lunch is now known as MacSlow [12:11] love dat android kernel [12:12] yes ! on g+ is a discussion runing now on his page [12:12] thank you very much popey [12:12] esigolo This is cool [12:13] yes it is ! [12:13] is a very good kernel [12:13] yes, I guess he is working very hard on it [12:14] sergiusens: seems 3g/gsm is busted in the 20130703 build for mako, are there some diags / logs I can get for bug 1197378 ? [12:14] bug 1197378 in phone-app "Cannot make calls on mako 20130703" [Undecided,New] https://launchpad.net/bugs/1197378 [12:14] sadly here an oppo is about U$850 [12:15] i'm on Brazil btw === dandrader is now known as dandrader|lunch [12:19] esigolo> Why do you want an oppo ? [12:21] is heard it is a solid phone ! [12:21] and the hardware is very good [12:22] i have dropped my nexus so many times [12:22] nexus 4 ? [12:22] yes [12:23] my nexus 4 smashed, the first time I dropped it [12:23] I now have a cover on it. [12:23] popey: cover aways saves me ! [12:24] the oppo does look like a nice device, and seems like they're happy to have CM support! [12:24] http://wiki.cyanogenmod.org/w/Find5_Info [12:24] I dropped my GNexus just once, the screen was "crashed" (sorry for my broken english) [12:24] would _love_ to have one of those running Ubuntu Touch! [12:24] i didn't unpacked my nexus 4 before i get the cover and the screen protection :D [12:24] popey: in fact that is what is holding [12:25] if I sell my nexus 4 [12:25] no ubuntu touch fo rme [12:25] and no Firefox OS neither [12:25] but I guess it isn't the point right [12:26] i got sell my galaxy s i9000B and my S2 === VDVsx_ is now known as VDVsx [12:35] netcurli: the autopilot vis is useful but very cumbersive as well. My listview results change for every search query, and the autopilot vis does not refresh/update when it does. === greyback is now known as greyback|lunch [12:41] ogra_: hey, do we have something hooked to the pending -> current symlink update for ubuntu-touch (flipped) images? === SkavenXXI-[OFF] is now known as SkavenXXI [12:42] lool, i know there is cdimage code that submits it somewhere, not sure that is used on the Qa/isotracker side yet though [12:42] lool, i guess cjwatson might know better [12:42] :q [12:42] nik90: yeah, if you have a changing tree of objects, you should probably use it only to inspect one state [12:42] Ups [12:45] can someone remind me the ssh command to make an ubuntu phone make a call remotely? [12:47] gema: ssh -t phablet@device my_command [12:48] sergiusens, maguro and grouper look fine ... i'll send the announcement if i got some mako feedback from you [12:48] ogra_: is there a bug for the upstart/ofono issue? [12:48] sergiusens: it is "my_command" I am looking for :) [12:48] sergiusens: ofono not starting > I'm thinking this ought to be catched by image smoke testing [12:48] popey, not sure, fix was uploaded a while ago though [12:49] netcurli: yup then I am in trouble since the listview results are very dynamic and so my autopilot needs to somehow magically figure out how many list headers and standard items are present to get the correct children to evaluate :( [12:49] sergiusens: would you know which smoke tests we currently run for touch before we expose an image to phablet-flash? [12:49] popey, bug 119722 accroding to the upload [12:49] bug 119722 in gstreamer (Ubuntu) "Sound Problem !!!" [Low,Invalid] https://launchpad.net/bugs/119722 [12:49] oops [12:49] ogra_: did we already disable the ubuntu-touch-preview crontab on cdimage? I only see the ubuntu-touch one [12:49] popey, bug 1197225 accroding to the upload [12:49] bug 1197225 in upstart (Ubuntu) "upstart-file-bridge assert failure: string.c:396: Assertion failed in nih_str_split: str != NULL" [Medium,Confirmed] https://launchpad.net/bugs/1197225 [12:49] (copy paste error) [12:50] lool, not to my knowledge, and we shouldnt until ports can use flipped [12:51] ogra_: hmm are the ubuntu-touch and ubuntu-touch-preview images built in different places? I only see "for-project ubuntu-touch", no entry for -preview [12:51] I guess the -preview one is still built on a jenkins instance somewhere [12:51] yes, -preview comes from the OEM infrastructure [12:51] ogra_: I don't have a mako [12:51] sergiusens, oh, manta then ? [12:52] lool: gema or plars would have the list [12:52] lool: preview is built on jenkins [12:52] sergiusens: thanks [12:52] ogra_: I'll do manta... [12:53] ok [12:53] popey, did mako look ok to you (except for the upstart issue) [12:53] lool: ubuntu-touch plain is build on cdimage but the android bits are pulled from -preview [12:53] gema or plars: Would one of you two have the list of smoke tests we run on ubuntu-touch /pending images before these get to be /current? Looking at whether this would catch the ofono not starting regression [12:53] ogra_: mako is the galaxy nexus right? [12:53] wwell, they are pulled directly from jenkins ... but yeah, same source as -preview [12:54] davmor2 nexus 4 [12:54] davmor2, nope, n4 ... maguro is the gnex [12:54] sergiusens: Thanks; mostly trying to check whether we have enough end-to-end QA on a phone call that "ofono not starting" prevents the image from being published [12:54] lool: I am almost certain we are not running the tests before the images get published yet === oSoMoN_ is now known as oSoMoN [12:54] ogra_: it was the 'in a nutshell summary' :-) [12:54] lool: and right now we don't have any test that would catch that, but we will soon [12:54] :) [12:54] gema: would you know what's missing for this and/or who is working on hooking such tests in? [12:54] ogra_: they should call them complete different things it would make life much easier :) [12:54] lool: we could add something simple , like checking the list of processes if you want [12:55] lool: I am working on some connected tests, that will establish a call and send an sms [12:55] gema: we have NM mocked tests; one for ofono would be ideal [12:55] lool: and fail if this cannot be done [12:55] that wont help much [12:55] ogra_: yes [12:55] lool: but I wouldn't block an image based on those, I'd run them after the image is published [12:55] i.e. ofono can run but not have a proper connection to rild on the android side [12:55] ogra_: if you have a suggestion on how to check we could addd one test today for that [12:56] which means it looks like you can do calls, but in facyt you cant [12:56] *fact [12:56] ogra_: we are going to make the call [12:56] ogra_: well it would have caught the ofono startup failure at least [12:56] and fail if the call doesn't make it [12:56] I'm happy if we manage to get an end to end real call happening, but also trying to avoid the perfect being the ennemy of the good :-) [12:56] lool: this is very basic stuff [12:57] lool: perfect is something else [12:57] gema: Not sure why we wouldn't block an image on such failures; I only mean the current vs. pending symlinks, not blocking upload to cdimage [12:57] gema, sounds good [12:57] lool: but if you guys have a command line way of checking for ofono that you want in place [12:57] gema: I was thinking more along the lines of the NM tests that pitti wrote [12:57] I will add that right now to the test suite that runs daily on the images [12:57] lool, well, being able to make calls is kind of essential on a phone :) [12:57] don't you have basic regression test before uploading ? [12:57] but on an image rather at package build time [12:57] ogra_: exactly [12:57] if we can make publication depend on that that would be great [12:58] lool: sounds good, will talk to pitti about that [12:58] ogra_: but I also know it's hard to have a real cell connection in a datacenter, and that also introduces a lot of noise in the QA process [12:58] lool: and that is why I don't wnat to make the images dependant on real calls [12:58] you can mute the speakers :P [12:58] (yes, understood, .... couldnt resist picking that up :) ) [12:58] lool: but I think it is a test we need [12:59] gema: whatever; today's image having a critical bug in using any telephony functionality seems pretty bad to me, so I'd like to stop such images from getting exposed to phablet-flash / daily updaters === greyback|lunch is now known as greyback [13:00] its not a critical bug using any telephony [13:00] gema: Alright, happy if you keep me in the loop for the cdimage publishing integration and actual tests :-) [13:00] lool: will do! [13:00] it is broken on exactly 50% of the supported phones :) [13:00] gema: lool ogra_ nmcli d should list the ril device [13:00] gema: lool ogra_ apt-get install ofono-scripts and run some of those [13:00] sergiusens, that doesnt tell me if the protocol works between ofono and ril [13:00] gema: ^ [13:00] hey popey, when you've got a minute, could you add music-app, sudoku-app, stock-ticker-app and dropping-letters to the touch-coreapps package? [13:00] sergiusens, now that sounds good ! [13:01] sergiusens, ogra_: Whatever would have caught today's failure to start ofono on some phones and stop the image from being published should get added to our smoke tests ASAP :-) if that's an ofono script, fine [13:01] sergiusens: ok, will try that and ask you some questions if they arise [13:01] lool, i guess that was a tired rsalveti catching it :) [13:02] dpm: if you are doing that you may want to consider dropping the dummy music app too to stop confusion [13:02] not sure how easily he is scriptable [13:02] :) [13:02] we can call his home number for the daily smoke testing and hook voice recognition to ack the build [13:02] ogra_: if you guys have any script to catch regressions , you may as well send them my way and we'll add them to the test automation [13:02] in any case the fix is uploaded and the image might even have built before i write the announcement [13:02] "Ricardo Salveti speaking" => build passes [13:02] "groans" => build rejected [13:02] :) [13:03] dpm: oh, they're not? [13:03] gema, yeah, i didnt know about ofono-scripts until a few mins ago [13:03] ls [13:03] popey, not in raring, at least. I haven't checked in saucy [13:03] you're right [13:03] esigolo, no such file or directory [13:04] davmor2, which dummy music app? [13:04] ogra_: I will hack something together with plars and let you guys know whenever it is in place [13:04] they're in the touch-collection one. which I always install with touch-coreapps, so never spotted it [13:04] ogra_: I aways do that [13:04] ogra_: remind me, where do the crontabs for livefs builds live again? [13:04] are these on buildds? [13:04] * pitti waves hello [13:04] lool, nusakan [13:04] dpm: there is a .desktop app on the phone that just opens a white page and nothing else [13:04] lool, sudo -u cdimage -i && crontab -l [13:04] hey pitti [13:05] ogra_: yeah, I see the crons to assemble the livefs into an image [13:05] davmor2, do you know which package that .desktop phone comes from? [13:05] ogra_: but how are the ones kicking livefs buildders to build images? [13:05] lool, evey command with the --live arg actually triggers a livefs build first [13:05] dpm: not a clue but popey might [13:05] ogra_: ah, thanks [13:05] sergiusens: rsalveti: now that we are getting the flipped image container, can we unblock unity8 new packaging by merging https://code.launchpad.net/~didrocks/qtubuntu/move_android_only/+merge/171965? [13:05] lool, cdimage changed a lot [13:05] sergiusens: rsalveti: once merged, I'll refresh the seed [13:05] lool: after a quick conversation with pitti he tells me we are not yet covering 3G/ofono with his tests and he doesn't know when we will [13:05] its all python now as well [13:05] :( [13:06] * ogra_ really loved the simplemness of the old shell architecture [13:06] lool: I'm currently working on mocking/recording/replaying tty device support in umockdev, so that we can take a real phone and mock it with modemmanager/ofono/etc; but I just started on that this week [13:06] lool: ofono doesn't currently have an autopkgtest; is there something which we can test about it that doesn't require actual hw? [13:07] gema pitti: Yeah, this is what I was guessing, but I think we ought to a) add a simple test that ofono starts (just a ping test) on a full image and b) make sure this test passes before we update /pending to /current [13:07] pitti, the kernels dont have tty's enabled usually [13:07] ogra_: oh -- how do we talk to modems/3G devices then? [13:07] (it breaks pixelflinger, so android kernels usually have ttys and consoles off ... and point console= to some serial device) [13:07] lool: do you think the ofono-scirpts that sergiusens mentioned are what you are looking for? [13:08] as a ping test? [13:08] ogra_: well, VTs != support for tty [13:08] pitti, modems usualy through trild (100% proprietary, differing completely by vendor) [13:08] ogra_: we certainly should have general tty/pty support, I guess/hope you mean we don't have a tty1..6 with getty? [13:08] pitti: Good question; I guess we will need device emulation anyway; not sure we can test anything useful without the hardware (would ofono even start without any kind of telephony hardware, maybe) [13:08] ogra_: you have an n7 and a galaxy nexus right? what happens when you click on the calculator app? [13:08] *rild [13:08] wow, Ubuntu-touch is made with C++ ? [13:08] pitti, right, no gettys [13:08] what isn't made with c++ [13:08] pitti, right, no tty1 either except on tegra [13:08] ogra_: that's fine [13:08] davmor2: known problem [13:08] gema, lool, ogra_: hi, just waking up so I'm not quite caught up yet and my coffee reserves are still empty... as gema said, we are in the middle of getting the pieces in place to test actual calling, but is there some simpler set of tests that could check for conditions like this? [13:09] popey: thanks [13:09] davmor2: bug 1196653 [13:09] bug 1196653 in Ubuntu Calculator App "calculator fails to start on device" [Undecided,In progress] https://launchpad.net/bugs/1196653 [13:09] doomlord> a lot of thing actually [13:09] plars: exactly what we're discussing; get a cup of coffee to parse the backlog :-) [13:09] popey: I just went to use it and got a white screen so tried my n7 instead and it happened there too :0 [13:09] doomlord: /whois popey [13:09] plars, nmcli d |grep ril ... that woould be a good start to see if the stack works [13:09] ogra_: usually any modem/3G device creates some new tty devnode which you use to talk to it (like ttyUSB0 for an usb dongle, etc.) [13:09] ops [13:09] Hello esigolo ☻ [13:09] i was trying to find popey on g+ [13:09] pitti, not in android :) [13:09] LOL [13:09] esigolo: popey.com/+ [13:10] ogra_: if the devices/kernel on android use something completely different, we are screwed of course [13:10] ogra_: that's on my list, let me guys know any other you want added [13:10] thanks [13:10] pitti, 3G devices show up as rmnet in NM though [13:10] pitti: also, I understand the autopkgtests are run standalone right now on dedicated machines; would we be able to leverage these on the actual cdimage images that we would like to publish to smoke test them? [13:10] ogra_: so what do they use to talk to the hw? [13:10] pitti, a proprietary daemon [13:10] ogra_: right, I mean how does that daemon talk to the hw? [13:10] which differs execpt for the binary name from vendor to bvendor [13:10] ogra_: the normal way would be to send AT commands through some /dev/ttywhatever [13:10] not even the protocol is guranteed [13:11] pitti, right, no what ofono does, it attaches to a socket that rild created when it was started by the android container [13:11] pitti: RILd is an android-side daemon; my vague memory is that we have some python ofono bindings to talk to it over binder but that these ought to be replaced with C [13:11] not binder [13:11] there is /dev/socket/rild [13:11] specific just for that [13:11] lool, ogra_: all that is way up in userspace and uninteresting for record&replay; I'm interested in the kernel interface that it uses [13:12] and rild manages everything rleated to the modem [13:12] i. e. device node, ioctl, mmap [13:12] pitti: I would think it uses multiple ttys indeed [13:12] pitti, then you have to ask the vendors for an NDA to look at their rild implementations [13:12] lool: I certainly hope so :) if it does, then we can use the record/replay stuff [13:12] pitti: There might be more weird stuff going on; I know there's complex machinery to allow some proprietary userspace drivers to talk to the hardware via RILd [13:13] pitti, but how do you plan to do such stuff anyway, it all happens in the container which you cant access at all [13:13] ogra_: the idea is that we record what rild sends to the device nodes, and replay that as part of a test suite; we wouldn't care about the particualr protocol [13:13] all we see on the ubuntu side are some of the devices, binder and /dev/socket [13:13] * ogra_ would rather suggest to hook somewhere into platform-api [13:13] ogra_: I guess then we should record what's going on in /dev/socket === zz_Namidairo is now known as Namidairo [13:14] pitti: I see a mock-ril in the source tree, looks promising [13:14] (and yay for this stuff to reinvent the whole world again..) [13:14] pitti, yeah, and forward it to a platform-api debug function you can switch on [13:14] popey: you're going to port Ubuntu Touch for find 5? say yes [13:15] ogra_,lool: I'll hook up the current/pending bits in cdimage as soon as somebody does it in jenkins. From the cdimage point of view it's trivial. [13:15] pitti, awe is pretty good with ril, wait for him to return next week, he can surely priovide you with some good ideas [13:15] lool: ok, seems this stuff is utterly non-standard, I'm afraid I now can't give you an estimate how long that would take, or what is necessary (or whether it's possible at all) [13:15] esigolo: lol [13:15] hi everyone [13:15] cjwatson: Yeah, I know you've built all the pieces there, I'm just not sure we have the smoke tests running anywhere yet [13:16] rigved: hi [13:16] are the touch images for galaxy nexus based on saucy now? [13:16] pitti, well, its a "over the thumb" standard (every vendor has a daemon called rild ... they not necessarily speak the same protocol though) [13:16] lool: Indeed. Just responding since my name was invoked ... [13:16] rigved, since several weeks [13:16] cjwatson: summoned! [13:16] esigolo: hi [13:17] pitti, i think you really want to hook into a higher level (ofono itself probably) [13:17] ogra_: so it sounds like we wouldn't want to know/record/replay what's going on in rild and below, and instead mock/record/replay what's going on on the rild API instead? seems we have zero control over what's beyond anyway [13:17] rigved: Hello [13:17]