[00:04] cwayne: I haven't seen an API or API docs for the infographic yet, no [00:17] w-flo, now i am booting into a empty screen without any reboot loop i wonder if it worked. [00:17] i cant' adb shell thought [00:18] ogra said it takes a long time for adb to come up on his device IIRC [00:18] hmm [00:19] I thought that it can't get stuck in the initrd (will spawn adbd if anything goes wrong), and if upstart fails to do its job, it should reboot... so if it's just a black screen.. what could have happened? upstart getting stuck? upstart failed to launch adbd? hum.. [00:20] I have the "boot to black screen without adb" when the /data flash fails halfway [00:20] you probably remember me saying that a few times already :D [00:22] how do i get to terminal on today's 19.2 image? [00:22] RobbyF, tap the "installed" header to expand the app list [00:23] that's how it was 2 days ago. not today though. [00:23] I just have 'more suggestions' [00:23] I guess I'm at 19.1 then. or 19.. [00:24] I think all 19 have been like that [00:24] w-flo, hmm ima re install /data then [00:25] annerajb, I think ogra recommends to check if there are files in "/var/log" [00:26] those are probably the last files getting flashed [00:26] so if there are files there, the flash was successful [00:28] there are files there [00:28] ima delete them boot once [00:28] and go into recovery to check if there are new ones [00:29] great idea :) [00:30] no black screen now -_- just the galaxy s and cyanogwen mod animation logo [00:31] quick question. where does the phablet-flash grab from? Ubuntu-touch or Ubuntu-touch-preview? [00:32] dixeflatline, ubuntu-touch [00:32] the one that says preview was for the old unflipped image [00:32] the ubuntu-touch has the new flipped image [00:32] unless you're using the community feature, dixeflatline [00:32] thank you. i thought so. [00:33] I was updating my N7 on MultiRom and i garbed the preview. Didnt seem right. [00:35] what's in the community feature? [00:36] yeah what is the community feature? bug reporting or something? [00:37] dixeflatline, it's for flashing ubuntu touch community ports (for unsupported devices) [00:37] so it doesn't apply to the nexus7 [00:37] hey [00:38] oops wrong channel. [00:38] thanks for the info w-flo [00:38] w-flo, so it appears that it wrote one log file /var/log/lxc/android.log but it's empty :( [00:39] IIRC there is a bug in upstart.. it fails to create /var/log/upstart if it doesn't exist. Though I wonder why you have no dmesg log.. [00:39] but still, that log file is a sign of life :) [00:40] so i guess it started running the rootfs :D [00:40] yes, I guess so! lxc sounds very much like it [00:40] maybe ogra can tell what's going on [00:40] manually create the /var/log/upstart directory and see if upstart puts anything in there :) [00:41] and BTW, the android.log is empty on my device, too. [00:42] w-flo, i recreated all the directories i deleted just in case they need to exist [00:43] meanwhile i am making progress on building a usb uart serial cable for the kernel :D [00:43] annerajb, your device has issues with partition labels, right? like the vision? maybe the update-fstab script fails... the upstart logs should probably tell === chriadam|away is now known as chriadam [00:45] w-flo, yeah my device has no partition labels and i coudnt find a fstab to remove nosu‎id [00:45] the /usr/lib/lxc-android-config/update-fstab script scans for partition labels to add data, system and vendor to /etc/fstab [00:46] It will fail, so maybe that's causing the hang.. I don't remember [00:46] http://status.ubuntu.com/coreapps-13.10/ubuntu-emailclient-dev-coreapps-13.10-month-4.html [00:47] I wonder how sudoku and dropping letters are a priority to an email client. [00:47] w-flo, it's booting upstart wrote files [00:48] yay! [00:48] w-flo, http://pastebin.com/qmsfPnaA [00:49] i think it should say counted cpus 1? lol [00:49] RobbyF: sukoku and dropping-letters were started independently and only later added to the core apps [00:50] I have the same ureadahead.. [00:50] Ty sir. [00:50] w-flo, which file has the CPU count? [00:50] errr... annerajb ^ [00:50] lol let me check [00:52] ureadaheadlog has that [00:53] contains only "ureadahead: Error while tracing: No such file or directory" for me.. :o [00:55] There's a timeout in /etc/init/ureadahead.conf .. 180s , then an additional 45s.. [00:55] that's good lol it means i should start in there [00:56] maybe if you wait 4 mins it continues booting? :D [00:56] hmm ima try that then [00:58] w-flo, do you have this container-detect stop/pre-start, process 253 [00:58] on the container-detect file [00:58] yes, the PID is 1093 / 1092 / 1089 though, not 253 [00:59] I guess that's fine :) [01:02] oh well, it's late and I must really go to bed now. I keep my fingers crossed of course :) === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [01:32] I'm following this guide I was linked to earlier https://wiki.ubuntu.com/Touch/PortingFlippedInProgress#Touch_Developer_Preview_-_Porting_Guide_2.0 and it seems I'm missing the breakfast command, what is the package or step I'm missing? === Ursinha is now known as Ursinha-afk [01:44] How would I call breakfast/brunch on an unofficially supported phone? The github repo for it is here I believe https://github.com/TeamApexQ/android_device_samsung_d2-common [02:21] hi everyone [02:22] hi [02:23] i have a question [02:24] can i install ubuntu tablet version into other tablet device? [02:24] if ported correctly I don't see why not. [02:27] let say i have samsung tablet [02:27] can i install it with ubuntu? [02:27] https://wiki.ubuntu.com/Touch/Porting [02:27] android? [02:28] clicking a few buttons probably not. dev work needs to be done. [02:28] mierol, Ubuntu tablet version, do you mean the saucy version? [02:28] see if it's in the list - https://wiki.ubuntu.com/Touch/Devices [02:29] Robby F, thanks [02:30] np, hopefully that puts in you the right direction. [02:34] What files do i download to install ubuntu touch on my galaxy note 2 verizon? I seen it was listed in a list of compatible devices. Any help would be greatly appreciated. [02:38] nmarriotti, you need two files [02:39] one of them is the rootfs http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ [02:39] the other one comes from whoever ported your device which is a zip file usually aroudn 50megabytes [02:39] you flash those two files with recovery [02:39] and reboot [02:41] Oh ok. Ill have to look into getting that 50mb file. Thanks for the help. === salem_ is now known as _salem === xIndirect is now known as Nick === xIndirect is now known as Nick === chihchun_afk is now known as chihchun [04:04] beuno: please file a bug with details. === Namidairo`bnc is now known as Namidairo [04:47] Does anyone know if there's a preferable way to ensure you're running the latest ubuntu touch? [04:48] I used apt-get update/upgrade and everything indicates current, but from what I'm seeing online, there's a working alarm clock, but I don't have it. [05:08] Kaleo: please e-mail me the patch / codereview link, especially if it's related to fixing the 5.1 blocker bug. === Namidairo is now known as Namidairo`bnc === Namidairo`bnc is now known as Namidairo === milli` is now known as milli === arunprasadr_away is now known as arunprasadr [07:00] good morning === Namidairo is now known as Namidairo`bnc === Namidairo`bnc is now known as Namidairo [08:29] mpt, hey, I've been looking a bit at https://wiki.ubuntu.com/Location#Phone and I've some questions [08:29] 1- that seem to assume you always have a gps, what happen when you don't (e.g on a desktop), is that a non supported case? [08:30] 2- the apps listed, it's written "It should include only those applications that have ever requested access to location data. " ... should that include uninstalled apps? [08:31] 3- would it make sense to have controls to turn on/off gps/wifi there since they are useful for location? === Namidairo is now known as Namidairo`bnc === Namidairo`bnc is now known as Namidairo === Namidairo is now known as Namidairo`bnc [08:47] seb128, 1. It's phone-specific and assuming a GPS right now, that's why it has a big heading "Phone" :-) [08:47] didrocks, https://code.launchpad.net/~thomas-voss/location-service/gcc-4.7-by-default [08:48] mpt, do all phone have a gps nowadays? [08:48] asac: in case you wonder why we revert the converged indicator-messages: [08:48] 10:45:01 seb128 | new indicator-messages sigtrap on start [08:49] tvoss: hi! Do you remember that we spoke about surface reparenting, once? I now got a more difficult case :-) [08:49] asac: so, clearly not ready, we revert [08:49] tvoss: here's the design: https://wiki.ubuntu.com/OnlineAccounts#phone-access [08:49] seb128, I was assuming it, but I haven't seen a hardware requirements document anywhere. [08:49] mpt, ok, wfm [08:49] mardy, btw, I'm halfway through replying to your reply about that [08:49] tvoss: that is, a dialog popping up in front of another window; will this be supported? [08:49] didrocks: and now? [08:49] mpt: thanks! [08:50] asac: well, tedg will be pinged [08:50] and sil2100 will list that as a note for the stack [08:50] let's first do the revert, already taking some time [08:50] mardy, it is not right now, and I would prefer if we had a list of use-cases that we have to support [08:50] as not enough integration tests on that didn't catch it [08:51] didrocks: so this means back to a ppa? [08:51] tvoss: OK, we'll first discuss the options with mpt and then come back to you, if needed :-) [08:51] seb128, 2. Good point, it should list only currently-installed apps. https://wiki.ubuntu.com/Location?action=diff&rev2=6&rev1=5 [08:51] didrocks: we really want this in the archive [08:51] even if we use a different source name [08:51] mardy, ack and thanks. It would help me if we could start noting down what we really need in an abstract way [08:52] asac, you don't really want something that sigtrap on start, no [08:52] mardy, tvoss: katie and I are working on use cases and patterns for dialogs this week. [08:52] asac, get the work properly done if you want it in... [08:52] seb128: I didnt say we want a crash [08:52] all i am saying is that we dont want to use a ppa version in official images anymore [08:52] mpt, ack, but this is a bigger issue: reparenting and cross process assembly of UIs [08:52] asac, yet at the same time you push to land stuff that didn't get the minimal level of qa/testing [08:52] seb128: wait :) [08:53] seb128: i always said: get it out of the ppa ... the fact that people decided to go for the converged indicator-messages [08:53] was THEIR choice [08:53] i always was saying: get it in using a different name or something [08:53] so you dont need to block on convergance [08:53] that's a waste of resources [08:53] seb128: having stuff in ppa is wasting resources as well ... its hard to sum things up [08:53] let's create extra work, duplicate sources, to converge back a week later [08:54] we don't have the luxury of throwing days of work away like that [08:54] seb128: a week? i think i said that over a month ago :) [08:54] now it might be a week ahead [08:54] well, everybody has ETOOMUCH [08:54] tvoss, we have the same issue on the desktop with PolicyKit dialogs. The PolicyKit API doesn't include "which window should the prompt be modal to", so sometimes it isn't focused properly, bug 877265. [08:54] bug 877265 in software-center (Ubuntu) "10.04 LTS: Focus does not go automatically to password field when 'install' button is pressed." [Low,Confirmed] https://launchpad.net/bugs/877265 [08:54] seb128: right. so it doesnt help either if suddenly ppas bust our images, making everyone go on a firedrill as well :) [08:55] mpt, yeah, that's why I think we need to think deeply about the cross-process reparenting stuff [08:55] seb128, 3. Why would you want location to be on but GPS to be off, or GPS to be on but location to be off? [08:55] tvoss, I guess on the phone it's easy because you can just assume that it's the app's frontmost window, which is usually its only window anyway [08:56] but on the PC, trickier. [08:56] mpt, good point, what about wifi? [08:57] asac: did it happen, do you have data about that happening (the ppa being bursted because of not being in the archive)? [08:57] didrocks: yep we had that twice :) [08:57] mpt, sure, but the phone is only the special case here and I would like to start working on a generic solution [08:57] asac: anyway, as I told, for indicator-messages, we can't split the package here, so if you want this, slangasek proposed his team IIRC [08:57] didrocks: that was the trigger for using saying we want to go for a zero-ppa approach [08:58] right now, we are reverting to get the image on shape again [08:58] asac: the only case I hard was because of unity8 [08:58] which is in the image now [08:58] didrocks: when did the indicator-messages? [08:58] didrocks: sure thats one case, next case will happen with very high certainty [08:59] and usually at the worst moment you can imagine [08:59] and with the componentn most painful to fix :) [08:59] right, I'm used to that :) [08:59] whats that law called again? [08:59] heh [08:59] murphy is my friend [08:59] it seems [08:59] right :) [09:00] but yeah, you have two options, gave them to you, right now, let's first get it back on shape [09:00] and ping people on why the revert [09:00] so hence, i have been long enough around to know if we dont go for zero-ppa we are FAIL :) [09:00] didrocks: yeah thats perfect [09:00] asac, if we land things without qa and tests we are FAIL as well [09:00] didrocks: when was the converged version landing? [09:00] tvoss, this is a special case of the origin function. For example, if a game has a "Tweet This High Score" button (my favorite example), ideally the Online Services prompt would emanate not just from that window, but from that button. [09:00] seb128: i have no control about what folks land [09:00] asac, today [09:00] asac: it landed today [09:00] asac, yet you pressure them to land [09:01] cool, then we can back out [09:01] like 3 hours ago in proposed [09:01] mpt, sure, but there is a difference between in-process dialogs (fine) and the case where the dialog originates from another process [09:01] 2h56 exactly [09:02] mpt, mardy, i think we should be using the design in this doc https://docs.google.com/a/canonical.com/document/d/1UwAQTXgEyZSD3di6fAUS0W18rKxh8TXb1TwsmkgbGG0/edit#heading=h.d5gufdir1r88 which uses sheets [09:03] tvoss: C++11 and friends I guess, right? (just ping me once you push the comment) [09:03] (already spotted some ABI breakage 2 cycles ago in gcc because of that) [09:04] tvoss, hey, I've some "location" question for you ;-) [09:04] didrocks, pushed [09:04] seb128, I'd rather leave it out to start with ... I don't think it'll be common for you to turn on wi-fi just to improve location detection. I could be wrong. [09:04] seb128, shoot [09:04] tvoss, 1- how do you turn the location service on/off (asking for system settings) [09:04] mpt, ok, wfm, thanks for the replies! [09:05] seb128, already clarified with charles, https://code.launchpad.net/~thomas-voss/platform-api/add-controller-for-location-service [09:05] tvoss, oh, right, he needs that for the indicator as well I guess ... thanks! [09:05] seb128, ack [09:05] tvoss, 2- https://wiki.ubuntu.com/Location#Phone has a list of apps that "have ever requested access to location data." [09:05] the corresponding dbus interface in the implementation will provide us with mediation hooks [09:05] tvoss, do you know if we have that info recorded somewhere? e.g what apps can access location [09:06] tvoss, and if we have a method to list those and their status (access allowed or not) [09:06] seb128, hmmm, that would require apparmor integration [09:06] seb128, so essentially, the click packages manifests contain that information [09:07] seb128, however, if it has ever been accessed might take more work for the security guys, let's check with them [09:07] tvoss, thanks, do you want to me to send an email on a list about that? or do you ping them? [09:08] seb128, will ping them, I'm interested in the more general question of tracking apps trying to access services and querying that information, too [09:08] tvoss, ok, let me know how that goes [09:08] tvoss, thanks! [09:08] seb128, ack [09:16] didrocks, https://code.launchpad.net/~thomas-voss/platform-api/location-service/+merge/180692 [09:17] tvoss: that will maybe take a little bit, I need to advance on system update [09:18] didrocks, ack :) [09:19] didrocks: hi, any news about getting ubuntu-ui-extras into saucy ? [09:19] nerochiaro: I asked robru about it, didn't get any feedback [09:19] nerochiaro: not sure if Mirv or sil2100 have spare cycle [09:20] nerochiaro: robru is travelling tonight (and is in the US timezone) [09:26] Mirv: sil2100: can you guys help with getting a package into saucy ? [09:27] nerochiaro: I'm queued with my other things including getting other packages into saucy :( at the moment, at least. [09:29] hi pete-woods [09:29] katie: hi [09:30] nerochiaro: I might be able to help with it, but later [09:30] sil2100: thank you [09:30] oSoMoN:^ FYI [09:30] pete-woods, i just filed a bug about the minimum value thing for the infographic [09:30] pete-woods, cos mine still isn't displaying properly.. :s [09:31] katie: okay, cool, I think it's worth implementing - I'm just a bit up to my eyeballs in bugs with other packages at the moment [09:31] pete-woods, i figured as much, that's why i filed the bug :) [09:31] :) [09:31] pete-woods, this is just an fyi [09:32] katie: tbh I'd much rather work on what you've asked, as it's a codebase that is clean, has practically 100% test coverage and is totally solid [09:48] hey guys, can someone tell me how to enter the pin for my sim card? [09:57] dednick, and the menu shows up on the session bus (verified with d-feet) ... the stand-alone example is able to pick up the menu-model without any issues [09:57] dednick, also using the same libqmenumodel in both cases (system-wide installed one) [09:57] @MacSlow do you know how to enter the pin for my sim card? [09:58] pureCenor, what? [09:58] pureCenor: https://fcns.eu/2013/07/26/ubuntu-touch-unlock-change-reset-pin/ [09:58] look in /usr/share/ofono/scripts [09:58] (found that by typing ubuntu touch sim pin into google) :D [10:02] pureCenor: you can ignore the wget lines [10:02] oSoMoN: gusch_: can any of you guys quickly check if you can access this ? https://jenkins.qa.ubuntu.com/job/ubuntu-ui-extras-saucy-armhf-ci/21/artifact/work/output/*zip*/output.zip [10:02] pureCenor: the stuff is pre-installed now [10:02] look in /usr/share/ofono/scripts [10:02] nerochiaro: no, and it’s expected, the jenkins instance is down, check your e-mail [10:03] oSoMoN: ok, missed that. thanks [10:03] that worked. then when i say: enter-pin "mypin" [10:03] it says: bash: enter-pin: command not found [10:06] pureCenor: ./enter-pin [10:06] aaawwwwwwww. i hate myself [10:12] pureCenor: ./enter-pin pin XXXX [10:15] perfectly working === chihchun is now known as chihchun_afk [10:27] Morning all === chriadam is now known as chriadam|away [10:39] ogra_: do we know when the new network indicator lands? This 3g issue is starting to drive me crazy .... well crazier :D [10:40] didrocks, can I get another top-approve? https://code.launchpad.net/~thomas-voss/location-service/gcc-4.7-by-default [10:44] davmor2: what 3g issue? the fact that you have to restart nm? === oreneeshy_ is now known as oreneeshy === matzipan_ is now known as matzipan [10:54] popey: yeap, that and it only working possibly 90% of the time after a restart. But it's hard to tell till the new code lands what issue is connected to what :) === MacSlow is now known as MacSlow|lunch [11:03] davmor2, no, ask cyphermox_ [11:04] it seems like gdb *always* reports two duplicated frames when trying to do backtrace :-( [11:05] diwic, oh, not sure you saw it yesterday ... http://paste.ubuntu.com/6003386/ was my last pacmd list run after the meeting [11:06] With hope that it could make a difference in the future, I report: [11:06] tvoss: done, interesting that the comment doesn't work this way, will note it down :) [11:06] ogra_, yes, I saw it, it looks good enough [11:07] ogra_, I flashed today's pulseaudio image here on Nexus 4, so I'm investigating how things work today... [11:07] didrocks, yeah, seems to only work if at the beginning of the line [11:07] I attempted to use Indigogo to fund but it is not working. I have JavaScript enabled for Indigogo but not others. [11:07] I suppose I may report it to Indigogo also. [11:09] ogra_, and the pulseaudio process hang when trying to make a voice call, and I can't use gdb to figure out where :-/ [11:09] :( [11:15] stupid proprietary blobs [11:15] you have to dance with them their way [11:16] or they hang or segfault [11:21] ogra_, I flashed a new image (without -b or --wipe), should that have brought in the new bluetooth user [11:21] ? [11:21] yes [11:21] well, not a user, just the froup [11:21] *group [11:22] ogra_, hmm, bluetooth:x:109:phablet [11:22] ogra_, (from /etc/group ) [11:22] no, thats definitely worng [11:22] oh, wait, that image is built on jenkins [11:22] it doesnt use the normal build tools [11:23] someone wit access to that part needs to add thesame change there [11:23] * diwic suspect that's why pulseaudio hangs [11:25] diwic, groupmod -g 1002 bluetooth [11:25] and reboot [11:25] (thats what tteh build change does too) [11:25] * diwic tries === tvoss is now known as tvoss|lunch [11:26] when is the touch phone out? [11:27] it will release together with ubuntu 13.10 [11:31] ogra_, okay, thanks, that worked [11:31] great, so the normal image has the fix [11:31] ogra_, do you know who to contact to have this fixed in the pulseaudio image? [11:32] i think sergiuens maintains that [11:32] (but he's not up yet) [11:32] ok [11:34] seb128: hi! I might be wrong, but I don't think you can comment out a line in a list of items if you have the "\" for the line continuations [11:34] mardy, what do you mean? my a11y panel being commented? [11:35] mardy, in the .pro? [11:35] mardy, oh, and hey ;-) [11:37] mardy, I built a deb here, worked fine... [11:38] ogra_, okay, so it seems phone calls are working here, so we should see if they are working on the Galaxy Nexus too. [11:39] ogra_, you don't need to download an entire new image, but an "sudo apt-get update" and then "sudo apt-get install pulseaudio" (so you get +phablet8 version) would be good. [11:40] Sup [11:41] anyone have an fstab file they can provide on a working system from /data/ubuntu/etc/fstab? [11:41] paste it somewhere? === hikiko is now known as hikiko|lunch [11:45] seb128: ah, ok; I thought it wouldn't work :-) [11:45] here is a question.. my device lacks device.fstab in the device folder, because its setup in the init file it seems... or somewhere else.. is having device.fstab a requirement for ubuntu to find your partitions? [11:45] Or does it go off the /dev/disk/by-id and label only? [11:46] seb128: mmm... I can't believe it works, unless qmake is very bizarre [11:47] mardy, calling me a liar? ;-) [11:47] seb128: lol, no, but maybe the deb does not contain all the plugins :-) [11:47] mardy, I'm happy to change it, but I bzr bd from the tree and the resulting deb built fine and doesn't have accessibility installed [11:48] mardy, http://paste.ubuntu.com/6006131/ [11:48] mardy, that's the debdiff of files between installed version and local deb [11:48] seb128: because you have some characters after the "\", which should always be the last character in the line if you want to concatenate the following line [11:49] mardy, well, are comment lines counted at all in concatenation? [11:49] seb128: right, they are probably excluded [11:49] mardy, but no big deal, just tell me what you prefer and I can change it [11:50] seb128: no no, it's fine, I just didn't expect qmake to be this smart :-) [11:50] ;-) [12:05] diwic, no tone at all on phone calls [12:06] ogra_, you mean ringtone? === MacSlow|lunch is now known as MacSlow [12:06] i mean nothing ... calling the phone doesnt work at all, if i make a call to my landline i cant hear otr speak [12:08] http://paste.ubuntu.com/6006187/ === tvoss|lunch is now known as tvoss [12:08] recent syslog [12:13] ogra_, okay...what's the current ports and profile? Try setting profile to "HiFi" and ports to "[Out] Earpiece" and "[In] Handset" [12:13] ogra_, pactl set-card-profile 0 HiFi [12:13] ogra_, pactl set-sink-port @DEFAULT_SINK@ "[Out] Earpiece" [12:13] ogra_, pactl set-sink-port @DEFAULT_SOURCE@ "[In] Handset" [12:13] oops [12:13] ogra_, pactl set-source-port @DEFAULT_SOURCE@ "[In] Handset" === alan_g is now known as alan_g|lunch [12:15] nope, no change [12:16] i can make calls but dont get any audio through once the connection is established ... [12:16] if i call the phone, the display turns on but there is no ringtone [12:17] (and no way to pick up the call intrestingly ... but it is apparently recieved (since the screen turns on) [12:19] paplay /usr/share/sounds/alsa/Front_Left.wav comes out of the earpiece btw [12:21] ogra_, that's weird...I think paplay and ringtone would use approximately the same mechanism === dednick_ is now known as dednick|lunch [12:21] ogra_, maybe if the earpiece is weak, the ringtone is also weak [12:22] ogra_, do you hear the ringtone better if you set the output port to "[Out] Speaker" ? [12:22] better ? [12:22] i dont have it at all [12:23] hello [12:23] i wonder if the phone app is simply broken [12:24] diwic, does it work for you ? [12:24] ogra_, I have working ringtone here, yes. And SMS notifications too [12:24] ogra_, they just worked so I haven't investigated what audio path they take [12:24] ogra_, the SMS notification is quite weak though, a small tone only [12:25] sms notification works fine ... but again, no sound [12:26] ogra_, hey i have good news. i got upstart to run, i got the dmesg to show on the cellphone screen, and i made a serial cable for my phone lol [12:26] oh, wait, sound out of the earpiece [12:26] but no ringtone :( [12:26] (sms notofication comes from earpiece now) [12:27] annerajb, dmesg to show on the cellphone screen ? then you did to much :) [12:27] make sure fbcon isnt loaded [12:28] ogra_, i had a warning on the kernel build saying to enable something i did and showed the dmesg but appeared to just hang there so i disable it. [12:28] ogra_, is fbcon a process or a kernel option? [12:29] annerajb, a module or fixed option you most likely have enabled [12:29] else you wouldnt have console output on the screen [12:29] ogra_, but anyway, it starts going wrong already here: module-android-audio-hal.c: set_parameters returned error 1 [12:30] yeah [12:30] its also weird that it seems to try to start pulse so often [12:30] ogra_, you mean this? [pulseaudio] pid.c: Daemon already running. [12:30] yeah [12:31] looks like it tries that many times [12:31] Mirv: yes it's critical to the 5.1 migration :) [12:31] Mirv: https://codereview.qt-project.org/#patch,all,63291,1 [12:31] Kaleo, surround apps ? [12:31] :) === hikiko|lunch is now known as hikiko [12:31] oSoMoN: woot [12:31] ogra_: woot [12:32] "the ubuntu UI, now in dolby digital" [12:32] woot woot? [12:32] ogra_: ok........ [12:33] ogra_, it only seems to happen at startup. [12:33] ogra_, something to investigate when I have some time left (never) [12:33] haha [12:34] ogra_, sorry. It should be investigated, but right now I'm just trying to get us ready for the bare minimum so we can switch. [12:34] right, as long as the startup doesnt influence that [12:35] ogra_, if you have a bootchart that shows a lot of pulseaudio processes taking up all of your startup time then maybe we should do somethign about it [12:35] ogra_, my device wrote some filse to /var/log/ among them ureadahead printed something that may seem like a probem [12:35] Counted 0 CPUs === rah_ is now known as rah [12:36] diwic, well, i only see one pulse process after booting ... [12:36] and i havent bootchart set up here atm [12:36] so it should be fine as long as it doesnt block the device [12:37] ogra_, anyway, so you have SMS notifications and normal audio playback, but not ringtone. I wonder if ringtone is done in some different way [12:37] we'll do a lot of bootspped analysis once lightdm is in the image [12:37] well, FSVO normal [12:37] ogra_, yeah, they'll quit themselves after noticing that the daemon is running [12:37] it comes out of the earpiece [12:37] and i cant pick up an incoming call ... display lights up, i can unlock but the phone app isnt there [12:37] ogra_, but if you switch port to "[Out] Speaker" it comes out of the speaker instead, right? [12:38] lets see [12:38] (or hear) [12:38] yup, it does [12:38] ogra_, and always, when you're not on a phone call, you should be on the HiFi profile. [12:38] so let me try another call [12:40] Kaleo: ok, filed https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/1214374 for it. I'm already looking at 5.1.1 a bit, but I'll try out a build of the declarative with that patch. I'll try a build of the current 5.1 with the patch and if I get that building and nothing obviously breaks, I'll copy it to qt5-beta-proper [12:40] Launchpad bug 1214374 in qtdeclarative-opensource-src (Ubuntu) "Include patch to Correct geometry for "@2x" pixmaps" [Undecided,New] [12:40] Mirv: thanks so much! [12:41] so ogra_ this is a cat * of all the files inside /var/log/upstart http://pastebin.com/qmsfPnaA see anything weird? [12:42] you miss a ureadahead patch in your kernel [12:42] not fatal though [12:43] barry: hey, do you have some time for a quick chat? [12:44] ogra_, well my device appears to get stuck there and never opens adb [12:45] ogra_, now that i have serial how can i send the upstart output to the serial console? [12:46] also i could never implement the extend patch on the commandline arguments this is how they are on the kernel "no_console_suspend=1 datapart=/dev/mtdblock3 console=tty1 " at the end the bootloader inserts the serial number === gusch_ is now known as gusch|akf [12:49] I would like to use 60-persistent-storage-dm.rules in the ramdisk at /lib/udev/rules.d [12:49] but it doesn't seem to be going into afect [12:53] ogra_: when will rsalveti be back? [12:53] do you know? [12:54] i thought today [12:54] ChickenCutlass should know [12:54] he was there yesterday [12:54] (incongnito) [12:54] :) [12:54] ogra_, oh, apparently Tuna needs you to specify sample rate. I need to push a new pulseaudio version [12:54] ok [12:58] hey ogra_ [12:59] so ... [12:59] yo [12:59] my wife's N9 (nokia phone) stopped booting [12:59] so I gave her my old android phone [12:59] I am now totally dependent on Ubuntu Touch [12:59] on my N4 [12:59] it's working really well [13:00] a few bugs here and there, but with a little polish in September/October, we're going to have an amazing product on our hands [13:00] ogra_, thanks for everything you do :) [13:00] rickspencer3, heh, i wouldnt do it if i wouldnt love it, and i love it because of the peoople i can work with :) [13:01] ogra_, yup ... a whole lotta people putting in a lot of great work [13:01] yeah [13:01] but every time someone comes here with a question or an issue ... [13:01] you're right there to help [13:01] (and popey too :) ) [13:01] hah [13:02] hi popey :) [13:02] well, we just try to teach the people to fish :) [13:02] rickspencer3: nice! (that you can survive with ubuntu only) [13:02] so, not to diminish anyone else's contributions, but I wanted to thank ogra and popey directly for all their help [13:02] Thanks. [13:02] asac, right, since popey showed me what a dope I was being with the authenticator app [13:02] I'm good to go [13:03] I have started creating a bunch of webapps that I personally use === Ursinha-afk is now known as Ursinha [13:03] made a webapps-demo-extras package which just contains .desktop and .png files really [13:03] rickspencer3: what app/approach are you using for that (2-phase?) [13:03] ? [13:04] tis very easy to do, but would be great if we could easily make webapps from sites [13:04] like the way the iphone does "add to home screen" [13:04] asac, mhall pointed me to an app called CanonicalAuthenticator [13:04] it more or less does what it says on the tin [13:04] ah wasnt that from aquarius ? [13:04] asac: stuart langridge wrote it [13:04] yes [13:04] rickspencer3: thats a touch app? nice [13:04] for use on his n9 ☻ [13:04] my poor N9 === _salem is now known as salem_ [13:05] popey, that phone was running meego [13:05] or was it moblin by then? [13:05] yeah, i do love the n9 [13:05] who cares [13:05] anyway, it was a sweet phone [13:05] tis a lovely piece of hardware [13:05] ++ [13:05] n9 was meego [13:05] or even maemo 6 [13:06] they did a really nice job with it, it was very useful [13:06] of course, Ubuntu Touch is nicer :) [13:09] ogra_, ok, pulseaudio version +phablet9 pushed to the ppa. I guess now we only have to wait two hours for it to build before we can see if it makes a difference [13:09] mpt, " “Last full charge” using the most appropriate unit: for example, “2 hours ago” or “51 minutes ago”" [13:09] bzoltan1, hey, do I still have to "enable developer mode" to run apps from QtCreator? [13:09] mpt, what should be displayed if that was "1h59" ago .. 1h or 2h? [13:09] seb128, 2 hours [13:09] diwic, bah, slow PPA builders ... it would build in 20min on the new machines [13:10] mpt, where is the limit ... what for 1h45? or 1h56? ;-) [13:10] seb128, because it's between 1h30m and 2h30m [13:10] mpt, ok, so it's rounded, not truncated [13:10] mpt, thanks [13:10] ogra_, will this work to get serial console while upstart boots? http://superuser.com/a/463085 [13:11] how can i start a touch app on through adb shell? [13:11] popey: can you confirm something, enable 3g via the indicator and restart n-m, disable wifi connect to a webpage (I recommend typing what's my ipaddress and let google handle the rest ), wait for it to load, re enable wifi, refresh the page, does wifi override the 3g is number one, then disable 3g now try again, now disable wifi and enable 3g again and try refreshing the page now [13:11] annerajb, except that upstart doesnt use inittab ... yeah, the ttyS0.conf job should work (but only if your serial devices is actually ttyS0) [13:11] seems just running the command doesnt work anymore [13:11] asac, are you phablet ? [13:12] ogra_: yeaw h... su - phablet [13:12] then run ubuntu-calc*app [13:12] wont work :P [13:12] you need sudo [13:12] rickspencer3: yes, you do. The new scripts package is not released yet [13:12] aha [13:12] ogra_, so if i had a ttySAC2.conf that should do it? [13:12] ok [13:12] su sets the environment wrongly [13:12] bzoltan1, pmcgowan tells me there is an easier way [13:13] bzoltan1, phablet-network-setup -i ? [13:13] ogra_, so it looks like the application phone-app-approver starts a stream to pulseaudio. When you have an incoming ringtone, does "pacmd list-sink-inputs" show the stream? [13:13] annerajb, make sure to change all of ttyS0 to ttySAC2 inside the job, and your kernel cmdline needs to point to it with the console= arg [13:13] rickspencer3: yes, there is, changing a script [13:13] I never tried phablet-network-setup -i [13:13] ogra_: i used sudo -u phablet bash [13:13] that one doesnt work either, but gives different errors [13:13] ogra_, what do you mean inside the job? [13:13] rickspencer3: I use only the QtC [13:14] bzoltan1, right, but that takes hours installed tons of build tools I never use [13:14] ogra_, upstart job? [13:14] asac, adb shell sudo -u phablet -i sh -c "calculator-app" [13:15] rickspencer3: you do not, but others might... releasing a new version with disabling all those what you do not like might break hundreds of SDKs out there [13:15] annerajb, the .conf file is an upstart job [13:15] rickspencer3: the "enbale development mode" does a bit more than just ssh authentication... even thou that is all you need as long you write simple QML apps [13:15] bzoltan1, remember what we talked about, we just need an option to enable ssh separately [13:15] so i rename all of the ttyS0 (inside of them to point to the same console) [13:15] bzoltan1, thats all I need to use qtc with qml apps [13:16] pmcgowan: of course... it is on the backlogs [13:16] ogra_: doesnt start the app either here :) [13:16] ogra_, btw can you verify my commandline is fine? since i recally you sanying console must be the last thing or something siilar when i boot the phone the serial number is always at the end [13:17] asac, adb shell sudo -u phablet -i sh -c "stop unity8;ubuntu-terminal-app && start unity8" [13:17] try that one then [13:17] pmcgowan: I can push the button to release the new "enable development mode" script what does the job only for QML apps ... it is a second. But the next day nobody will not be able to use QtC for developing QML plugins for example [13:18] bzoltan1, why don't you just make them 2 different commands? [13:18] bzoltan1, right its a new option [13:18] or even just run the script when the user chooses Run on Device? [13:18] enable ssh [13:18] asac: if you're using a mir image, you'll need to run apps with desktop_file_hint, like this "ubuntu-calculator-app --desktop_file_hint=/usr/share/applications/ubuntu-calculator-app.desktop" [13:19] asac: else unity8 will reject the application, and app will crash [13:19] mpt, should 29h be "29h ago" or "1 day ago"? if it's the second, does it change to 2 days after 36 hours? [13:19] pmcgowan: I can do that... but that will need a touch on the QtC UI what trigers the whole release cycle. [13:20] seb128, yes and yes ... There really should be a library function for this. :-) [13:20] mpt, there should but there isn't... ;-) [13:20] bzoltan1, is it not a change to the plugin? [13:20] rickspencer3: starting apps on the device is slow enough, chekcing for openssh packages before every app start would slow it down [13:20] mpt, thanks [13:20] bzoltan1, what are you saying? [13:20] pmcgowan: The UI of the QtC is not [13:20] you have to run the command at some point anyway [13:21] may was well run it when the user obviously needs it [13:21] rickspencer3: Yes, but preferable not too often [13:21] instead of not working until the user discovers the magical tab and the "actually works" button [13:21] seb128, the same library that contains a function for stripping Unicode direction characters from untrusted strings before displaying them in UIs. :-) [13:22] heheh [13:22] rickspencer3: I did not want to just install ssh server on the device automatically. We can do it if that is what you want. [13:23] bzoltan1, what do you think users want? [13:24] rickspencer3: I think I know what they want :) but checking for the openssh on the device before every app start would be an unacceptable overhead in my opinion. [13:26] rickspencer3: The solution will be to offer two kind of developer mode: 1. light one for only QML apps 2. heavy weight for those who need more ... these are going to be two buttons on the Device Tab [13:26] bzoltan1, why would you do it every time? [13:26] that makes no sense [13:26] if you run the command and it fails [13:26] rickspencer3: (2013-08-20 16:18:27) rickspencer3: or even just run the script when the user chooses Run on Device? [13:26] bzoltan1, c'mon, think about it [13:26] QtCreator knows that the command didn't work [13:27] so you have a clear signal [13:27] rickspencer3: I udnerstand [13:28] rickspencer3: do you want that QtC installs ssh on the device without user confirmation? [13:28] greyback: i am using the normal image [13:28] cdimage-touch [13:28] bzoltan1, what do you think users would want? [13:28] greyback: should i switch to MIR image? [13:28] asac: you've installed ppa:phablet-team/mir ? [13:29] greyback: grep -r mir /etc/apt/sources.list.d/ -> zero hits [13:29] rickspencer3: I think they want to know that an ssh backdoor is just set up on their device [13:29] rickspencer3: but I could be wrong.. I do not maintain strong opinion on that [13:29] bzoltan1, so then how would design the feature so that users can be successful but also have that knowledge? [13:30] asac: ok, then you're not using the updated unity8 etc. Does unity8 come up at all? [13:30] rickspencer3: direct them to the Device Tab ... they need to clone the network anyway manually [13:31] bzoltan1, may suggest that you offer it more directly [13:31] rickspencer3: For the same security reason I did not want to clone the network profile without confirmation [13:31] don't tell them where to go, just offer to do it for them [13:32] rickspencer3: That is a fair suggestion. That will need a bit more interactive QtC plugin ... [13:32] rickspencer3: like when a script fails we offer help in a form of dialog. Doable. [13:32] asac: phablet-team/mir ppa contains a build of unity8, plus its unity-mir dependency, plus ubuntu-touch-session which tweaks some things. But with that, you have working unity8, plus applications [13:32] greyback: i see the launcher and can use it [13:32] yes] [13:34] asac: is this env var set "QT_QPA_PLATFORM=ubuntumirclient" ? === dednick|lunch is now known as dednick [13:35] ogra_, http://pastebin.com/FAcFVRx8 line 412 forward you see anything strange there with ureadhead ? [13:35] asac: but I'd recommend installing the PPA, as it contains what will land for unity on mir [13:35] greyback: QT_QPA_PLATFORM=ubuntu [13:36] so yeah. guess i have not migrated to the MIR/unity8 image [13:36] asac: that image is still using surface-flinger. Adding that PPA will convert it to Mir only [13:37] greyback: right. but for now i just want to start an app from a command line :) ... you say that i need to have MIR in order to do that nicely nowadays? [13:37] annerajb, looks fine to me ... the ureadahead message is normal if you dont have the kernel patch === alan_g|lunch is now known as alan_g [13:37] ogra_: where can i get this MIR image? [13:37] asac, no idea, likely on the broken jenkins machine :P ... ask ricmm [13:37] ok... nevermind. i guess i will wait with my experiment a few days then [13:38] asac: no, launching an app on standard image /should/ just work. I don't know what's wrong there === rachelliu_ is now known as rachelliu === robbiew1 is now known as robbiew === Ursinha is now known as Ursinha-afk [13:54] boiko, gusch|akf, could I ask a favor from either or both of you? Could you please provide a review for https://codereview.qt-project.org/#change,63026 to get it going through the review process? [13:55] fginther: yep sure [13:55] boiko, thanks! [13:56] fginther: in a meeting right now, will do after that === gusch|akf is now known as gusch [13:57] boiko, no rush === Ursinha-afk is now known as Ursinha === chihchun_afk is now known as chihchun === jono is now known as Guest90491 === greyback is now known as greyback|needsfo === greyback|needsfo is now known as greyback|food [14:14] ogra_, at what point during boot is adb started? [14:15] ogra_: I have an app (uReadIt) that won't launch from the Dash anymore, any suggestions on how to figure out why? [14:16] rickspencer3: a quick relief for you -> https://code.launchpad.net/~bzoltan/qtcreator-plugin-ubuntu/lighter_dev_mode_enabler/+merge/181052 [14:17] or maybe I should ask Saviq [14:17] Saviq: I have an app (uReadIt) that won't launch from the Dash anymore, any suggestions on how to figure out why? [14:17] rickspencer3: it is not a final solution, but will easy the process [14:18] mhall119, nothing off the top of my head [14:18] mhall119, how can I try? [14:19] mhall119, can you see any interesting output in ~/.cache/upstart/unity8.log ? [14:20] Saviq: Unable to activate "ureadit.desktop" [14:20] hmmm, maybe because I have capital letters in my .desktop? [14:21] mhall119, capital letters? [14:21] uReadIt.desktop [14:21] no ureadit.desktop [14:21] annerajb, look in the upstart job in /etc/init/ ... i dont remember [14:21] not [14:22] mhall119, yes, that's most probably it, but sounds like a bug nevertheless [14:22] ogra_, ok thanks [14:22] yup, that was it [14:22] mhall119, something somewhere lowercases the name [14:22] mhall119, can you please file against unity8, we'll see where to redirect it [14:23] sure [14:27] Saviq: https://bugs.launchpad.net/unity/+bug/1214428 [14:27] Launchpad bug 1214428 in Unity "Apps with capital letters in their .desktop file name won't launch" [Undecided,New] [14:27] mhall119, thanks [14:34] mhall119: are you back on flipped images or are you using system image for your devices? === greyback|food is now known as greyback === alan_g is now known as alan_g|tea [14:44] hey Kaleo do you have a minute ? [14:45] mpt, question about the power panel, what should happen the screen brightness slider when the screen is on autodim? should it disactive autodim if you move it? be disabled? [14:46] mpt, or nothing? e.g let the user change it even if the system set it back to the judged best value just after that? [14:47] seb128, ah yeah, I was going to draw a graph to show how that would work [14:47] popey: I went back to flipped [14:47] hm, okay [14:47] popey: the partition scheme on the system image meant that there wasn't enough room to use apt-get even after enabling developer mode [14:48] oof [14:48] greyback, I posted some images regarding the mail app issue. did you see them? should i report this as a bug? http://www.phabletusers.com/t/cant-get-back-to-inbox/19/3 [14:49] seb128, basically, if you adjust the brightness manually when it's on Automatic, that should adjust the Automatic curve -- automatic brightness = f(ambient light) -- so that it goes through your manually set point. [14:49] mpt, hum, k ... [14:50] sforshee, ^ is powerd handling anything like that? (I've no idea how the auto value works on touch) [14:50] seb128, but I hadn't thought much about what happens at the extremes ... Maybe it should be a curve drawn through your last three manually set points [14:50] seb128, it's times like these I wish I was better at maths. :-) [14:50] chrisd_: yes, please log bug, that simply isn't right [14:51] greyback thanks [14:51] seb128, but the simple answer is, the slider should remain sensitive [14:52] mpt, ok ... I've no idea how the auto adjustement works, but I've the feeling it's not going to do the complex things you wish [14:52] seb128, mpt: I'm working on adding autobrightness support to powerd, but the situation with the android kernels is a bit complicated [14:52] mpt, e.g settings the slider on a point is going to change the current value [14:52] that's about it [14:53] chrisd_: thanks to you [14:53] seb128, mpt: the lux values from the ALS don't map to any standard range, and each device supplies a lux->brightness mapping in an xml file [14:53] hmm, my /usr/share/qtcreator/ubuntu/scripts is empty. where'd my sdk go? [14:54] seb128, sure, I understand that. The point is to remember what your manual setting was for a given level of ambient light, so that whenever you have that level of ambient light in the future, it's automatically set to the same level. [14:54] seb128, mpt: in the short term I'm just trying to get this basic functionality working, so I don't expect to implement anything like what was described right now [14:54] Kaleo: i am experience a bug with the Pagestack component, https://plus.google.com/u/0/photos?pid=5912883833171629858&oid=101694416703170881163 and here is the code to reproduce it http://paste.ubuntu.com/5994664/ [14:55] sforshee, do ambient light sensors have a maximum detectable brightness? If so, do measurements get anywhere near that maximum in real-world use? [14:55] mpt: undoubtedly they do, the brightness must come from the hardware in registers with limited number of bits after all [14:56] mpt: but the specific value will vary with the hardware [14:56] and the binary blob used [14:56] the values from the drivers already vary between devices for the same ambient conditions [14:56] (if there is one) === alan_g|tea is now known as alan_g === alecu_ is now known as alecu === freeflying is now known as freeflying_away [15:10] <_5m0k3> Do the buttons in the Installed section of the Applications tab (on ubuntu touch) not function on devices that I've written and pushed to the device? I can open the app from the terminal, but not from the launcher [15:12] ogra_, how do you define the order how upstart starts process? [15:12] by the "start on/stop on" stanzas [15:13] ?? [15:13] inside the jobs [15:13] mhall119: is that a bug? [15:13] oh you mean [15:13] start on filesystem or runlevel [2345] [15:13] yeah [15:13] filesystem is where it's going to start at? [15:14] how can i move it to a earlier stage? where do i get the list of stanzas or stages? [15:14] that means either if the filesystem is mounted or if runlevel XYZ is reached [15:14] start on startup [15:14] thats the earliest you can start something [15:15] well it's definetly not reaching that point :( [15:15] but you need to take requirements of your app into account [15:15] since i dont get a adb shell [15:15] did you read the upstart log for it ? [15:15] there is no log. [15:15] it should show why it cant start [15:15] ev, do you have a sec to check the 19.2 image with mir? iirc you also had seen mir not come up [15:15] tvoss, kgunn, rsalveti, ^ [15:15] olli_: can do! [15:16] ogra_, none of the files in /var/log/upstart give a error about anything. [15:16] olli_: this one? http://s-jenkins:8080/job/ubuntu-touch-phablet-image-saucy-mir/lastSuccessfulBuild/artifact/saucy-preinstalled-phablet-armhf.zip [15:16] ev, not sure, rsalveti? [15:16] I reflashing 19.2 with -b to have a clean one, and will try to reproduce, but could only get the failure with aug 5th [15:17] also, did you test mir after disabling surfaceflinger? [15:18] ev: http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20130819.2/ [15:22] popey: is what a bug? [15:22] mhall119, has anybody expressed interest in the Oppo Find 5 contest yet? [15:22] _5m0k3: did you install the app on the device, or just running it on the device via QtCreator? [15:22] smartboyhw: we've had several people take a crack at it [15:23] mhall119, all failed?:O [15:23] evidently someone had the Raring images booting on it back in April, but they're having trouble with Saucy [15:23] Is there a working alarm clock yet? [15:23] smartboyhw: one of the XDA devs is going to be here for dholbach's porting clinic on Thursday, hopefully it'll get claimed then [15:23] <_5m0k3> mhall119: It's a qt/c++ app, so i built and installed on device [15:24] mhall119, oh wow;) [15:24] Tragyn: not yet, the alarm API just recently (like, yesterday) landed so the clock developers can use it [15:24] _5m0k3: out of curiousity, what's the .desktop file name? [15:24] Well, I would want to try once, but deeming to the current situation that you guys keep failing on making it work, hmm [15:24] <_5m0k3> BlackJack.desktop [15:25] _5m0k3: rename it to blackjack.desktop [15:25] _5m0k3: same bug I ran into: https://bugs.launchpad.net/unity/+bug/1214428 [15:25] Launchpad bug 1214428 in Unity "Apps with capital letters in their .desktop file name won't launch" [Undecided,New] [15:25] <_5m0k3> Much appreciated [15:25] mhall119: thanks. Wasn't sure as I've seen videos/screenshots of other people with Nexus 4's who seem to have a later version of the OS. [15:25] mhall119, BTW how did it fail to boot? [15:26] smartboyhw: "you guys" means XDA community,since nobody at Canonical is working on a Find 5 image [15:26] mhall119, oh [15:26] mhall119, but you have the phone to try right? [15:26] How did it fail? [15:26] smartboyhw: not sure, when I was at the XDA DevCon someone said it had somethign to do with the init files and the build script [15:26] smartboyhw: yes, I have a phone [15:27] <_5m0k3> mhall119: Worked like a charm. Thanks [15:27] the init files and the build script. Got it. [15:27] _5m0k3: cool, hopefully that'll be fixed soon too, since it used to work [15:27] Saviq: so it unity8 using upstart to launch apps now? [15:29] sergiusens: ping [15:29] tsdgeos: pong [15:29] rsalveti: I had been (presumably) disabling it by overriding init.rc, yes [15:29] sergiusens: how do i install image 20130819.2 using phablet-flash? [15:29] it failed both before and after I did that [15:30] right [15:30] smartboyhw: I don't know much about porting, so I didn't understand what they were saying [15:31] smartboyhw: but the impression I got was that the standard build script(ubuntu or cyanogen I'm not sure) wasn't putting the right init files in the right place, which was preventing Ubuntu from booting [15:31] mfisch, mhr3_: Poked jodh on the upstart settings we could use; he pointed out there's a --confdir flag to extend the list of dirs with upstart jobs, but for system jobs it's an override and needs to be on kernel cmdline which is hard for custom channels (since cmdline is in the boot.img) [15:31] mfisch, mhr3_: Here's what he said: [15:31] 17:12 setting vars for user sessions is already handled. See http://upstart.ubuntu.com/cookbook/#initctl-commands-summary, specifically 'initctl set-env --global'. [15:31] 17:13 having Upstart read jobs from multiple directories is also handled but for user sessions (init --user) only atm. See "--confdir" in init(8). [15:31] tsdgeos: ping [15:32] mhall119: yes? [15:32] lool: and I am in the middle of having the same conversation [15:32] lool: we will need a system level job to run dconf update [15:33] mhall119: space on the device after enabling developer mode [15:34] nerochiaro: hi~ [15:34] lool, mfisch, bind mount? [15:34] sil2100: hello [15:34] nerochiaro: hah! I see you've been cheating in lp:ubuntu-ui-extras ;) [15:35] sil2100: oh really ? ;) [15:35] nerochiaro: there is an autopilot package but the test is empty ;p! It's just asserting true on true! [15:35] everything is possible with bind mounts :) [15:35] nerochiaro: anyway, can you create some real integration tests till the end of this week? [15:35] sil2100: oh, that was just part of the example component, to give a skeleton to people who want to add their own components. there's really nothing that can be tested in there [15:36] sil2100: in fact now that i added the share component i can probably remove the example component [15:36] sil2100: and the share component doesn't really anything that can be autopilot tested either, but it has some unit tests already [15:36] mhr3_: lol [15:36] mhr3_: can we add a separate dir under /etc/init? [15:36] will upstart use it I wonder [15:37] nerochiaro: ah, ok then, good to know [15:37] mfisch: what we could have is a single pre-session start job that does a run-parts on stuff in /custom [15:37] mfisch: basically, don't start the session until these scripts from /custom have been run [15:37] nerochiaro: then I'm not adding any integration tests to be ran - I guess the autopilot package and the autopilot directory can be removed soon [15:38] concerning user environment, this might be a bit nicer to handle; it might just be a case of adding a confdir via some ligthdm switches or something [15:38] mhall119, not yet, but with Mir it will [15:38] lool: you dont want to have a smaller system level job and then whatever else gets copied into ~/.config/upstart? [15:38] mhall119, initially we'll support the usual --desktop_file_hint [15:39] mhall119, but when we build the machinery between upstart and unity8, we'll switch to an upstart-only model [15:39] mfisch: yeah, that would be a way [15:39] mhall119, where upstart will authorize an application with Mir to be able to display anything [15:39] nerochiaro: also, I gave to review to the other packaging guys the final packaging review: https://code.launchpad.net/~sil2100/ubuntu-ui-extras/packaging_review/+merge/181075 [15:40] sil2100: excellent. to clarify, do you need me to remove the example component or just its autopilot tests ? [15:40] mfisch, mhr3_: So let's call this pre-session-custom-scripts.conf and have a system upstart jobs that run-parts /custom/pre-session.d/* if present [15:41] and then starts lightdm [15:41] lool: +1 [15:41] lool: I didn't get a sense of where to change DCONF_PROFILE from jodh though [15:41] mfisch: so one way would be to set it in the session environment via some early session upstart job [15:42] or in fact, since this isn't user specific we could have it set system-wide for touch images [15:42] lool: right it [15:42] lool: right it's not user specific [15:43] nerochiaro: hm, I guess my only problem is the autopilot package - since it gives the impression that it's 'autopilot tests for the package' [15:43] sil2100: so is it something i should remove before it can be approved ? [15:43] nerochiaro: so I guess just the autopilot tests remove is enough, so removing the -autopilot package [15:44] it's perhaps a bit ugly to set DCONF_PROFILE system-wide for all jobs, but that would work [15:44] nerochiaro: it would be best if it was removed before we release [15:44] mfisch: yes, you can add subdirs below /etc/init and upstart will load those jobs. Just be careful to specify "start subdir/job1" for example when referring to /etc/init/subdir/job1.conf. [15:45] jodh: oh that's pretty cool [15:45] sil2100: ok, i'll send up an MR to do that [15:45] Thanks! [15:46] lool: so we could bind mount, but the pre-start custom job is cleaner I think [15:46] yeah, I was trying to think of a non-ugly way to leverate a /etc/init/custom [15:46] maybe if we could declare this from the custom tarball [15:47] mfisch: do you know from stgraber if we can declare extra bind-mounts we need for /custom in the custom tarball? [15:47] mfisch: e.g. we would create an empty /etc/init/custom/ and bind-mount /custom/etc/init/ to it [15:47] (completely unrelated to the rw-bindmounts currently handled in new images) [15:47] lool: I haven't tried [15:48] ssweeny: did you ask stgraber about bind-mounting /etc/default once? [15:48] mfisch: anyway, the initial idea seems safest to implement this week [15:48] lool: agreed [15:48] it's less generic, but it's at least doable [15:49] rsalveti, olli_: it works! I have a unity8 process and a UI to play with. [15:49] much appreciated [15:49] phablet 1695 8.3 0.5 83436 10492 ? Sl 15:48 0:01 /usr/bin/python /usr/share/unity-scopes/scope-runner-dbus.py /usr/share/mockmusic/mockmusic.py [15:49] Saviq: have the way we launch applications changed in the last month or so? [15:49] lool: so the main point of that job will be to run dconf update, anything else? [15:49] ^ srsly? [15:50] Saviq: on the phone. is QtUbuntu not used anymore? [15:50] mfisch: we need to worry about where dconf update will write its files; otherwise, seems good [15:50] greyback: ^ [15:51] mfisch: Ah I just realized we don't even need another upstart job to set DCONF_PROFILE, we can leverage the first one to run a script that will initctl set-env [15:51] Wellark: yes it is [15:51] lool: in my plan it wrote out to /var/custom [15:51] lool: ok [15:51] since we don't really need to arrange for the dconf profile to be set before lightdm starts [15:51] I mean, it's enough if it's set just before lightdm starts [15:52] right [15:52] greyback: ok. I'm investigating some HUD breakage and at least can't get any results from the gallery-app [15:52] so I suspected that if the way we launch apps has changed then the new launching entity might not set the $APP_ID [15:53] lool: I'd like to keep the option still to copy in user jobs so that we can do stuff like make dbus calls as the user [15:53] mfisch: should we start a simple touch specific package to enable these /custom hacks? [15:53] Wellark: no such change has been made [15:53] lool: I've been thinking about that, a package is probably the best bet [15:53] greyback: ok. good. one thing down from the checklist then [15:53] lool: want me to start one with what I have? [15:53] mfisch: for user jobs, I'd think we want to add a lightdm hook to add a --confdir when launching init [15:53] lool: rather than a copy? [15:53] lool: I guess if we can get it in this week its ok [15:54] mfisch: yeah; initially I thought we'd dump pieces in d-conf itself, but it's actually a much better idea to put it in a specific package that we only install in touch images [15:54] lool: well it [15:54] lool: well it's supposed to be a tarball [15:54] that way a) we don't pollute / put at risk other images b) can ask for a standing FFE on that package without disrupting release of other flavors [15:54] lool: but the tarball can come from the package later [15:54] mfisch: sorry, I meant for e.g. the extra upstart job [15:54] lool: ah, sorry yeah [15:55] didn't really think about the best way to generate the tarball [15:55] tsdgeos: sorry for late reply [15:55] probably just bzr branch + jenkins job I guess === tvoss is now known as tvoss|afk [15:55] tsdgeos: was in a meeting [15:55] no worries [15:55] lool: I dont know enough about jenkins, but we can lay the files out in bzr easily [15:55] diwic, FYI, no change in call behavior with new pulse (no matter what i select as profile or device) i'll do a re-flash [15:56] mfisch: ack [15:56] ev, you might want to grab thostr for that, it (python) will go away [15:56] tsdgeos: phablet-flash cdimage-touch --ubuntu-path http://cdimages.ubuntu.com/ubuntu-touch/daily-preinstalled/20130819.2/saucy-preinstalled-touch-armhf.zip --device-path http://cdimages.ubuntu.com/ubuntu-touch/daily-preinstalled/20130819.2/saucy-preinstalled-touch-armel+[device].zip [15:56] ah, so they have [15:56] ogra_, ehm, that's not expected, what's the syslog like now? [15:56] lool: I'll probably push what amount i have today to bzr so ssweeny and sfeole can start adding stuff [15:56] mfisch: gtg to another meeting, but seems we have a good next iteration [15:56] lool: ok [15:56] olli_: was somewhat worried that they were long living python processes [15:56] sergiusens: wow :D [15:56] mfisch: will you need me to review or upload stuff? just ping [15:57] lool: are you still doing a BP so we can split this up? [15:57] mfisch: yes [15:57] Hey guys. I've seen videos/screenshots of other people who have an extra "tab" at the bottom for "People". Right along where music, home, apps, and videos are. Is that available? === chihchun is now known as chihchun_afk [15:58] ev, iirc, scopes-runner-* is just a wrapper to work around some life cycle issues with scopes, tagged to be addressed post 13.10 [15:58] iirc [15:58] * ev nods [15:58] cheers [15:58] ogra_, oh...stupid me, I forgot to change from AUDIO_FORMAT_DEFAULT to AUDIO_FORMAT_PCM_16BIT [15:58] ogra_, apparently the Tuna audio HAL must have exactly the latter [15:58] tsdgeos: there is no cdimage backend/manifest that can guarantee file formats [15:58] ogra_, argh, there goes another 1-2 hours [15:59] diwic, ah, well ... [15:59] tsdgeos: I got tired of playing catchup and doing if < 2013XXXX file is X else Y [15:59] it will all get better if you can upload to the archive again .... the builders are 4x as fast [16:00] ogra_, I can upload to the saucy archive but that'll break the builds for i386 and amd64 at the moment [16:00] mfisch: https://blueprints.launchpad.net/ubuntu/+spec/touch-customizations as a start [16:00] ogra_, so I don't think people are going to be happy about that [16:00] diwic, who cares ... unimportant arches [16:01] :) [16:01] ogra_, but I need it to flash my arm device! [16:01] :-) [16:01] through pulse ?!? [16:01] :) [16:01] ogra_, or does phablet-flash work on arm too? :-) [16:01] nerochiaro: once you're merge is ready, I'm enabling it for daily, so waiting ;) [16:02] sil2100: sorry, got caught up, i'll do that in a minute [16:02] tsdgeos: but don't forget -b on the end as well...i think that was the needed step [16:02] "fresh" [16:02] diwic, it surely does, i use it from my chromebook all the time :) [16:03] diwic: it works on my chromebook [16:03] diwic: but not as the target ;-) [16:03] but flashing through pulse indeed avoide to have to install that extra package :) [16:03] *avoids === Ursinha-afk is now known as Ursinha [16:06] ogra_, ok, pushed to the ppa, now dinner time [16:06] enjoy [16:06] sil2100: is it ok if i leave an empty tests/autopilot even if there's no actual tests in there ? or is it better to remove the entire autopilot directory for now ? [16:07] sil2100: and i guess the debian package as well [16:07] Hey guys. I've seen videos/screenshots of other people who have an extra "tab" at the bottom for "People". Right along where music, home, apps, and videos are. Is that available? [16:07] Tragyn, i think thats gone since a while [16:08] nerochiaro: let's remove the whole tests/autopilot directory for now [16:08] there is a separate app for it now [16:08] We can re-add it later when there is need [16:11] Oh, ok. Just looked like they had more stuff. For example, on the left side-bar, when you scroll over an icon/launcher, a label pops up. [16:11] Maybe that's gone, too. [16:11] i think it is [16:11] i havent seen it in a while [16:12] barry: FYI, just added some precisions: https://wiki.ubuntu.com/ImageBasedUpgrades/Client?action=diff&rev2=20&rev1=19 otherwise perfect! [16:13] sil2100: so the deb package for autopilot tests starys but it is empty ? [16:13] didrocks: looks great, thanks! [16:13] nerochiaro: no no, let's remove it ;) [16:13] nerochiaro: since it wasn't released yet, and releasing empty packages are a bad idea - so removing tests/autopilot and the -autopilot package [16:15] sil2100: https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-extras/ubuntu-ui-extras-no-ap-tests/+merge/181091 [16:16] barry: can't wait for playing with the mock then! Now that we have the right API I think :) [16:17] nerochiaro: thanks! Sorry that I bothered you with this, it's because of the meeting - reviewing it in a moment ;) [16:18] sil2100: no worries. thanks to you for helping [16:19] The Nexus status spreadsheet is almost a month old. Is there a more current one? [16:19] nerochiaro: modified it a bit and now approving! [16:23] sil2100: excellent [16:33] ev: so is mir always working for you now? [16:33] so far, yes [16:34] Tragyn: http://bit.ly/12AQV53 there are images build tests now :) [16:35] Thanks davmor2 [16:35] ev: rsalveti ....i was also trying...so i flashed the 19.2 specifically with -b [16:36] and now i am following the "Easy way" here https://wiki.ubuntu.com/Touch/Testing/Mir [16:36] is that all correct ? [16:36] it says easy, not correct :P [16:36] wondering about the dist-upgrade (do i really want all packages updated) [16:36] ogra_: :) [16:37] kgunn: that image should work, but I didn't yet test that [16:37] ricmm: should know better [16:37] all I'm testing is 19.2 with the mir packages instead [16:37] oh, 'easy way', not easiest [16:37] :-) [16:38] hi everybody [16:39] i am with a problem to the install ubuntu touch in Asus Nexus 7 (mobile). Can someone help me? [16:39] rsalveti: do you mean you're testing 19.2 with mir in-built ? (sorry confused) [16:39] or are you having to add mir packages on top [16:39] kgunn: yeah, just testing mir_demo_server with 19.2, with the package already available in the archive [16:40] after disabling surface flinger [16:40] but following that wiki should give you all you need [16:40] mir demo server working every time now, so will stop working on this and get back to the possible races between udev/ueventd === ChickenCutlass is now known as ChickenCutlass_l [16:41] rsalveti, anything you did to make it work ? === ChickenCutlass_l is now known as ChickenCutlass [16:41] ogra_: nothing, works all the time with 19.2 [16:41] or just another heisenbug ? [16:41] heh, k [16:42] should be the race, or just heisenbug indeed [16:42] maybe the stars are aligned differently now [16:44] diwic: ogra_ this will have the guid change for bt once built http://10.97.2.10:8080/job/ubuntu-touch-image-saucy-pulseaudio/31/ [16:44] rsalveti: wasn't the magic seemingly -b (making it fresh) [16:45] rsalveti: 19.2 working for everyone? [16:45] too many eyes observed the bug [16:45] tsdgeos: did it work for you ? [16:45] kgunn: on it [16:45] sergiusens, if i now knew if that is what i'm pulling since 20min ... i indeed only picked latest [16:45] ricmm: yeah [16:46] ah, its still building [16:46] well, it did crash with aug 5th, but didn't yet crash with 19.2 [16:47] sergiusens: can you refresh me on using upstart to start something like sudoku? I can't remember what the format of APP_ID is supposed to be [16:48] plars: ^^ you know? [16:50] kgunn: doesn't seem to be working here :-/ [16:51] tsdgeos: even after flashing with -b? [16:51] doanac`: I think we sorted out something last week after lots of trial/error, but then there was the issue of it starting behind unity. I don't recall for sure, but I think it was something like "com.ubuntu.sudoku_sudoku_0.4.2 [16:51] rsalveti: yes [16:51] doanac`: or something like that [16:51] can you paste the output of /system/bin/logcat? [16:51] doanac`: tedg maybe would know? ^ [16:53] rsalveti: sure give me a sec [16:53] rsalveti: didn't work for me either [16:54] kgunn: can you also paste the output of logcat? [16:54] lool: do you know if there's agreement yet on the path? /custom? I've heard /var/custom' [16:54] doanac`: let me send you the email I sent to thomi [16:54] lool: that's what I've been using [16:55] stupid adb shell [16:55] copied all in one line :D [16:55] http://paste.ubuntu.com/6007099/ [16:55] not very useful [16:55] mfisch: /custom [16:55] * tsdgeos ssh's [16:55] rsalveti: the thing mentions surfaceflinger for some reason [16:56] but i've followed https://wiki.ubuntu.com/Touch/Testing/Mir#Switch_from_SurfaceFlinger_to_Mir [16:56] tsdgeos: run as root [16:56] tsdgeos: your paste is empty [16:56] rsalveti: that is as root [16:56] sergiusens: is not empty, it's a very long line as i said [16:56] tsdgeos, adb shell /system/bin/logcat -d | pastebinit [16:57] tsdgeos: yeah, I see it at the end now... :-) [16:57] tsdgeos, that will auto submit it and have proper linebreaks [16:57] oh, it's all in the same single line [16:57] sergiusens, ack, thanks [16:58] rsalveti: sergiusens: http://paste.ubuntu.com/6007109/ [16:58] ogra_: tx [16:58] tsdgeos: better [16:58] rsalveti: mine looks corrupt....i gotta run for a bit === alan_g is now known as alan_g|EOD [16:59] rsalveti: if the servicemanager is waiting for surface flinger i am guessing that can be the camera service toots [16:59] rsalveti: tsdgeos indeed, ti may be the camera..... E/mm-camera( 517): Cannot init Gesture library [16:59] yeah, but that's fine [16:59] rsalveti: fine as fix later? [16:59] since surfaceflinger is disabled by hand [17:00] rsalveti: are we doing android builds for the mir image? [17:00] sergiusens: fix later [17:00] rsalveti: we might want to start a branch for it and remove all traces of flinger [17:00] tsdgeos: so mir is running fine in there [17:00] will be nice for pulse stuff too [17:00] it's not crashing [17:00] well [17:00] that's been forever [17:00] it's always run well [17:00] i even get a nice 60fps counter [17:00] yup, but we don't need a parallel build, we cna work on disabling it by default [17:00] i just don't see anything on screen [17:01] lool: some of these jobs will be run once, I'll leave that up to the jobs to figure out on their own [17:01] that's another issue :_) [17:01] another ? [17:01] that's the only issue i've ever had [17:01] tsdgeos: yeah, sf is not related to your issue [17:01] bug https://bugs.launchpad.net/mir/+bug/1211694 [17:01] Launchpad bug 1211694 in Mir "Black screen on Nexus4" [Critical,In progress] [17:01] comment #4 [17:01] that was the crash I was getting with an older image [17:02] ok, so someone hijacked my bug? [17:02] can you just start mir_demo_server and then the triangle example? [17:02] mfisch: ack [17:02] wonder if that's powerd related (like turning off the screen) [17:03] tsdgeos: right, getting a black screen now as well with 19.2 [17:04] * tsdgeos needs to EOD, stuff to do and 1 hour past my time already [17:04] if you guys need want me to do something on my tomorrow morning, send me an email or comment in the bug [17:04] * tsdgeos waves === wellsb is now known as _5m0k3 [17:21] lool: can I have edit access to the BP? [17:23] mfisch: oh wow [17:23] mfisch: you can't edit whiteboard? [17:23] I thought everyone registeredcould [17:23] lool: one sec [17:23] lool: sorry I've been having LP issues, I was signed out when I loaded the page [17:24] np [17:42] the hibernate/lock screen with the circle that says "No Data sources available" what is that for and is the configurable for anything yet? [17:43] dixeflatline: take some photos ☻ [17:44] hmmm camera isnt launching... [17:44] what if i added photos too ~/Pictures [17:45] are you guys interested in mir bug reports for ported devices yet? Or should I wait to see if it starts working until it lands officially? [17:46] lool: start on starting lightdm won't work, at least not today [17:47] w-flo: i dont see a problem with a beta bug reports [17:48] think that would help getting more people involved [17:49] well, maybe support for strange/old/random devices is not even expected at this time [17:49] Or additional, device-specific work is required that I haven't done [17:49] add-apt-repository ppa:pahablet-team/ppa is this still a valid repository? [17:50] hmmm... strange and old? [17:50] strange, yes! [17:50] old, no [17:50] m2c [17:51] jdstrand: hey, you around? [17:52] ralsina: do you happen to know when the Click scope will be able to actually install a click packaged app on the device? [17:52] mhall119: "since friday" :-) [17:52] mhall119: it's in the main apps scope right now [17:53] mhall119: however, the two apps that are available in the production server (including yours) have an error in the manifest, and will install but not run [17:53] mhall119: beuno was working on a re-upload fix earlier today [17:53] sorry [17:53] I didn;'t follow through [17:53] ralsina: I think I fixed the XDA app, but it doeesn't appear to have installed when I clicked Install [17:53] ralsina, could you tell mhall119 what he needs to change? [17:53] and I'll fix my app [17:54] sergiusens: hey, yes [17:55] jdstrand: I guess I'm using the old format here http://pastebin.ubuntu.com/6003148/ for the manifest, but removing the top level app_name has the same effect, any ideas? [17:55] beuno: I'll ask alecu to do it since he's the one that knows best. Alecu? ^ [17:57] sergiusens: it should look like this: http://paste.ubuntu.com/6007303/ [17:57] sergiusens: well, except that you need to specifiy "networking" as a policy group [17:58] jdstrand: this is what I have http://paste.ubuntu.com/6007305/ [17:58] sergiusens: yes. that is what you want, but you need to rerun the hook to regenerate the profile [17:59] jdstrand: reinstalling the click package won't do that? [17:59] sergiusens: sudo click hook remove apparmor ; sudo click hook install apparmor. that will actually overwrite the file in /var though [18:00] sergiusens: not for the same version of the click package I don't think [18:00] sergiusens: can you give me the click package? [18:00] I'd like to look at it for some reviews I'm doing [18:00] "aa-clickhook -f" fixes profiles for default click apps after updating the ubuntu template for my device [18:00] jdstrand: ok, that explains a lot, thanks!, works now [18:00] ogra_, +phablet10 of pulseaudio has finished building [18:01] great, will install soon [18:01] w-flo: yes, that would work. did you file a bug fr the change you made? [18:01] jdstrand: one sec, uploading [18:02] jdstrand, are we supposed to send bug reports / merge requests for ports? The comments seem to indicate you are trying to find a better solution [18:03] w-flo: the more information the better. knowing how things are being handled on ports may influence how we fix it [18:03] jdstrand: http://people.canonical.com/~sergiusens/com.ubuntu.stock-ticker-mobile_0.3.7ubuntu1_all.click [18:03] w-flo: or at the very least, we can tell ports people what to do [18:03] jdstrand, okay :) I can send a merge request later, or would you prefer a bug report? [18:04] w-flo: bug report with debdiff is preferred (diff from the MR is fine) [18:05] alright :) [18:07] sergiusens: thanks! so, couple things: a) description is not specified in the manifest and b) shouldn't the pkgname be com.ubuntu.developer..stock-ticker-mobile? [18:08] ie, com.ubuntu.developer.surgemcgee.stock-ticker-mobile [18:08] jdstrand: I can easily add a description [18:09] jdstrand: regarding package names, since these are going to be part of ubuntu and when I talked to cjwatson we agreed on using com.ubuntu for core apps [18:09] sergiusens: it looks like 'title' would be the description and you just need a short title then [18:09] jdstrand: title is deb['description'], [18:10] I will be changing these and unlinking from deb some more [18:11] sergiusens: there is some ambiguity here that I brought up on the mailing list. title and description are both optional fileds in click, but we will want them required for the app store. DEBIAN/control's Description looks to be generated from only the click manifest 'title', but I think it might need to be a combination of both 'title' and desription [18:12] sergiusens: we should bring this up with cjwatson when he is back from holiday [18:13] sergiusens: I'll adjust me test for the pkgname [18:13] jdstrand: ok, the use of deb['description'] on these packages doesn't involve cjwatson, this was something I did end of June (IIRC) to start building these [18:13] s/me/my/ [18:14] once the format was stabilized I would do proper manifests for all [18:15] sergiusens: I might suggest this then: title = deb['package'].partition('com.ubuntu') [18:15] description = deb['description'] [18:15] ralsina, mhall119: I'm also trying to install the latest xda-developers-app, and it seems to get successfully installed: [18:16] this is shown on the logs: "Installed com.ubuntu.developer.mhall119.xda-developers-app-0.1.2" [18:16] sergiusens: that should set the title to 'stock-ticker-mobile' [18:16] but the .desktop file for the app is not created, so we can't start it [18:17] jdstrand: sure, but I'm taking this info out of the original packaging... [18:17] package_name = deb['package'].strip('-app') [18:17] manifest = {'name': 'com.ubuntu.%s' % package_name, [18:17] ralsina, mhall119: it's not even created if I install it manually: http://pastebin.ubuntu.com/6007382/ [18:17] alecu: let me phablet-flash to the latest and try again [18:18] jdstrand: anyways, I'll make you happy and be done with it ;-) [18:18] oh, no desktop file created? [18:18] alecu: so perhaps it's the hook? [18:18] mhall119: right [18:18] sergiusens: I see, then just set title to package_name [18:18] weird, there's a desktop file in the project [18:19] sergiusens: I think that is the best that can be done since the original packaging only has the package_name and description with no title [18:19] this is the installer running with debug enabled: http://pastebin.ubuntu.com/6007385/ [18:19] sergiusens: I'll make an adjustment to the maintainer check for core apps [18:19] sergiusens: thanks for you help! :) [18:19] alecu: ah, in the project it's xda-developers.desktop, but the package name is xda-developers-app [18:20] could that be the problem [18:20] ? [18:20] does the .desktop name need to match the app/package name exactly? [18:20] mhall119: I believe it is based on some other email thread [18:20] mhall119: btw, I have a number of packaging improvements for you [18:20] mhall119: I don't know about the .desktop file. The thing that has to match, from my pov is the app name in the webservice and the one in the manifest [18:20] jdstrand: I just clicked the button in QtCreator, so improvements should go to bzoltan [18:21] mhall119: I'm boostrapping the review scripts [18:21] mhall119: ah, perfect, thanks [18:21] jdstrand: but I'm happy to make any changes to see if they fix problems before asking bzoltan to make the changes to the packaging script [18:21] jdstrand: additional question, not sure if you recall what I mentioned last week about testing, but will your package check break if there's ancilliary data in the manifest? [18:21] mhall119: I am just about to eod after 16hours :) so please mail me if you need something [18:22] only 16 hours? slacker [18:22] sergiusens: can you give me an example [18:22] is jenkins down for everybody or just for me? https://jenkins.qa.ubuntu.com/ [18:22] mhall119: or feel free to send an MR https://launchpad.net/qtcreator-plugin-ubuntu [18:22] alecu: the public jenkins was down earlier IIRC [18:22] mhall119: yeah, I know.. i am a lazy bugger :) freerider [18:23] sergiusens: so, it will likely report the odd entries. [18:24] jdstrand: http://paste.ubuntu.com/6007398/ [18:24] jdstrand: it's my way of getting the autopilot tests into the phones from an installed click package [18:25] sergiusens: the script would complain about 'source'. is that a new click field? [18:25] jdstrand: to support the qa infrastructure [18:25] jdstrand: it's not a new click field [18:25] sergiusens: let me look up something... [18:26] mhall119: I just tried manually installing the weather app, (as I got it from jenkins on Jul 26), and it installed fine, including creating the .desktop in ~/.share/local/applications [18:27] I'd like to test with a newer package when jenkins gets back [18:27] sergiusens: would it make sense to change 'source' to 'x-source'? from the click docs: 'Keys beginning with the two characters "x-" are reserved for local [18:27] extensions: this file format will never define such keys to have any [18:27] particular meaning.' [18:27] alecu: you don't need jenkins http://people.canonical.com/~ubuntu-archive/click_packages/ [18:27] sergiusens: great, thanks [18:28] jdstrand: sure, I control the whole thing so changes are easy, I could even place this out of band if it were necesasry, but this seemed like an easier approach [18:28] jdstrand: making the change now, thanks [18:28] sergiusens: there seems to be many old files there, named like "*-app_*" [18:29] sergiusens: I think 'x-source' would be fine. I need to update the scripts for 'x-' anyway [18:29] alecu: just use the new ones... I'm not deleting old apps [18:29] ack [18:30] alecu: oh, I'm only copying the ones we install... that's why, if you have acces to the internal jenkins instance... I think you do, then... [18:30] alecu: http://10.97.2.10:8080/view/click/? [18:31] alecu: I'm going to rebuild all of them in like 20 mins with the recs from jdstrand [18:31] seb128: so I have a snail background image now.....how do I change that? [18:31] mhall119: latest ubuntu-clock seems to install just fine too, so I guess it must be something different in your package [18:32] alecu: the scope says it's installing, ubt then when I go out of the preview and back in the button is reset to "Install" [18:33] looks like it downloaded it... [18:35] mhall119: yes, the button being reset is a bug in my code of the apps scope [18:36] alecu: http://paste.ubuntu.com/6007437/ is my unity-scope-click.log after telling it to install xda-developers-app [18:39] sergiusens: fyi, review tools won't fail on 'x-*' now (but will report them) [18:39] mhall119: yes, I'm getting the same on my phone. And it looks that the install succeeds, but the .desktop file to start your app is never created by the click installer on ~/.local/share/applications [18:39] alecu: what are you using to install the app? [18:39] mhall119: that is also not created if I run the click installer from a console [18:39] jdstrand: pkcon [18:40] alecu: where does it install the click package to? [18:40] jdstrand: so this is what you want http://paste.ubuntu.com/6007453/, right? [18:40] mhall119: /opt/click.ubuntu.com/ [18:40] alecu: that's what I was saying earlier :) when you said those two apps are broken :) [18:40] sergiusens: that looks extremely sensible :) [18:41] davmor2: yes, but mhall119 has corrected his, and it's still not installing right [18:41] ok, I'm comitting this and rebuilding [18:41] alecu: :( that's bad, I blame the polar realignment of the sun :) [18:42] alecu: curious if it work with: sudo click install --force-missing-framework --user=$USER ./*click [18:42] jdstrand: I'll try that. I can't use "sudo" from the scope, though, so I'm forced to use pkcon [18:42] alecu, ralsina, so, my app, hello_world seems to have the same namespace in the click package and in the search index [18:42] jdstrand: and also, weird thing is that using a different .click, (the clock app) it gets installed ok. [18:43] alecu: right-- just saying, if it worked with sudo, there might be a bug with pkcon. [18:43] alecu: interesting. can you upload the click package somewhere? [18:43] jdstrand: yes but shhhh [18:43] alecu: I can run it through the review tools/see if I spot something [18:44] jdstrand: that would be great [18:44] jdstrand: the working one is: http://people.canonical.com/~ubuntu-archive/click_packages/com.ubuntu.ubuntu-clock_0.5_all.click [18:44] jdstrand: the one that does not work... let me upload it [18:45] sergiusens: did you generate ^ [18:45] jdstrand: not yet, that's what I said I was going to do :-) [18:46] sergiusens: no, I meant-- did that one come from you in the first place, not if it is fixed [18:46] hmmm, is there a click uninstall? [18:46] jdstrand: yes [18:46] jdstrand: everything com.ubuntu is from me up until now at least [18:46] jdstrand: this is mhall's app: http://ubuntuone.com/5bSZgfX145WwN74dUFg5Bj [18:47] sergiusens: apparmor/ubuntu-clock.json should not have "ubuntu-clock": {}. just put policy_groups and policy_version toplevel [18:47] mhall119: jdstrand https://code.launchpad.net/~sergiusens/stock-ticker-mobile-app/apparmor/+merge/181123 [18:47] mhall119: afaik, there's not a click uninstall yet [18:47] sergiusens: not sure if that was one of the changes you said you were going to make [18:47] jdstrand: yeah, I removed that just now and rerunning it for all apps [18:47] cool [18:48] mhall119: cjwatson may know the state of the "uninstall" feature, but he is on vacations [18:49] alecu: well, I can say for sure that it is using the old format for the manifest [18:50] jdstrand: alecu here's the latest clock http://10.97.2.10:8080/view/click/job/clock-app-click/34/console (you can see the updated json's at the end) [18:50] and here's the actual app http://10.97.2.10:8080/view/click/job/clock-app-click/34/artifact/com.ubuntu.ubuntu-clock_0.5_all.click [18:50] ah jeez, I have to connect to the vpn? :P [18:51] mhall119: ^ jdstrand says the app still has "old manifest format" [18:51] jdstrand: there's no public jenkins avail yet... :-/ [18:51] sergiusens: I'm just kidding [18:52] jdstrand: good, some people complain for real :-) [18:52] mhall119: it might be the case that qtcreator still uses it [18:52] alecu, mhall119: so the problem is that the manifest is old and does not contain 'hooks' === racarr is now known as racarr|lunch [18:52] beuno: what were the changes that you did to the manifest for the demo? [18:52] alecu: jdstrand: can you give me an example of what to look for and where in the click package? [18:52] alecu: do you want me to paste changes that would work? [18:53] I'll update my qtcreator+plugins and try again [18:53] mhall119: right, this was one of the things I was talking about when you pointed me at bzoltan :) [18:53] ok [18:53] this is the format: https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement/Manifest#Click [18:54] mhall119: you can look at http://paste.ubuntu.com/6007398/ [18:54] mhall119: let me give you a diff, hold on [18:54] alecu, replaced the absolute paths in the .desktop file [18:54] mhall119: and the MR I just posted to you ;-) [18:54] alecu, keep in mind that if the app has a _ things will blow up as well [18:54] (fixing in my app now) [18:55] jdstrand: is there a way to see logs of what the hooks are doing? [18:56] mhall119, gsettings reset org.gnome.desktop.background picture-uri [18:57] mfisch: ack on lightdm; I guess it's another job for now, but will soon be it [18:57] alecu: no-- that is up to the hook to decide atm. I know it is a todo for the apparmor hook to do more logging [18:57] seb128: do we have an ETA on when the image chooser will work in system-settings? [18:57] mhall119, ubuntu-backgrounds is not installed on the touch image so the default background is not available, and the config panel is setting the snail as a fallback when the config is not available ... the fix just got merged, should be on tomorrow's image [18:58] mhall119, when the content picker is available, kenvandine said they target a working version for system settings by end of august [18:58] ok, cool [18:58] thanks seb128 [18:58] jdstrand: is each hook a separate lp project? I'd like to open bugs to have some logging from the apparmor hook and for the hook that creates the .desktop files [18:58] mhall119, yw! [18:59] plars, Did you guys figure it out? [18:59] In what piece of the stack does the proximity sensor in Touch get activated? Like, say I'm on a call and hold up the phone to my ear. [19:00] alecu, ralsina, I've re-uploaded my app, take a crack at installing it now [19:01] * alecu triesx [19:01] -x [19:01] also, testing out unicode support now ;) [19:02] Hi , anyone working on sony xperia z ? [19:02] beuno: I get an empty square after "Hello world", so it seem that the char is not supported in that font on the phone! [19:02] alecu: each hook is separate, yes. there are 3 right now. the lcick-desktop in click itself. that is going away. then there is upstart-app-launch which provides a hook to replace click-desktop. there is also click-apparmor [19:03] alecu, http://www.charbase.com/32db-unicode-circled-katakana-si [19:04] beuno: the .desktop file is not created for your app either, just like for mhall119's app above [19:04] beuno: it seems that the click manifest format changed [19:05] ah [19:06] so that's what the sdk is generating [19:06] beuno: yes, it seems to be broken [19:06] beuno: the apps that sergiusens is packaging in jenkins seem to be working fine, though [19:06] sergiusens tends to cheat a lot [19:07] he doesn't get caught, though [19:07] beuno: classic argentinian [19:07] * sergiusens doesn't use the sdk [19:08] beuno: it would be great if sergiusens' apps show up in the click index in some automagical way [19:08] sergiusens: that's what they meant by cheating [19:08] mhall119: well the sdk can't be used from jenkins ;-) [19:08] alecu: the core apps click packages? [19:08] <_5m0k3> Can somebody tell me how many gridunits a N7 is? height and width [19:08] mhall119: right: those are working if I install them from the cmdline [19:09] mhall119: they don't show up on the click index though [19:09] _5m0k3: 14 [19:09] <_5m0k3> 57 x 73? [19:09] alecu, we can create the official canonical account and start uploading [19:09] _5m0k3: /etc/ubuntu-touch-session.d/grouper.conf is where it's set [19:10] oh, how many gridunits wide? [19:11] mhall119: http://paste.ubuntu.com/6007556/ [19:11] beuno: uploading from the canonical account would be awesome, to have some apps on the prod index that work [19:11] mhall119: note, you don't have to put the security manifest in apparmor/ -- just make sure that the click manifest use the right file [19:12] mhall119: also note, that is a diff between unpacked click package trees, not your bzr branch [19:12] ok [19:12] jdstrand: I may just wait until the SDK is fixed [19:13] yeah, we need the SDK to work properly to open the contest [19:13] unless you're unsure if this will fix it, in which case I'll make the change to test it [19:13] that should work [19:14] alecu: do you want an updated package with my changes to test? ^ [19:14] jdstrand: yes, please [19:16] alecu: http://people.canonical.com/~jamie/click/com.ubuntu.developer.mhall119.xda-developers-app_0.1.2+jdstrand1.click [19:17] ogra_: i think the jenkins for dashboard is back up [19:17] \o/ [19:17] alecu: note, that package uses the same internal version [19:17] now we have to figure if we will get the results for the old images still... [19:17] plars: doanac`: ^^ [19:18] ? [19:18] alecu: so you might want to uninstall the other one before trying that one [19:18] jdstrand: how do I unpack this dang .click file? [19:18] mhall119, rename .tar :) [19:19] boiko, om26er: heyo -- do either of you know what bits of code trigger the screen to go dark when the phone is raised to one's ear? (i know the proximity sensor lives in platform-api, but I don't know what drives it) [19:19] mhall119: dpkg-deb -R /some/dir [19:19] doanac`: plars I've added a test app to https://code.launchpad.net/~sergiusens/phablet-tools/click_test_provision/+merge/179835 [19:19] mterry: ChickenCutlass probably knows it better [19:19] mterry: powerd [19:20] boiko, mterry what do I know? [19:20] :) [19:20] thanks jdstrand [19:20] np [19:20] beuno: tar command didn't like the file [19:20] jdstrand's command works through [19:20] ChickenCutlass: the proximity sensor, where is the code that triggers the screen to go black [19:20] its an ar file [19:20] ChickenCutlass, do you know what bits of code trigger the screen to go dark when the phone is raised to one's ear? (i know the proximity sensor lives in platform-api, but I don't know what drives it) [19:20] boiko, powerd [19:20] mterry: ^ [19:20] jdstrand: and the command to re-pack it? [19:20] one could use ar x [19:20] mterry, lp:powerd [19:20] sergiusens, ChickenCutlass: ok, thanks. I thought I searched in that [19:20] will look again [19:21] mterry: you can also turn that sensor on yourself with powerd-cli [19:21] mterry: let me find the line [19:21] oh, did not look in that [19:21] that would be a good example of a call to follow-in [19:21] dpkg-deb -b --nocheck /some/dir /path/to/$pkgname_$version_all.click [19:21] So does the phone-app specifically turn the feature on? Or is it always on? [19:21] mhall119: ^ [19:21] mterry: its always on for now [19:22] mterry: powerd listens on dbus for an ofono signal and enables it then [19:22] thanks jdstrand [19:22] mfisch, ah... [19:23] sergiusens: can you review this MR to enable ubuntu plugin for maliit: https://code.launchpad.net/~bfiller/maliit/enable-ubuntu-conf/+merge/181139 [19:23] mterry: on_ofono_voicecall_signal(), I left a huge comment explaining it there, powerd.cpp [19:23] bfiller: let me test it [19:24] mfisch, thanks, looking around [19:24] sergiusens: ubuntu-keyboard packages has landed in saucy, so after landing this change we can update the seed to pull ubuntu-keyboard instead of maliit-keyboard [19:24] mterry: the sensor enablement is done via a display state request, so thats lower down the stack [19:24] mterry: let me know if you have more questions [19:24] bfiller: I think we can pull both and nothing bad would happen? Let me try that too [19:25] sergiusens: true [19:25] will work [19:26] bfiller: ok, so let's do that first for a seamless migration [19:26] mterry: that code should be moved elsewhere at some point, if that's your question ;) [19:26] sergiusens: I'll do an MR for the seed then [19:26] mterry: this isn't related to the lock screen appearing during phone calls, is it? If so I think you're barking up the wrong tree. [19:26] mfisch, I'm trying to figure out why a SysPowerStateChange(0) signal is emitted during a proximity event, which makes the greeter appear [19:27] sforshee, oh really? [19:27] mterry: oh yeah there's a bug on that one [19:27] bfiller, here's the code that you sent me, but inserted into my applifecycle app ... http://bazaar.launchpad.net/~rick-rickspencer3/+junk/app-lifecycle/view/head:/ContactsTab.qml [19:27] well, and slightly tweaked with some comments [19:28] mfisch, so you're saying I could use powerd-cli to fake an event? [19:28] mterry: powerd will turn off the screen and possibly suspend the device, but it can't directly cause anything to be changed in the ui [19:28] mterry: something might be responding to one of the signals powerd emits, either about suspending or about turning the display on or off [19:28] mterry: you can cause an event to happen [19:28] sforshee, sure, but the shell listens to powerd signals, and shows the greeter on a SysPowerStateChange(0) event. Which doesn't seem like it should be emitted in this case... [19:29] thats odd [19:29] mterry: we might suspend, so if we do that signal is appropriate [19:29] mfisch, I just want to confirm that my guess about the signal is right (it's only thing that would show greeter, but still, just want to confirm) [19:29] mterry: Sys state or display? [19:29] mfisch, so I'd want to replicate the proximity event stuff [19:29] mfisch, we don't care about display state, only sys state [19:30] sforshee, we'd suspend in the middle of a call? [19:30] mterry: you can use 'powerd-cli listen' to see exactly what signals powerd is emitting [19:30] mterry: yes [19:30] sforshee, yeah but I don't know how to replicate the situation without actually being in a call [19:30] rickspencer3: thanks [19:30] sforshee, was hoping for an easy way to reproduce the bug [19:30] mterry: you can affect states via the cli, but there's no direct call that forces a signal [19:30] mterry: run it w/o args [19:30] mterry: it also has a signal listening mode which will help you [19:30] sforshee, maybe I don't understand what suspend means on Touch [19:31] mterry: you can make a display request that enables the sensor, let me get the semantics [19:31] mterry: telephony is generally handled by a separate CPU from the main application processor, so it's possible to suspend the CPU doing all the ubuntu stuff and keep the phone call active [19:31] mterry: you can control the proximity sensor with powerd-cli [19:31] mterry: powerd-cli display dc proximity and then in another terminal, powerd-cli listen [19:31] 'powerd-cli dispaly proximity' [19:32] yeah, what mfisch said. I forgot the dc [19:32] sfeole: did we drop on/dc? ok.. [19:32] mterry: we can G+ if you'd like [19:33] sforshee, mfisch: I believe the DisplayStateChanged signal has a flags argument? Which indicates if a proximity event is the reason for the change. Maybe the SysPower event needs that too then [19:33] I dont think proximity sensor is forcing the phone off [19:33] that doesnt make sense [19:34] mfisch, sforshee says it might [19:34] sforshee: when proximity turns the screen off though... it could? [19:34] mterry: when the screen goes off then there could be nothing left which prevents the phone from suspending [19:34] screen on does block suspend [19:34] so when that's removed powerd could decide that it's okay to suspend [19:35] and I think that's 100% legitimate [19:35] right === chihchun_afk is now known as chihchun [19:35] mterry has a problem in that he cannot distinguish between a power button press and screen off during a call [19:38] asac: the dashboard should sync up with the results missed over the past 2 days I think [19:38] sergiusens: thanks for the test app! [19:39] doanac`: np, wrt to tedg's comment, I'd say this is mostly blocked on mir and it might be good to test this with the mir image [19:39] mterry: why not have a policy that you never push the lock screen over the phone app? [19:40] Sorry, got disconnected [19:40] sforshee, so if I understand correctly... proximity turns screen off; then the system might look around and say "nothing stopping us from suspending" and suspend. So it's not *directly* because of proximity but rather suspends only indirectly due to proximity event === chihchun is now known as chihchun_afk [19:41] mterry: correct [19:41] sergiusens: do we need to be able to support non-mir usage, or will getting this sorted out for mir be sufficient for now? [19:41] mterry: when the screen is off, we stop the activity timer and so nobody is left [19:42] mfisch, sforshee: OK. So sounds like the shell needs to keep track of whether the display is stopped for proximity reasons to know if we should respond to suspend signals... [19:42] doanac`: well mir was supposed to be switched to this week... I say block on that [19:42] well last week [19:42] mfisch, sforshee: do you know if there's a way to tell if the system is auto-suspending vs a user pressing the button while in proximity [19:42] but you get the point [19:43] sergiusens: yep. thanks [19:43] i'll let thomi know [19:43] mterry: not via powerd I dont think [19:43] doanac`: that's however doesn't prevent us from working on the autopilot feature and since thomi is in the mir team, he can grab the right people too [19:43] Those should be treated differently... [19:43] correct [19:44] doanac`: I'm aiming for preemptive here since all rocks are going to be falling on QA once they land ;-) [19:44] mfisch, how malleable is the powerd signal interface? Do we know if many people are using it? Is it standardized or something we can change? [19:44] It seems to be under a com.canonical interface...; [19:45] mterry: some of the details here will change when powerd and mir/unity get fully integrated together [19:45] mterry: thats a sforshee question, I'm off doing other fun stuff now, but the users are limited at this point [19:45] unity may be consuming it though [19:45] maybe this is one of the things that can get communicated via that interface === racarr|lunch is now known as racarr [19:46] sforshee, it seems reasonable to add a flags signal to SysPowerStateChange. DisplayPowerStateChange has one [19:46] sforshee, and then we could say "proximity is on / this is an auto-suspend vs user-suspend" [19:49] mterry: I'm not sure that it would be possible to do that reliably [19:49] mterry: better to get the data from the display signals I think [19:50] sforshee, but we still wouldn't be able to tell whether a suspend that happens during proximity is due to auto-suspend (which we should not show greeter for) or due to user press (which we should show greeter for) [19:51] When I android-chroot it cannot find sh and its because /proc/pid/root/system is empty.. when I copied my system partition files over to that folder, then I found root@android [19:51] sforshee, does powerd do the auto-suspending? [19:52] anyone have and idea as to why /proc/pid/root/system is empty? [19:52] mterry: so the plan in the future is that unity will decide to turn off the screen on power button press, not powerd (though it will go through powerd) [19:52] mterry: powerd does do the auto suspending, but all that code knows is whether or not anything is requesting a state > than suspend [19:53] not why it is being requested [19:53] those are logically separate in powerd [19:56] sforshee, but when the system suspends due to idle, we'd still want to show the greeter. So we want to show greeter in all suspend cases but auto-suspend-while-proximate [19:56] sforshee, we actually used to listen to power button press, but changed it to listen to powerd because we weren't showing greeter during idle-suspend [19:57] Gah, internet [19:57] sforshee, but when the system suspends due to idle, we'd still want to show the greeter. So we want to show greeter in all suspend cases but auto-suspend-while-proximate [19:57] sforshee, we actually used to listen to power button press, but changed it to listen to powerd because we weren't showing greeter during idle-suspend [19:57] mterry_: the plan as far as I know is that unity _will_ handle the power button and idle timeout itself, and just communicate that information to powerd [19:57] mterry_: it just hasn't all been implemented yet [19:58] mterry_, do you want to allow suspend during a active call? or just screen dim? [19:58] sforshee, hrm. OK [19:58] pmcgowan, not that I want it, but apparently something we do. apparently we can suspend during an active call because of different CPUs for the call stuff than normal stuff [19:58] pmcgowan: if we suspend during a call it means less power used, so longer talk time [19:59] mterry_, sforshee makes sense [20:00] sforshee, OK, sounds like maybe my best bet is to way then for unity8 to subsume this job... [20:01] mterry_: that sounds likely. Some of the stuff powerd does right now are just stop-gap measures. [20:05] sforshee, if the proximate sensor turns off the display, but we don't suspend for some reason (maybe a process is active), and the user presses the power button while remaining proximate, what happens? Turn on the display or go to suspend? [20:07] sergiusens: here is MR for seed to add ubuntu-keyboard https://code.launchpad.net/~bfiller/ubuntu-seeds/touch-add-ubuntu-keyboard/+merge/181148 [20:08] bfiller: no need [20:08] bfiller: look at the comment I just added to the keyboard MR [20:08] ok [20:08] mterry_: the power button will override display state requests, but not power state requests. So if some other process in the system is requesting the active state, powerd will not suspend. [20:09] bfiller: it's right above == Shell == [20:09] sforshee, OK, so if it doesnt auto-suspend, it won't button-suspend either? [20:09] sergiusens: ah, didn't see that [20:10] bfiller: the ubuntu keyboard rotates horribly on manta [20:10] mterry_: correct [20:10] sergiusens: worse than maliit-keyboard? [20:11] bfiller: yes... it is positioned locked and rotates keys but trims [20:11] mterry_: and the proximity sensor does fundamentally the same thing as the power button -- forces the screen off but doesn't override system state requests [20:11] bfiller: let me get a screen [20:11] mterry_: there's a system state request associated with having the screen on, and as soon as the screen is turned off that request is released === iBeliever is now known as iBelieve [20:13] sforshee, OK, then we could probably handle this today by just watching signals to see if we're in proximate mode and ignore all suspends during. I don't have to worry about a manual button suspend happening. We'll either auto-suspend or not at all [20:14] sforshee, ah... so all suspends are auto-suspends in that sense. Nobody is directly suspending, everyone just shuts the screen off? [20:15] sergiusens: actually the rotation is busted on maguro too, keyboard rotates in the shell and it shouldn't [20:15] bfiller: http://people.canonical.com/~sergiusens/ubuntu-keyboard/ [20:15] sergiusens: like if you click search and start typing in dash search field and then rotate the keyboard rotates and shouldn't [20:15] mterry_: more or less ... powerd only suspends as a result of all active state requests being released, but the display code isn't the only thing that can take active state requests [20:16] bfiller: yup, but look at landscape-2.png in that link [20:16] it starts doing some weird rotation [20:16] sergiusens: all messed up [20:16] sigh [20:16] well at least we caught it [20:17] guess we shouldn't make it the default yet [20:17] agreed on that [20:17] sergiusens: guessing it was his latest rotation "fix" that caused this issue. I might revert that and then we can try again [20:18] <_5m0k3> New screenshots of blackjack app: https://plus.google.com/110587021591787005452/posts/if6UrgTAJgc [20:18] bfiller: ack, on manta in portrait there's that small hover effect too [20:19] sergiusens: what do you mean? [20:19] <_5m0k3> If somebody could test it on a N7 and send me a screenshot of landscape and portrait, I'd appreciate it [20:19] sforshee, gotcha, but display state is what both proximity sensor and power button are hooked up to, it sounds [20:20] mterry_: yes [20:29] sforshee, btw, thanks for walking me through all that! :) [20:30] bfiller: http://people.canonical.com/~sergiusens/ubuntu-keyboard/portrait.png [20:30] bfiller: see how the keyboard is not really at the bottom? [20:32] sergiusens: I see [20:32] sergiusens: I'm filing bugs [20:33] sergiusens: assuming rotating works correctly if you switch back to maliit-keyboard in server.conf? [20:33] one sec [20:33] it was from recollection but let me check again [20:35] bfiller: it's not a regression bug [20:36] bfiller: floating thing in portrait happens on both [20:36] sergiusens: what about other issues? [20:37] bfiller: the rotation whack is just ubuntu-keyboard [20:37] k [20:37] bfiller: I can log them if you want [20:38] sergiusens: already did, if you want to add comments https://bugs.launchpad.net/ubuntu/+source/ubuntu-keyboard/+bug/1214578 [20:38] Launchpad bug 1214578 in ubuntu-keyboard (Ubuntu) "rotation broken on Manta" [Critical,New] [20:39] and https://bugs.launchpad.net/ubuntu/+source/ubuntu-keyboard/+bug/1214575 [20:39] Launchpad bug 1214575 in ubuntu-keyboard (Ubuntu) "keyboard rotating when shell is not" [Critical,New] [20:46] doanac`: updated MR [20:49] sergiusens: ah - didn't think about setup.py needing an update also :) === jo-erlend_ is now known as jo-erlend === freeflying_away is now known as freeflying [21:54] sergiusens: reverted latest orientation changes on ubuntu-keyboard, when jenks builds we can test https://code.launchpad.net/~bfiller/ubuntu-keyboard/revert-orientation-fix/+merge/181163 [21:55] bfiller: ack === kentb is now known as kentb-out [22:06] sergiusens: any idea where renato push his code for QOrganizer EDS plugin for calendar? not sure if you were involved or not [22:07] bfiller: I know how to get to it [22:07] bfiller: lp:qtorganizer5-eds [22:07] sergiusens: nice [22:07] salem_: ^^^ [22:08] bfiller, ah cool. didn't know this project existed. [22:10] jdstrand: did you email bzoltan the changes needed to the click build script in the SDK? [22:10] sergiusens, salem_ : looks like it landed in saucy too yesterday [22:11] mhall119: I am working on filing bugs, etc now [22:11] mhall119: I will email after that [22:14] thanks jdstrand, I'll follow up with him in the morning [22:15] Hi [22:15] Is Ubuntu Touch based upon android? I read somewhere that it uses SurfaceFlinger instead of Mir for driver compatibility [22:15] alfonsojon: currently it does, yes [22:16] Alright [22:16] Mir will be replacing SurfaceFlinger soon though [22:16] Will this break compatibility with most Android devices though? [22:16] currently some android bits like that run in an LXC container on top of Ubuntu Touch [22:16] I'm a novice Cyanogenmod developer so I was curious [22:16] alfonsojon: no, Mir will talk to the same hardware (using the same drivers, IIRC) as SurfaceFlinger === freeflying is now known as freeflying_away [22:17] But Mir is not compatible with X11 or Wayland drivers [22:17] What makes Surfaceflinger special? [22:17] mhall119: well, latest update and look what we have here... http://popey.com/~alan/device-2013-08-20-231706.png [22:17] we need a better way to show app names, especially for long apps [22:17] and hello world seems to have a unicode character or something in iti [22:17] popey: I saw === freeflying_away is now known as freeflying [22:18] XDA De...ers [22:18] lol [22:18] and yeah, I was thinking we should talk to design about allowing 2 lines of text [22:18] alfonsojon: I think wayland and Mir are both capable of using SurfaceFlinger drivers [22:18] Huh... [22:18] That's honestly surprising [22:19] alfonsojon: something about EGL that's way over my head [22:19] Does Utouch support mice? [22:20] hmmm, that I don't know, I haven't tried [22:20] * popey gets his usb cable [22:20] I mean, you can run Unity 8 on the desktop and use mouse and keyboard there [22:20] I was wondering because I have a Manta-based laptop [22:20] so in theory is does [22:20] It's essentially a Nexus 10 laptop [22:21] The ARM Chromebook has exactly the same SoC that the Nexus 10 uses [22:21] my nexus 4 doesn't power up the mouse on the usb port [22:21] Just with a trackpad, keyboard, and lower res display [22:22] popey: speaking of the new Click scope, you can technically use it to install apps right now (but click packages build from QtCreator are broken) [22:23] popey: I'm hoping that tomorrow I'll get the XDA app installable [22:23] One more question [22:23] Can UTouch be compiled with U-Boot support? [22:23] mhall119: I was thinking og CCing a mailing list so app developers would be aware of the issues. which should I use? [22:23] alfonsojon: U-Boot? [22:23] Yeah [22:23] It's an EFI-based firmware & bootloader [22:23] jdstrand: I wouldn't worry about it yet, since the click store hasn't been announced yet [22:24] alfonsojon: if Ubuntu supports it, then Ubuntu Touch should support it [22:24] fginther: plars: doanac`: all the clicks will have x-source in tomorrows build, you can give it a go and we can happrove it after that [22:24] sergiusens: are there touch images with Mir enabled, or is it something I have to set up by hand? [22:24] doanac`: hand [22:25] sergiusens: you have directions anywhere? [22:25] This is going to be fun [22:25] :) [22:25] ARM Chromebook + Utouch [22:26] = yay? [22:26] mhall119: root@ubuntu-phablet:/opt/click.ubuntu.com# ls [22:26] com.ubuntu.developer.mhall119.xda-developers-app [22:26] alfonsojon: why not run Ubuntu desktop on the chromebook? [22:26] mhall119: it installs ☻ [22:26] popey: yeah, but the .desktop file isn't put into ~/.local/share/applications/ [22:26] and it's not pointing to the right place either, so you can't just copy it [22:26] ah [22:27] so close ☻ [22:27] was nice to see that there though, nice work! [22:27] mhall119: Because it uses X11 [22:27] popey: yeah, that's what jdstrand is going to send to bzoltan, so hopefully will be fixed soon [22:27] i updated minesweeper btw [22:27] alfonsojon: for now :) [22:27] Mir doesn't like it either [22:27] The drivers just flat out suck [22:28] doanac`: https://wiki.ubuntu.com/Touch/Testing/Mir#Switch_from_SurfaceFlinger_to_Mir [22:28] sergiusens: thanks! [22:28] So, I figured if it's identical to a Nexus 10 on the insides, then Surfaceflinger should run fine [22:28] http://popey.com/~alan/device-2013-08-20-232806.png i suck at minesweeper [22:28] lol [22:28] popey: have you tried moving launcher icons yet? [22:28] nope [22:28] I'm getting my hands on an official Cyanogenmod device tomorrow [22:29] = shouldn't be hard to port Utouch / could have already been ported [22:29] popey: long-press, then pull them off to the side [22:29] oooh [22:29] I'm so excited :) [22:29] alfonsojon: what device? [22:29] !devices | alfonsojon [22:29] alfonsojon: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [22:29] popey: can you get this in? https://code.launchpad.net/~sergiusens/stock-ticker-mobile-app/apparmor/+merge/181123 [22:29] * popey clicks [22:30] approved [22:30] VS920 [22:31] oh my gosh [22:31] it's supported unofficially :) === salem_ is now known as _salem [22:34] plars: if you are still there, how about we put the current threshold at 98.0 to get started [22:34] ? [22:34] asac: down from 99/99.25? [22:34] plars: or 98.5 ... i think thats currently achieved in all runs seen so far [22:34] plars: yeah [22:34] plars: i believe that values above 98 are ok for autopilots [22:35] even lower most likely [22:35] but still important to check whats going on [22:35] asac: agree, if we have a runaway process, 95% even should catch it [22:35] i kind of dont like that lots of stuff is happening during test runs [22:35] yeah. but dont want to get too low [22:35] asac: agree [22:35] i looked at all and those that had succesfful autopioots [22:35] asac: I'll set it to 98 [22:35] always made 98 [22:35] i still like those logs :) [22:36] asac: good [22:36] plars: i am a bit scared with 98 though [22:37] plars: http://reports.qa.ubuntu.com/smokeng/saucy/image/3661/ubuntu-weather-app-autopilot/287489/ [22:37] that was was pretty noise, but made 98 [22:37] plars: http://reports.qa.ubuntu.com/smokeng/saucy/image/3661/ubuntu-weather-app-autopilot/ [22:37] well, but we know that those failures are network failures [22:37] plars: how is the ip filter ticket going? :) [22:38] popey: thanks [22:38] asac: tried it out with the guest network last week, and eventually got it to work, but that router is junk. IS is going to add an additional one without the restrictions, hopefully this week from the last I heard [22:38] asac: we'll use that [22:38] its really intersting how busy the system is [22:38] i dont see it like that here at all [22:38] asac: which device are you testing with locally? [22:38] plars: maguro [22:40] plars: will you land this change tonight? [22:40] would be cool to see it tomorrow greenish :) [22:41] asac: yes [22:41] another ueventd runaway: http://reports.qa.ubuntu.com/smokeng/saucy/image/3657/calendar-app-autopilot/287507/ [22:41] http://reports.qa.ubuntu.com/smokeng/saucy/image/3657/calendar-app-autopilot/ [22:42] plars: that one seems to be safely below 98 :) [22:42] asac: well, we knew we would see a lot of that [22:42] its good [22:42] want to see just that for now :) [22:42] e.g. looping processes [22:42] we can carea bout other noise in a separate test i guess [22:43] plars: so maybe 97.5 is safer for the before/after ones? [22:43] i guess so [22:43] lets do that [22:43] really just want to catch the big fish [22:43] asac: and for the one in default smoke tests? [22:44] plars: well, i thought at some point we might want to run it like 4 times: systemsettle-80 systemsettle-95 systemsettle-98 systemsettle-99 [22:44] maybe strip the 99 for now [22:44] i assume the 80 will in almost all cases succeed on first run [22:45] same for 95 [22:46] asac: hmm, why so many? If we expect it to settle to 99% idle, then we should just set it at 99% idle. If we add all these others for 80, 90, 95, then we are just adding extra results for things that would either pass and we don't care (because it's not idle enough) or would fail, but also fail at 99% [22:48] plars: yeah. well. lets use the same value i guess [22:49] plars: lets be nice and go 97.5 on both fronts and be happy [22:50] asac: ack [22:50] keep noise low, but catch big problems [22:50] revisit if we still get too many false alerts [22:51] plars: oh the .1 build is building [22:51] plars: i sasume we cant sneak the change in that even :)? [22:51] * asac guesses not [22:53] asac: done [22:54] wow :) [22:54] nic [22:54] e [22:54] asac: well, no [22:55] ah :) [22:55] k [22:55] asac: looks like the 20.1 build came out a while ago [22:55] ok i just saw first job on dashboard [22:55] so thought it had still way to go :) [22:55] asac: and things don't look good on it [22:55] :( [22:56] i better reboot now... have problems [22:56] bbiab [22:56] hmm [22:56] phablet-flash failed to download [22:59] plars: retry :) [22:59] asac: yeah, I am [22:59] asac: looks like the image said it was there, but it wasn't [23:00] plars: is the image now there? [23:01] asac: seems to be, yes [23:01] asac: they are already running again [23:01] plars: so i guess it means our tweak might get picked up? [23:01] hmm [23:01] * asac hopes [23:02] asac: yes [23:02] cool [23:04] * plars goes back to the 12.04.3 marathon [23:11] anybody around that knows how lxc works? [23:13] especifically who calls the /var/lib/lxc/android/rootfs/init.rc and in what order [23:20] w-flo, i think i foudn my issue [23:20] i never renamed the 70-vision.rules to 70-epicmtd.rules [23:26] annerajb, oooh. [23:26] yes. [23:26] that may explain why it only wrote a minimal amount of lxc output before getting stuck [23:26] you shouldn't just rename it, it's better to create your own file. I guess it will boot with my file, but some hardware will fail [23:27] fail = not work until you fix the file [23:34] i am on the process of flashing cyanogen mod to run your script [23:34] i renamed it on the rootfsmodifier since i dont need your device lol [23:37] hm.. my script? :) [23:37] I haven't flashed cynanogenmod in a long time, so I doubt it's my script [23:43] it's posted on the wiki under your notes w-flo [23:43] adb shell cat /var/lib/lxc/android/rootfs/ueventd*.rc|grep ^/dev|sed -e 's/^\/dev\///'|awk '{printf "ACTION==\"add\", KERNEL==\"%s\", OWNER=\"%s\", GROUP=\"%s\", MODE=\"%s\"\n",$1,$3,$4,$2}' [23:43] which is wrong because it's using the path for the rootfs instead of the cyanogen mod path [23:43] ah, ogra_ added that I guess [23:44] also i think it has a type with the new line [23:44] do you have adb in the rootfs? [23:45] yeah but i dont get a shell because i believe the permissions are not there [23:45] ie. i think adb never starts and it fails early [23:45] which makes sense since lxc cannot set permisions for adb [23:46] that shell command doesn't work for me [23:46] adb shell cat /var/lib/lxc/android/rootfs/ueventd*.rc|grep ^/dev|sed -e 's/^\/dev\///'|awk '{printf "ACTION==\"add\", KERNEL==\"%s\", OWNER=\"%s\", GROUP=\"%s\", MODE=\"%s\"\n",$1,$3,$4,$2}' | sed -e 's/\r//' this is better [23:50] yeah i had to delete the ^M on vi [23:50] ima edit the wiki to use yours [23:50] even thought i think the path is still wrong [23:50] I've done that already [23:50] :D [23:50] :D [23:51] On the vision, ubuntu booted fine with an empty 70-vision.rules file [23:51] lucky you [23:52] I then hacked together a custom python script to do the conversion :) seems like ogra has a simpler solution [23:52] i am downloading the latest tested rootfs so i can repack everything with the new rules file [23:52] yup mighty bash [23:55] can someone post a dang /etc/fstab for me to look at [23:55] would REALLY help me [23:55] please! [23:55] I am curious what vendor says [23:56] lxc-android-boot is adding /system and /data but nothing for /vendor [23:56] and then upstart/lxc-android-boot.log is complaining it cant find vendor [23:56] in my canse and i think w-flo we commented # /data [23:56] ? [23:57] so vendor partition in fstab is point to your /dev/device_data_partition? [23:57] our devices dont have partition labels so it cannot moutn /data [23:57] or a bind mount? [23:57] no vender should point to /system/vendor [23:57] since the vendor stuff is on the system partition [23:57] ahhh [23:57] exactly [23:57] so a fstab entry would be a bind? [23:58] "/system/vendor /vendor bind defaults,bind 0 0" [23:58] ? [23:58] Curious what you guys actually have [23:59] for some reason the fstab detector.. automatic.. things.. detects my data and vendor.. but nothing for vendor.. and then the log says its missing in mtab or fstab [23:59] so I am curious what you guys have [23:59] so I can add it myself, and remove that update-fstab