[00:05] <Nothing_Much> Hello, is it possible to use libhybris on the Ubuntu desktop?
[00:12] <slangasek> Nothing_Much: what would you use it for on an Ubuntu desktop?
[00:12] <slangasek> libhybris is a wrapper around hardware-specific android libraries; you probably don't have any of those on your desktop
[00:17] <Nothing_Much> slangasek: For an arm device that doesn't have drivers yet on it.
[00:17] <Nothing_Much> *for Linux
[00:17] <slangasek> so you are running Ubuntu desktop on an ARM device?
[00:18] <Nothing_Much> Yes I am
[00:18] <slangasek> ok
[00:18] <slangasek> so /in theory/, you could use this to provide Ubuntu desktop through XMir on top of Mir, routing through libhybris
[00:18] <slangasek> but I think you'll be the first to actually do this if you succeed :)
[00:19] <Nothing_Much> Ah well, I was going to try Wayland, though now that you mention it, I could try both, at different times.
[00:20] <Nothing_Much> Since this is Ubuntu though, how would I go about using XMir?
[00:21] <slangasek> apt-get install unity-system-compositor, I believe
[00:21] <Nothing_Much> that's it?
[00:21] <Nothing_Much> oh uh oh
[00:21] <Nothing_Much> I forgot to mention I'm using uh.. Xubuntu
[00:21] <slangasek> XMir won't care ;)
[00:21] <Nothing_Much> Oh really? Cool
[00:21] <slangasek> https://wiki.ubuntu.com/Mir/Installing
[00:24] <Nothing_Much> I'd also like to see if Wayland would work, no offense.
[00:24] <slangasek> none taken ;)
[00:24] <slangasek> I believe you'll find the wayland story is a bit more do-it-yourself, though
[00:24] <Nothing_Much> oh dear
[00:25] <Nothing_Much> yeah that's a problem when you're dealing with a dumb consumer like myself
[00:27] <Nothing_Much> Okay so I got a warning and Xmir won't launch
[00:28] <Nothing_Much> [ 11314.194] (WW) "xmir" is not to be loaded by default. Skippin
[00:28] <Nothing_Much> g
[00:29] <slangasek> hmm, I'm not sure where that's configured
[00:29] <slangasek> do you already have the libhybris stuff installed?
[00:29] <Nothing_Much> Yes
[00:29] <slangasek> ok
[00:30] <Nothing_Much> Do I need libubuntu-application-api1?
[00:31] <slangasek> not 100% sure, but I think you only need libhardware2
[00:31] <Nothing_Much> I got that installed already
[00:31] <Nothing_Much> lemme try restarting lightdm again
[00:32] <Nothing_Much> still software rasterizer :(
[00:32] <slangasek> yes
[00:33] <slangasek> you need to configure the X server somehow so that it knows to use xmir on your hardware
[00:34] <Nothing_Much> would that require an xorg.conf? :(
[00:34] <slangasek> probably :)
[00:34] <Nothing_Much> hmm..
[00:35] <Nothing_Much> well, so far no driver exists on Linux for an Exynos 5 atm, that should be where Libhybris takes its place
[00:41] <Nothing_Much> What would I put for the "driver" section of the conf?
[00:42] <slangasek> I don't think you need to specify a driver, only to specify that the mir module should be loaded
[00:42] <Nothing_Much> Oh
[00:42] <slangasek> but you may not even need to do that; it's possible all the configuration lives in lightdm
[00:42] <Nothing_Much> How would I do that?
[00:42] <Nothing_Much> Well I did sudo lightdm restart twice
[00:42] <Nothing_Much> Maybe a full restart will do it
[00:42] <Nothing_Much> brb
[00:44] <Nothing_Much> still nothin'
[00:44] <slangasek> Nothing_Much: yes, you need additional config somewhere, I just don't know where :)
[00:44] <Nothing_Much> Oh darn
[00:44] <slangasek> out of the box, Mir+Xmir will only run on known supported drivers
[00:45] <slangasek> is lightdm displaying for you?
[00:45] <Nothing_Much> Yeah, lightdm is here but not under xmir
[00:46] <slangasek> so if you've restarted lightdm, it *should* be displaying on top of mir
[00:46] <slangasek> because of /etc/lightdm/lightdm.conf.d/10-unity-system-compositor.conf
[00:46] <Nothing_Much> well it's not, it says this: [    10.061] (WW) "xmir" is not to be loaded by default. Skipping.
[00:46] <slangasek> that's XMir
[00:46] <slangasek> that's not related to whether Mir is running
[00:47] <slangasek> lightdm talks to Mir directly, it doesn't rely on XMir
[00:47] <slangasek> do you have an X server running when at the lightdm greeter?
[00:47] <Nothing_Much> You mean.. a gui?
[00:48] <slangasek> I mean, is there an 'X' process running
[00:49] <Nothing_Much> I'm not sure?
[00:49] <slangasek> ps waxuf | grep X
[00:49] <slangasek> if there is, Mir is failing to start
[00:49] <Nothing_Much> Yeah there is
[00:49] <Nothing_Much> X that is
[00:49] <Nothing_Much> Not Mir
[00:51] <slangasek> what happens if you run 'unity-system-compositor' from the commandline?  (best not to try this from an existing X session, you may want to stop lightdm first)
[00:51] <Nothing_Much> hmm..
[00:52] <Nothing_Much> I'll risk it
[00:52] <slangasek> well, even if it doesn't crash your X server, it may not tell you anything useful :)
[00:52] <Nothing_Much> ERROR: Throw location unknown (consider using BOOST_THROW_EXCEPTION)
[00:52] <Nothing_Much> Dynamic exception type: boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >
[00:52] <Nothing_Much> std::exception::what: assign: Bad file descriptor
[00:52] <Nothing_Much> Just got that
[00:52] <slangasek> you need to run it as root
[00:52] <slangasek> and you really want to run it from console
[00:53] <slangasek> well - console, or remotely
[00:55] <Nothing_Much> Ah
[02:25] <nothing_much> So I apparently flashed this for no reason.. Oh well, does anybody know an xorg.conf for XMir?
[04:24] <nothing_much> Anybody know how to get libhybris to work on an Ubuntu desktop with xmir?
[04:27] <cwayne> nothing_much: what are you trying to do?
[04:28] <nothing_much> cwayne: I'm trying to run (X)Ubuntu desktop on an arm device.
[04:28] <nothing_much> It's not a tablet btw
[04:28] <nothing_much> Just a very tiny arm PC :)
[04:28] <cwayne> something like a panda board?
[04:28] <nothing_much> Yeah
[04:28] <cwayne> ah
[04:29] <nothing_much> Except it uses an Exynos 5
[04:29] <cwayne> sorry, I don't know anything to help you, I was just curious :)
[04:29] <nothing_much> Ah darn
[04:29] <nothing_much> Anybody else? I'm using an Odroid-XU
[04:29] <nothing_much> Trying to get libhybris to work
[04:29] <cwayne> nothing_much:  more people are active in this channel in EU timezones
[04:30] <nothing_much> oh really?
[04:30] <nothing_much> well luckily I'm nocturnal
[04:30] <nothing_much> in the US
[04:31] <cwayne> ha, me too
[04:32] <cwayne> to be honest nothing_much, i'm not sure anyone's really tried anything with libhybris re: desktop
[04:32] <nothing_much> you sure about that?
[04:32] <nothing_much> there should be a way to
[04:32] <nothing_much> since it's basically utilizing the android driver.. things
[04:32] <nothing_much> right?
[04:35] <cwayne> i'm not saying it's not possible
[04:35] <cwayne> i'm just saying to my knowledge nobody's tried yet
[04:35] <cwayne> (people may very well have, just not to my knowledge)
[04:35] <nothing_much> yeah
[05:12] <Mirv> mhall119: ok, I can do a branch
[09:48] <JamesTait> Good morning all; happy Friday and happy X-Ray Day! :-D
[09:48] <Nothing_Much> JamesTait: woo!
[10:00] <oSoMoN> dpm, hey, regarding your question on robru’s askubuntu answer about the term size, it’s probably because the image is RO by default, try "touch /userdata/.writable_image" and then reboot
[10:06] <dpm> oSoMoN, oh, but that will make my image writable and I'll have to reflash to get back to RO, right? That makes that answer not really useful
[10:06] <dpm> actually, I should probably say it is useful, but only applies to RW images
[10:07] <oSoMoN> dpm, yeah
[10:07] <oSoMoN> dpm, I would answer on the askubuntu page directly, but apparently my reputation is too low and I’m not allowed to answer to inline questions
[10:09] <lool> seb128: Hey, when I open system settings > date & time with a touch nexus 4 I've just updated to latest devel-proposed, I get an empty list of settings, and the title doesn't say "Date & time" but "System settings"; is this known?
[10:09] <Laney> lool: It's https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1248646
[10:09] <seb128> lool, yes, https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1248646
[10:10] <lool> Ok thanks
[10:10] <seb128> lool, it has not been a good week for the uitk, ken also told me that their toolbar change is creating issues as well
[10:10] <lool> ack
[10:11] <t1mp> 19:18:00 < robru> t1mp, ping. any ETA on those ui-toolkit patches landing?
[10:11] <t1mp> robru: which patches? We have a bunch of MRs that are ready to be merged, but waiting for AP1.4 to land before we approve stuff
[10:17] <t1mp> seb128: there was no UITK release since 11 Oct until now, so a very large amount of changes that were made in the past month only came to the apps now
[10:18] <seb128> t1mp, yeah, that's "suboptimal"
[10:18] <t1mp> seb128: a bunch of good ones, but also the bugs that popped up :(
[10:19] <t1mp> did AP1.4 land? i.e., can we land fixes in the UITK now?
[10:20] <popey> sil2100: ^^
[10:23] <sil2100> t1mp: AP 1.4 landed, but probably Mirv will know best regarding landings in UITK
[10:23] <sil2100> Mirv: ?
[10:23] <t1mp> sil2100: ok, thanks. I'll wait for Mirv's "go"
[10:32] <Mirv> t1mp: yes, so as discussed I know of two regression fixes https://code.launchpad.net/~nicolas-doffay/ubuntu-ui-toolkit/selector-api-break-fix/+merge/194313 + https://bugs.launchpad.net/ubuntu-rssreader-app/+bug/1248759 - those can go in as soon as possible
[10:33] <Mirv> (the latter if applicable)
[10:35] <Mirv> t1mp: then after we are sure there are no regressions anymore compared to the 20131016 ui-toolkit release, and nothing that is required for fixing app AP tests, and we have an image that has the new ui-toolkit release, the trunk can be "really" opened
[10:38] <t1mp> Mirv: happroving https://code.launchpad.net/~nicolas-doffay/ubuntu-ui-toolkit/selector-api-break-fix/+merge/194313
[10:44] <Mirv> happroval accepted
[10:56] <seb128> MacSlow, hey, could you review https://code.launchpad.net/~larsu/notify-osd/update-sync/+merge/194364 ?
[11:00] <ogra_> seb128, do we have any migration process for user settings on upgrades in place yet ? (i.e. if at some point LC_ALL gets set, will the user setting get updated so he gets the fix)
[11:00] <ogra_> if not, we should definitely work out something ...
[11:00] <seb128> ogra_, what do you mean with LC_?
[11:01] <ogra_> seb128, thats just an example
[11:01] <ogra_> i know that you set the locales in ~/,pam-environment
[11:01] <ogra_> we will need a way to update the user settings if there are system fixes coming in
[11:02] <ogra_> so users that upgrade get the fix too in their setups
[11:02] <seb128> ogra_, man session-migration
[11:02] <ogra_> ah, k
[11:02] <ogra_> (no man on the phone :P )
[11:03] <seb128> ogra_, http://blog.didrocks.fr/post/Announcing-session-migration-now-in-ubuntu
[11:03] <ogra_> yeah, i remember it
[11:03] <seb128> ogra_, http://manpages.ubuntu.com/manpages/raring/man1/session-migration.1.html
[11:03] <ogra_> i wasnt aware we use it on the phone
[11:04] <seb128> ogra_, we don't yet afaik, but no reason we couldn't if we had a need for it
[11:04] <ogra_> right, we should
[11:04] <seb128> ogra_, there is no config migration to do yet afaik though
[11:04] <ogra_> there will surely be :)
[11:04] <seb128> ogra_, I'm not convinced, but let's see
[11:05] <ogra_> i just want to make sure that we dont forget about it before the first stable to stable update happens ...
[11:06] <ogra_> seb128, well, how will i get 24h clock settings if you dotn migrate my ~/.pam-environment once that bug gets fixed ?
[11:06] <ogra_> (for example)
[11:08] <davmor2> Morning all
[11:09] <seb128> ogra_, to be fair I don't care, we have enough issues, it's a v1, you can go to settings and pick a locale to fix your clock
[11:09] <ogra_> how would i know about this ?
[11:10] <ogra_> it wont be the last settign we have to migrate and it will likely also affect v2 or v3 once we have changes in settings
[11:11] <ogra_> which gets particulary intresting with asac approach that you shoould be able to switch channels back and forth being up or downgraded at your aill
[11:11] <ogra_> *will
[11:11] <ogra_> (since there will be settings that arent backwards compatible or wheer the format of the file in the homedir changed etc)
[11:13] <MacSlow> seb128, on it
[11:16] <asac> maybe we should make settings - similar to APIs - something that we dont treat as an internal thing that we can just change and refactor as we feel, but rather something that should be managed, discussed and once agreed, frozen forever with an SETTINGS scheme version etc. :)
[11:16]  * asac doubts its doable for the past, but maybe something to think about future way we handle setting schemes
[11:17] <ogra_> asac, well, we have the boot hooks to handle such stuff, and i suppose running the session -migration automatically against the current version of whatever setting you have for each up-downgrade should work
[11:17] <yarre> I want a good mail/calendar/contacts application with support for getting it from a server (as in not local) for ubuntu.. why arent there any?
[11:18] <ogra_> yarre, because you didnt write one yet !
[11:18] <ogra_> :)
[11:18] <asac> ogra_: i think its not so much about migrating away... its keeping backward compatibility...
[11:18] <asac> consider we use a setting called background which might be just a string now
[11:18] <asac> in future we decide we need to make a structured setting out of that (e.g. a tupel)
[11:18] <ogra_> asac, well, its keeping compatibility ... in either direction
[11:18] <asac> so shouldnt we continue to keep the other setting working?
[11:19] <ogra_> at least if you want to allow the either-direction-channel-switch approach
[11:19] <asac> right compatibilty ... i think settings should be come a part of our API that we version and not change without keeping in mind that those that use the old setting scheme still need to be able to continue to do so
[11:19] <asac> anyway. i haven't thought about this problem enough to have any sane input :)
[11:20] <ogra_> right, me neither, thats why i asked seb128 :)
[11:20] <asac> seb128: what are we using to store the settings? dconf?
[11:21] <ogra_> dconf or dot files ... depending on the app
[11:21] <ogra_> s/app/setting/
[11:21] <asac> seb128: are apps supposed to access settings at all?
[11:21] <asac> or are they confined to just have access to its own settings?
[11:21] <ogra_> only through the API
[11:21] <asac> ogra_: what kind of API? a generic key value  look up? or rather strictly typed etc. functions?
[11:22] <ogra_> well, except language and locale settings for example ... they just are session wide set
[11:22] <asac> e.g. get ("background") or getBackground()
[11:22] <asac> ?
[11:22] <ogra_> something like that, yeah
[11:22] <asac> where is the settings API?
[11:22] <ogra_> iirc you can currentlly reqest info about stuff like aut-rotation defaults and such
[11:22] <asac> i guess that should naturally get embraced by platform API
[11:22] <seb128> asac, ogra_: apps setting is a topic we didn't tackle yet
[11:22] <ogra_> somewhere in the QMl stuff ... Ubuntu.Components or so
[11:23] <ogra_> but i would think app settings are a matter of the app devs
[11:23] <seb128> asac, ogra_: system settings are a mix of gsettings and file, e.g /etc/timezone, we basically use whatever was in place
[11:23] <ogra_> not our problem
[11:23] <seb128> ogra_, asac: app settings are likely going to be qsettings with a qpa using e.g u1db
[11:23] <asac> seb128: right. thats the backend, but how do apps/unity etc. interact with that? do we have an API?
[11:23] <ogra_> what i care about are the system defaults that live in ~/
[11:23] <asac> or do they go directly to /etc/timzeone etc. as needed?
[11:24] <davmor2> ogra_: this has happened a couple of times to me now on maguro it looks like it is suspended ie black screen, you press the power button to wake it and nothing you have to pull the battery to get it to power up again
[11:24] <ogra_> davmor2, that only happens to be if it drained the battery completely
[11:25] <seb128> asac, we have a mix, gsettings-qt is the API to access gsettings, for /etc files we either use custom backends to talk to dbus services (e.g timedated) or direct file editing from cpp
[11:25]  * ogra_ has seen that as well ... but i usually need to charge it then before i can actually boot
[11:25] <davmor2> ogra_: no this is on 40% when it happened this morning
[11:25] <davmor2> so last night must of been on 80-ish%
[11:25] <ogra_> davmor2, well, file a bug, attach syslog and stuff :)
[11:26] <davmor2> ogra_: I was hoping it would happen again so I could see if I could adb into it in the broken state
[11:26] <ogra_> ah, well, then do that
[11:27] <ogra_> mguro is slowly moving to lower prio though ...
[11:29] <asac> seb128: gsettings-qt is basically a key look up? or is that something more meaningful?
[11:29] <asac> e.g. get("background) rather than getBackgroundInfo
[11:29] <asac> seb128: smells like we should hide all of that behind the platform API and make a decent API there for all our system settings
[11:30] <asac> i will connect you to ricmm
[11:30] <seb128> asac, it's a key lookup yes, e.g
[11:30] <seb128>     GSettings {
[11:30] <seb128>         id: desk
[11:30] <seb128>         schema.id: "org.gnome.desktop.sound"
[11:30] <seb128>     }
[11:30] <seb128> print(desk.eventSounds)
[11:30] <asac> right. think right thing is to really hide all that stuff behind a decent API that we can manage, discuss and support forever :) ... let's see if we can experiment with that as part of platform API v2 and discuss/see where we would hit walls etc.
[11:31] <asac> but just an idea to put up there for now :)
[11:31] <seb128> asac, wfm; though the number of components accessing system settings is limited (it's basically the settings app)
[11:31] <seb128> asac, so I'm not sure it makes sense to have an API for it
[11:31] <ogra_> well, you will still need to migrate whatever lives in the homedir
[11:31] <seb128> migrate from what to what?
[11:32] <asac> seb128: aren't all components != app a potential client for those settings? e.g. unity, mir, etc.?
[11:32] <ogra_> seb128, from image 100 saucy to image trusty 10 and backwards
[11:32] <seb128> asac, for some yes, but those are shared are mostly stored in gsettings or accountsservice
[11:32] <ogra_> seb128, asac wants to be able to move the release back and forth underneath the stable and devel aliases ...
[11:32] <seb128> ogra_, well, what do you want to migrate when switching between those images?
[11:33] <ogra_> ad as well allow users to do that randomly as they like
[11:33] <yarre> ogra_, Im not old enough.. somebody should have done it already >_<
[11:34] <asac> note: going back is not a big priority for now :) ... but its an interesting test for many things :)
[11:34] <ogra_> seb128, well, if image 1 has keys x, y and z ... and image 10 renames then to xa, xb and xc ... once you roll back you need to migrate them back to the old names
[11:34] <seb128> ogra_, having compat in our storage backends is not going to be easy, image an app using a sqlite db and changing the table structure in a new version
[11:34] <seb128> ogra_, that's a difficult topic, good luck tackling it
[11:34] <ogra_> seb128, i dont care about apps
[11:34] <asac> its not about apps so myuch. yeah. those would just get disabled
[11:34] <ogra_> thats (as i said above) a matter of the app devs to keep compatibility
[11:35] <asac> also an app has to decide if they support downgrading etc.
[11:35] <seb128> asac, so you would e.g loose your addressbook contacts?
[11:35] <ogra_> its the system settings ... imagine we rename "background" to "wallpaper"
[11:35] <seb128> or webbrowser bookmarks?
[11:35] <ogra_> this name needs to be tied to an image version then
[11:35] <ogra_> and the upgrade mechanism needs to know about it
[11:36] <ogra_> and change it accordingly
[11:36] <seb128> ogra_, system settings is such a ridiculous small part of that issue
[11:36] <ogra_> app devs need to define their own settings api
[11:36] <seb128> ogra_, where it gets tricky is not system settings, it's e-d-s and contacts, or webbrowser and bookmarks
[11:36] <ogra_> right, system apps fall under system settings for me
[11:37] <seb128> they don't for me
[11:37] <seb128> they are thing I've no clue about and I'm not interested in resolving
[11:37] <ogra_> well, we need to resolve it for the whole of the system
[11:37] <seb128> supporting format changes in both direction is not an easy problem
[11:37] <ogra_> but not for any apps that dont come preinstalled
[11:37] <seb128> ogra_, is any OS out there doing that?
[11:38] <ogra_> seb128, i dont know any OS that supports going back and forth at your will
[11:38] <seb128> yeah, because it's not an easy problem
[11:38] <ogra_> i know
[11:38] <ogra_> :)
[11:38] <ogra_> i think it would be worth a vUDS discussion
[11:38] <seb128> well, not my call, but it seems a lot of efforts ... not sure that's our first issue to tackle
[11:39] <ogra_> it is something that will influence our future ... and it will be hard to fix once we have to much stuff established, better do it right from the start so we dont have to hack around issues later
[11:39] <seb128> the issue is that we don't start from scratch
[11:40] <seb128> we have lot of components coming from out there
[11:40] <ogra_> i.e. via defining an API version for settings or somw such ... as asac suggested (not sure thats a good idea, but i have no better one)
we should never have used eds :)<unreasoonablerantmode>
[11:40] <seb128> which don't support that and don't plan ot
[11:40] <seb128> asac, sure, we can forget about opensource and just start an OS from 0, not reusing anything existing ;-)
[11:40] <ogra_> seb128, it doesnt have to live in the components ... rather in an upper layer
[11:40] <asac> seb128: I didnt say that :)
[11:41] <ogra_> in the abstraction above the actual apps
[11:41] <seb128> asac, well, that's sort of what we are doing though, we want to change/rewrite almost everything than exists
[11:41] <asac> right
[11:41] <asac> i think there is a pattern
[11:41] <seb128> ogra_, create the abstraction and then apps eventually will see the benefit and start using it? :p
[11:42] <ogra_> seb128, right, but we need to define the abstraction layer ... it doesnt exist yet ;)
[11:42] <seb128> ogra_, ok, wfm, let me know when it exists so I can look at using it ;-)
[11:42] <ogra_> seb128, i dont see that as a "your team alone" issue, it spans across all teams ...
[11:42] <seb128> I don't even see how you could abstract things so different in a same API
[11:43] <asac> right. lets ignore this topic for now :)
[11:43] <ogra_> if you add a system setting with your app, you add the setting and possible values to a db ...
[11:43] <ogra_> if the setting naming changes the db needs to reflect this
[11:43] <seb128> is /etc/timezone a db?
[11:43] <asac> i will have our smart archtects think about this a bit and see what they come up with :)
[11:44] <ogra_> and the session migration uses the db to migrate to the matching setting of a specific version
[11:44] <seb128> ogra_, the issue is that not every transformation is reversible
[11:44] <ogra_> most will be though
[11:44] <seb128> ogra_, you might have stuff doing sql db updates on upgrade that you are not able to reverse later if you want to downgrade
[11:44] <seb128> e.g e-d-s for contacts storage
[11:44] <ogra_> there might be settings in v2 that weren in v1 ... these wouldnt be touched on going backwards
[11:45] <seb128> so you would loose your addressbook on downgrade
[11:45] <ogra_> then you need to keep a backup of the old db and move the data
[11:45] <ogra_> and the API needs to know this (as well as the migration tool)
[11:45] <asac> addressbook synched to the cloud might be an answer
[11:45] <asac> :)
[11:45] <asac> same for bookmarks
[11:46] <ogra_> or that ... but that forces you to be online
[11:46] <seb128> ogra_, if you do that you loose any edition done while running the new version
[11:46] <ogra_> why ?
[11:46] <seb128> because the new version is not going to edit the old db you move
[11:46] <ogra_> i read the data from the new db and push it into the old structure
[11:46] <seb128> lol
[11:46] <seb128> good luck doing that
[11:47] <ogra_> the tool needs to know both structures
[11:47] <seb128> I think you underestimate how much work is in there
[11:47] <ogra_> and know that they are incompatible and how to solve this
[11:47] <seb128> no app dev is going to want to support retro compat for their past formats
[11:47] <ogra_> i dont think i do ...
[11:47] <ogra_> i know it is a big thing
[11:48] <seb128> is that a so compelling feature that it's worth the investment?
[11:48] <ogra_> app devs (for click package apps from the store) wont have to do that
[11:48] <ogra_> their settings are bound to the app, not to the system
[11:48] <ogra_> when they get settings from the system that already happens through the ubuntu API
[11:49] <ogra_> seb128, well, not that compelling for going backwards (i dont like that idea anyway) but it is surely very important that we update the user settings when going forward
[11:51] <ogra_> s/user settings/user-dir stored system settings/
[11:51] <seb128> ogra_, we spent some time thinking about upgrades issues in the past and came with some solution, but it's not an easy topic ... I don't even want to think about handling downgrades and both way transitions
[11:51] <seb128> ogra_, right, upgrade is not something new
[11:51] <seb128> ogra_, and we already support those
[11:51] <ogra_> seb128, right, asac brought up that going back thing yesterday ...
[11:52] <seb128> downgrade is another topic
[11:52] <ogra_> which made me think about tieing settings to the image version somehow
[11:52] <seb128> that doesn't help you much...
[11:53] <ogra_> the above db and tool would ... but it would require a lot of developer discipline to keep it up to date
[11:53] <seb128> you need to know how to transform a configuration described in a new format to one that the old app can read
[11:53] <seb128> and that's just simply not always possible
[11:53] <ogra_> yes
[11:54] <seb128> sometime the changes made create a situation were you don't have enough infos to create something the old app version would understand
[11:54] <ogra_> ti is always possible to migrate data ... as long as you know both formats
[11:54] <seb128> if the new format is rich enough
[11:54] <seb128> let's say you dropped a field
[11:54] <seb128> and are adding new entries
[11:55] <seb128> what would happen with the old app that use that field?
[11:55] <seb128> would you just "invent" values to populate the config on downgrade
[11:55] <ogra_> liek i said, you might have to create a new db in the old format and feed the content of the new db into it ... dropping all data for unknown keys
[11:55] <seb128> how buggy is that going to look in the app?
[11:55] <seb128> well, what if it's the other way around
[11:55] <seb128> if the old config has more infos
[11:55] <seb128> like timestamps of when the records are created
[11:55] <seb128> and the new one doesn't
[11:55] <ogra_> you have the new db around already
[11:55] <seb128> and you don't have the info to retro fill it
[11:56] <ogra_> and upgrade the content that changed in the old one since you migrated
[11:56] <seb128> well, the new db might not have the infos you need
[11:56] <seb128> cf my timestamp example
[11:56] <ogra_> the new db already has it
[11:56] <ogra_> when i went forward it got all the imafo
[11:56] <seb128> not for new record you added since the update
[11:57] <seb128> since the new version stopped collecting those info, because it got simplified
[11:57] <ogra_> if i go back and make changes in the old db and then go forward again the changes just need to be fed into the new db again
[11:57] <ogra_> if it got simplified that simplification will happen again going forward
[11:57] <seb128> yeah, the issue is going back to the old one when you did change on the new one
[11:57] <ogra_> as you said, forward we already have a migration
[11:58] <seb128> you might be missing info to retrofit in the old config
[11:58] <seb128> because you just stopped to need those in the new world
[11:58] <seb128> so you simply don't have them
[11:58] <seb128> and you have no way to guess them
[11:58] <seb128> (like a timestamp of an event that happened and didn't register)
[12:08] <Mirv> oSoMoN: is https://code.launchpad.net/~osomon/gallery-app/hide-toolbar/+merge/194094 still worked on / going in, or is it abandoned in favor of UI Toolkit side https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/ap-toolbar-open/+merge/194122 ?
[12:09] <Mirv> oSoMoN: ahum, I see gallery-app AP:s themselves are passing already without it..
[12:12] <oSoMoN> Mirv, it’s on hold until we have a resolution on the UITK side indeed
[12:15] <yarre> ogra_, KDE Kontact does all of the things i asked ;)
[12:15] <ogra_> yarre, so you just need to port it to QML then :)
[12:17] <yarre> ogra_, nope no porting... just run and be happy ;)
[12:17] <ogra_> wont work
[12:17] <yarre> i just need it on my regular desktop
[12:18] <ogra_> oh, i thought you complained about mail on the phone
[12:18] <yarre> no just linux in general :)
[12:19] <ogra_> well, this is the ubuntu-touch channel, nobody expects general linux mail questions here :)
[12:19] <ogra_> (#ubuntu is the channel for general ubuntu questions ;) )
[12:24] <Mirv> oSoMoN: alright, thanks
[13:23] <Mirv> seb128: do you want to test the fixed ui-toolkit version, settings was affected?
[13:24] <seb128> Mirv, I sure can, and yes, setting was affected, some of the panels wouldn't load, easy to test ;-)
[13:25] <Mirv> seb128: yeah I noticed, too. http://pastebin.ubuntu.com/6382000/
[13:25] <Mirv> and error alredy :)
[13:25] <seb128> Mirv, thanks
[13:25] <Mirv> daily-build PPA, that is
[14:11] <seb128> MacSlow, thanks for review the notify-osd changes, how is that looking? ;-)
[14:11] <seb128> MacSlow, oh, you just commented while I was writing that it seems
[14:12] <MacSlow> seb128, just commenting on the bug itself with my solution/alternative (while waiting for design-input) so it still can move forward regardless the outcome.
[14:16] <Mirv> seb128: I went through all of the settings panels, seem to work fine
[14:16] <karni> What's the sdk team channel? (besides ubuntu-sdk which has like 2 folks in it ;) )
[14:16] <seb128> Mirv, great, thanks, I'm about to test as well if you still need my ack
[14:17] <karni> nvm got it
[14:17] <Mirv> seb128: if you happen to test I'm happy to get an ack. I'm running various AP:s still.
[14:23] <MacSlow> seb128, just posted my alternate solutions as a further comment
[14:23] <seb128> MacSlow, thanks
[14:24] <seb128> Mirv, it's way better but there is still a bug there
[14:24] <seb128> Mirv, the ringtone/messaging sound subscreens are empty with that version, they work if you downgrade to the saucy toolkit version
[14:37] <Mirv> t1mp: ^
[14:37] <Mirv> dandrader: ^
[14:38] <Mirv> seb128: ok, thanks for testing. I haven't found any regressions so far, so I'll probably be releasing it (or kenvandine / robru will if I won't) as is, but a new bug would be needed for the remaining problem
[14:39] <seb128> Mirv, do you have ringtones listed in the corresponding panel?
[14:39] <seb128> Mirv, that's not a regression compared to the buggy trusty version but it's still once compared to a week ago
[14:40] <Mirv> seb128: I'll check once my current AP is finished
[14:40] <Mirv> seb128: I'm now comparing just regressions to #15 image
[14:40] <seb128> Mirv, ok, seems an improvement over that one indeed
[14:47] <mterry> tedg, you mentioned in Oakland wanting to be able to use the greeter DBus API in unity8 sooner rather than later.   But you'd only be using it in phone_greeter mode, right?  Which we don't ask for yet
[14:48] <tedg> mterry, Well, that depends what the greeter is on the desktop.
[14:48] <tedg> mterry, And in the phone greeter case I'd prefer to just ask "what's the current user?" and then never get a change.
[14:48] <mterry> tedg, you mean unity8-greeter or unity-greeter?
[14:48] <tedg> mterry, So if it changes form "phablet" to "phone" we don't have to care.
[14:49] <mterry> tedg, sure, but in phone greeter case, you won't use that indicator code until I split the greeter right?
[14:49] <ogra_> why would the greeter differe pn phone/tablet or desktop ?
[14:49] <ogra_> *differ on
[14:49] <tedg> ogra_, Multi user by default.
[14:50] <ogra_> tedg, right, but why would the greeter differ ? :)
[14:50] <tedg> mterry, I'm confused.  We'd still have the same code path.  We'd just be asking you for which user to ask account service for.
[14:51] <tedg> mterry, We wouldn't be using the switching, but we still need to know the user name.
[14:51] <ogra_> tedg, i dont think we should have different greeters, but instead one that can detect if there is more than one user and show the right stuff
[14:51] <mterry> ogra_, it's not the greeter, it's the indicators
[14:51] <mterry> ogra_, the greeter does detect that
[14:51] <tedg> ogra_, That's the plan, but we're not there yet.  And I think mterry is trying to prioritize.
[14:51] <ogra_> tedg, if i have a fully converged phone one day i might wat to have multiple users on this
[14:51] <mterry> ogra_, we support that!  :)
[14:51] <mterry> ogra_, oh you mean on the indicator sid
[14:52] <ogra_> you guys discussed having different greeters above
[14:52] <ogra_> based on multi/single user
[14:52] <tedg> No, it's the unity7 legacy greeter vs. the unity8 new greeter.
[14:52] <mterry> tedg, I'm just saying, in Oakland, you said you'd like me to port the desktop DBus API to unity8 so the indicators could at least start using that before I split the greeter out.
[14:52] <ogra_> tedg, aaaah !
[14:53] <mterry> tedg, but "phone" mode (c.f. phone_greeter) shouldn't be using the DBus API, right?
[14:53] <tedg> mterry, I guess all I really want is that you implement enough of the API that you can return "phablet" as the current user.
[14:53] <mterry> tedg, and all we ever ask is for "phone" mode right now
[14:53] <tedg> mterry, The indicators display different UIs, but they dont' have different modes.
[14:54] <tedg> (plus or minus, but mostly)
[14:54] <mterry> tedg, so the indicator will always look for a greeter and ask for the current user, even in a user session, where it will just fail to find the DBus name?
[14:54] <tedg> mterry, So, yes, it'll have a "lightdm" mode.  But not a form factor mode.
[14:55] <szymon_w> hi! What was a command to perform test on .click package before I submit it to USC ?
[14:55] <Mirv> seb128: no the ringtones/messaging submenus do not show the list
[14:56] <mterry> tedg, OK.  But "lightdm" mode isn't being used in phone yet, until I split, right?  I'm just not seeing why landing DBus stuff now in unity8 (ahead of split) would be useful to you
[14:56] <Mirv> otherwise fine
[14:56] <seb128> Mirv, save here
[14:56] <seb128> same
[14:56] <jdstrand> chrisccoulson: hey, on bug #1249326, is that something you have started or something others could look at?
[14:57] <tedg> mterry, Ah, yes.  No, until it is split no.  But it should be pretty trivial, no?
[14:57] <chrisccoulson> jdstrand, i've not started looking at that yet. it should just be a case of adding another build step
[14:57] <mterry> tedg, the DBus API?  Sure.  I started a branch with support, but realized halfway through that you wouldn't even be using the code yet
[14:59] <seb128> chrisccoulson, hey
[14:59] <davmor2> szymon_w: you are better asking that on #ubuntu-app-devel
[14:59] <seb128> chrisccoulson, did you see my ping earlier? ;-)
[14:59] <chrisccoulson> seb128, yeah
[15:00] <jdstrand> chrisccoulson: ok, well, if it seems enablement-y, I was thinking rsalveti might be able to provide some insight (fyi only)
[15:00] <seb128> Mirv, I've a testcase, http://paste.ubuntu.com/6382461/
[15:00] <chrisccoulson> seb128, it's not the sort of thing i'd normally do an out-of-band update for, particularly as it's not a new bug, and IIUC it's not likely to be something that affects most people
[15:01] <seb128> chrisccoulson, ok, that seems fair enough to me, I'm just going to unsubscribe sponsors then
[15:01] <seb128> chrisccoulson, is that going to be in 26?
[15:01] <chrisccoulson> seb128, i also disagree with the assertion that there's no regression potential - this code has no tests, and has been quite fragile in the past
[15:01] <seb128> chrisccoulson, that got commited but I'm not sure how to see what serie
[15:02] <chrisccoulson> seb128, i'm not sure which branch it's in yet
[15:02] <szymon_w> davmor2, thanks
[15:03] <seb128> chrisccoulson, ok, no worry, I'm just going to unscribe sponsors with a comment saying it's coming with one of the next security updates when $whatever_version_inlcuding_the_fix lands
[15:03] <seb128> chrisccoulson, thanks
[15:06] <Mirv> seb128: please file a bug for sdk team and ping bzoltan / t1mp with it
[15:08] <mterry> Cimi, did you get anywhere with your wizard-cmake branch?
[15:13] <davmor2> ogra_: we having any new images today dude?
[15:13] <ogra_> davmor2, leater, yeah
[15:13] <davmor2> ogra_: nice thanks
[15:30] <Cimi> mterry, it works
[15:30] <Cimi> mterry, working on wifi
[15:34] <jdstrand> chrisccoulson, rsalveti: fyi, I created bug #1249387 and assigned it to rsalveti based on the meeting this week. it is a bit sparse, feel free to adjust
[15:35] <mterry> Cimi, seb128 mentioned to me in Oakland that he'd be interested in us landing an incomplete version in system-settings trunk soonish, then we can propose branches to fix it up
[15:35] <mterry> Cimi, are you working out of the wizard-cmake or welcome-wizard branch?
[15:35] <chrisccoulson> and we haz mailing list now:
[15:35] <chrisccoulson> https://lists.launchpad.net/oxide/
[15:35] <chrisccoulson> ;)
[15:35] <jdstrand> \o/
[15:36] <Cimi> mterry, wizard-cmake
[15:37] <popey> chrisccoulson: how do you join?
[15:38] <chrisccoulson> popey, https://launchpad.net/~oxide
[15:38] <popey> not https://lists.launchpad.net/oxide-developers/ ?
[15:38] <popey> confusing ☻
[15:38] <ogra_> isnt oxide just another word for rust ?
[15:38]  * ogra_ wonders if the name choice was so clever :P
[15:39] <mterry> Cimi, OK, is it in a relatively clean state?  Could we propose for merging?
[15:40] <chrisccoulson> popey, yeah, this is an unfortunate side effect of launchpad requiring you to be a team member. oxide-developers have full commit access
[15:40] <Cimi> mterry, if we don't add the wifi, yes
[15:40] <chrisccoulson> popey, so that team has to be restricted
[15:41] <mterry> Cimi, but wifi isn't enabled now, right?
[15:41] <Cimi> mterry, don't remember, was playing with it
[15:41] <davmor2> ogra_: only if iron is in front of it
[15:42] <Cimi> nope that I remember
[15:43] <mterry> Cimi, alright.  I'll look at taking the current branch and propose it.  And maybe drop the extra po file?  I still feel like we don't need a second one
[15:44]  * mterry looks into it
[15:44] <Cimi> mterry, yeah
[15:46] <cwayne> popey: had a chance to try out that app by any chance?
[15:47] <popey> cwayne: no because i dont want to make my phone r/w
[15:47] <popey> can i do it any other way?
[15:48] <oSoMoN> Mirv, hey, can you confirm that we can resume autolanding for apps that are back to green in the dashboard?
[15:48] <cwayne> popey: not yet :/
[15:48] <cwayne> popey: we need either the account plugin to land in the image, or the ability to install it as a click
[16:27] <Cimi> seb128, I'm trying to run dpkg-buildpackage on the phone for system-settings, and I get cp: cannot stat 'debian/tmp/usr/share/upstart': No such file or directory
[16:27] <seb128> Cimi, what branch do you try to build?
[16:28] <Cimi> seb128, my wizard-cmake
[16:28] <seb128> Cimi, grep for upstart in it?
[16:28] <seb128> do you have a .install listing that?
[16:28] <Cimi> seb128, debian/ubuntu-system-settings.install:usr/share/upstart
[16:28] <seb128> that's why
[16:28] <seb128> drop that
[16:29] <Cimi> ok
[16:42] <sergiusens> fginther, can you run this agains your maguro? http://paste.ubuntu.com/6382941/
[16:44] <fginther> sergiusens, yes
[16:45] <sergiusens> fginther, let's start a hangout in about 2 hours to debug, further, sound good?
[16:47] <fginther> sergiusens, sure, I may be babysitting and have to move it a little later
[16:48] <sergiusens> fginther, later is better actually ;-)
[16:48] <fginther> sergiusens, I'll ping you when I'm available
[16:48] <sergiusens> sounds good
[16:55] <sil2100> oSoMoN: ping :)
[16:55] <oSoMoN> sil2100, pong
[16:57] <sil2100> oSoMoN: regarding the gallery-app failures, we're waiting for a fix from the UITK side, right? Did you hear anything about the status of that?
[16:57] <sil2100> Hope it won't be changing the API and require many AP test modification?
[16:58] <oSoMoN> sil2100, nope, not heard back from it, but I think the proposal was not modifying the API at all
[16:58] <sil2100> t1mp: are you the one working on the toolbar ?
[16:58] <t1mp> sil2100: yes, mostly me
[16:58] <oSoMoN> sil2100, basically, the implementation of click_button() would ensure that the toolbar is up before actually clicking the button
[16:59] <sil2100> THat would be awesome
[16:59] <sil2100> t1mp: how far is it from being complete?
[16:59] <sil2100> t1mp: the autopilot management of the toolbar ^
[16:59] <oSoMoN> sil2100, while you’re around, can you confirm that we can resume autolanding for apps that are back to green in the dashboard?
[17:03] <t1mp> sil2100: can you be more specific?
[17:03] <t1mp> sil2100: there are autopilot emulators
[17:04] <t1mp> sil2100: if you are referring to this bug specifically: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/ap-toolbar-open/+merge/194122
[17:04] <t1mp> sil2100: I meant this bug: https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1248487
[17:04] <t1mp> sil2100: for that bug, this MR is ready: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/ap-toolbar-open/+merge/194122
[17:04] <t1mp> sil2100: waiting for jenkins/CI to approve, and then we can merge it.
[17:08] <t1mp> sil2100: jenkins just rejected it :(
[17:08] <t1mp> sil2100: https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/3261/console
[17:08] <t1mp> to me it seems there is some write error AFTER the tests are executed. Or am I overlooking something?
[17:11] <oSoMoN> sil2100, ping, re-resuming autolanding
[17:23] <cwayne> mhall119: is there an api to integrate with the messaging indicator?
[17:23] <cwayne> qml bindings, that is
[17:35] <cwayne> tedg: hey, i noticed when you click a link in google that's a tel:// link, it doesn't open the dialer
[17:35] <cwayne> is that because it's not tel:/// ?
[17:36] <tedg> cwayne, perhaps
[17:36] <tedg> cwayne, Not sure which links the webbrowser forwards on.
[17:36] <cwayne> tedg: so this is likely a webbrowser-app bug?
[17:37] <tedg> cwayne, Uhm, I'd start there.  They'll pass it on if it's not :-)
[17:37] <cwayne> tedg: sounds good to me :)
[17:52] <mandel_> barry, ping
[18:01] <sergiusens> fginther, are you still looping?
[18:05] <fginther> sergiusens, yes
[18:06] <mandel_> kenvandine, ping
[18:06] <fginther> sergiusens, 26 iterations
[18:07] <sergiusens> fginther, the you change the sleep to 60 (or did I forget); and is this the same system?
[18:11] <fginther> sergiusens, it just failed on iteration 31
[18:11] <fginther> this is the same system
[18:12] <sergiusens> fginther, great, so what's the output of
[18:12] <sergiusens> fginther, adb devices
[18:12] <sergiusens> fginther, can you adb shell after?
[18:12] <fginther> sergiusens, it shows up on adb, I can shell in
[18:13] <sergiusens> fginther, ah, it's the dumb failure of mtp reconfiguring the bus and disconnecting adb midway then
[18:13] <fginther> sergiusens, http://paste.ubuntu.com/6383448/
[18:13] <fginther> sergiusens, did you want a 60 second sleep, the script is only using 5
[18:14] <sergiusens> fginther, yeah, just because you have a 60 in there
[18:15] <fginther> sergiusens, restarted
[18:16] <kenvandine> mandel_, pong
[18:17] <mandel_> kenvandine, I was wondering if you could point me to a qmake.pro that install header for development, I've starting to improve u-d-m to have a client lib
[18:17] <mandel_> kenvandine, and I need to create a -dev package
[18:17] <barry> mandel_: pong
[18:18] <mandel_> barry, wanted to let you know that gatox will be taken care of showing the descriptions for the system updates, he is doing big changes in system settings
[18:18] <mandel_> barry, so he will take care of it in the right way
[18:19] <barry> mandel_: awesome, thanks.  will you be assigning #1215586 to gatox?
[18:19] <sergiusens> fginther, this problem started with flipped images or mtp; do you recall?
[18:19] <mandel_> barry, good point, yes I will
[18:20] <barry> mandel_: thanks.  i'm tracking that bug (on system-settings) to know when to commit my branch
[18:20] <fginther> sergiusens, I don't have good records, we've only been doing this reflashing since August
[18:21] <fginther> sergiusens, when did MTP start?
[18:22] <mandel_> barry, lp seems not to let me, can you set it for diegosarmentero
[18:22] <mandel_> barry, that is gatox username in lp
[18:22] <gatox> mandel_, barry yes, diegosarmentero
[18:23] <barry> gatox, mandel_ huh.   when i search for that in assignee, it comes up empty
[18:23] <sergiusens> fginther, ok, so it's flipped
[18:23] <mandel_> barry, awesome, so it was not me being stupid..
[18:23] <sergiusens> fginther, flipped happened in July
[18:23] <mandel_> gatox, can you try and assign it to you?
[18:23] <mandel_> gatox, https://bugs.launchpad.net/ubuntu-system-settings/+bug/1215586
[18:23] <gatox> done
[18:24] <sergiusens> fginther, the problem with flipped is that the usb connection is reset 3 times (2 without mtp)
[18:24] <barry> mandel_, gatox i was able to *unassign* those bugs, but nope, i couldn't find gatox
[18:25] <kenvandine> mandel_, look at lp:libaccounts-qt
[18:26] <mandel_> kenvandine, awesome, thx
[18:26] <kenvandine> np
[18:26] <barry> mandel_, gatox ok, so lp will let me assign the ubuntu-system-settings (ubutu) bug task, but not the project bug task.  i guess you have to be a member of the team owning that project to do that
[18:26] <kenvandine> mandel_, let me know if you have questions
[18:26] <mandel_> kenvandine, will do
[18:26] <mandel_> barry, oh well... so we have gatox name there, right? at least we have one to track
[18:27] <mandel_> barry, I'll ping didrocks on monday at CET time then
[18:27] <barry> mandel_, gatox yes.  just be sure to change the status on the right bug task when you commit the fix ;)
[18:27] <barry> sounds good
[18:27] <gatox> barry, ack
[18:28] <barry> oh hahahaha.  gatox now you're assigned to both.  thank you lp for clearing things up <wink>
[18:32] <mterry> Cimi, OK, I filed lp:~unity-team/ubuntu-system-settings/welcome-wizard against trunk.  I made some small cleanups, including unifying the pot file
[18:32] <mterry> Cimi, you may want to merge back into your working branch
[18:32] <mterry> Cimi, and let's see how seb128 and Laney like it
[18:35] <fginther> sergiusens, another failure in the loop, but I can still see it and shell in
[18:35] <fginther> http://paste.ubuntu.com/6383545/
[18:35] <sergiusens> fginther, back to adb; when it failed; did you ever try restarting adbd on the device?
[18:35] <fginther> sergiusens, I did not try that. How is that done?
[18:35] <sergiusens> fginther, that error is  after sleep 60?
[18:36] <fginther> sergiusens, yes
[18:36] <sergiusens> fginther, I know that wait for device is a crappy thing to rely on; eevn the scripts in the android tree have sleeps preceding the call to it
[18:36] <fginther> sergiusens, by the way, I'm available to hangout if you want
[18:37] <sergiusens> fginther, let me strt one
[18:37] <sergiusens> fginther, I'm not sure how to start a new hangout these days :-)
[18:38] <fginther> sergiusens, ring ring
[18:39] <sergiusens>  Sorry! The voice chat with Francis failed because of a problem with our servers at 3:39 PM. Please reload the page and try again.
[18:52] <cwayne> cjwatson: can click packages have dependencies on other click packages?
[18:59] <kenvandine> cwayne, no
[18:59] <kenvandine> just the sdk version
[18:59] <kenvandine> cwayne, at least that is my understanding... i am not an expert on click packages :)
[19:00] <cwayne> kenvandine: then how would a click app for an account-plugin work?
[19:00] <kenvandine> good question
[19:00] <kenvandine> maybe they will support it?
[19:00] <kenvandine> i guess we do need cjwatson :)
[19:04] <cwayne> i mean it could be packaged with an app, but that would kind of defeat the purpose
[19:06] <kenvandine> yeah
[19:09] <mhall119> cwayne: as far as I know there aren't QML bindings yet for any of the Unity APIs
[19:10] <cwayne> mhall119: ah, thanks
[19:10] <cwayne> was just looking for some new qml bindings to play around with :)
[19:10] <cwayne> kenvandine: speaking of click apps and account-plugins: https://code.launchpad.net/~cwayne18/+junk/account-plugin-click-hooks
[19:14] <kenvandine> cwayne, you can always play with the content hub bindings :)
[19:16] <cwayne> :)
[19:17] <kenvandine> cwayne, so what provides those hooks?
[19:17] <mhall119> +1 for content hub
[19:17] <cwayne> maybe i'll make it so you can choose a picture to make your fitbit picture
[19:17] <cwayne> kenvandine: a new package
[19:17] <kenvandine> i have a branch that works for starting from the export side
[19:17] <kenvandine> but tests fail
[19:17] <kenvandine> i need to finish that
[19:18] <kenvandine> mhall119, btw, those fixmes are all gone in trunk now, if you could regenerate the docs
[19:18] <cwayne> kenvandine: it seems like system-settings is hardocded to look in /usr/share/accounts/qml-plugins?
[19:18] <kenvandine> cwayne, but what does the hook accomplish?  did mardy get click installed plugins working?
[19:19] <cwayne> kenvandine: it copies the account-plugin into ~/.local/share/accounts/
[19:19] <cwayne> kenvandine: that's what we're trying to get working now :)
[19:19] <kenvandine> yeah, but i think the current container only loads them from the one directory
[19:19] <kenvandine> cool, mardy has made progress?
[19:19] <cwayne> not sure
[19:19] <kenvandine> i thought he was on holiday this week :)
[19:19] <cwayne> ssweeny and i are working on it now
[19:20] <kenvandine> oh, awesome!
[19:20] <kenvandine> so get me a patch for ubuntu-system-settings-online-accounts to make it load from multiple paths :)
[19:20] <cwayne> running into some issues now..
[19:28] <ssweeny> kenvandine, looks like it at least recognizes files in ~/.local/share/accounts/providers
[19:29] <cwayne> and services
[19:29] <kenvandine> ssweeny, that is probably unrelated to  ubuntu-system-settings-online-accounts
[19:29] <kenvandine> the backend stuff is probably smart
[19:29] <kenvandine> about DATADIR
[19:29] <kenvandine> etc
[19:29] <ssweeny> right
[19:29] <kenvandine> but the UI elements load from the one path
[19:30] <ssweeny> ah
[19:30] <kenvandine> there is this container for system settings that mardy wrote
[19:30] <kenvandine> that finds plugins by path
[19:30] <kenvandine> and qml components for them
[19:31] <kenvandine> not sure if  ubuntu-system-settings-online-accounts uses the API for that container from ubuntu-system-settings to find them
[19:31] <kenvandine> or if it has it's own code
[19:31] <kenvandine> probably it's own... you just need to make it smarter about where it loads
[19:31] <kenvandine> internally it might even just mangle the QML2_IMPORT_PATH
[19:33] <kenvandine> oh, maybe it's more complicated than that
[19:33] <kenvandine> src/module/OnlineAccountsPlugin.pc.in defines that path
[19:34] <kenvandine> oh, that's just some account plugins can find where to install
[19:34] <kenvandine> i guess signon-ui must be the consumer end of it
[19:34] <kenvandine> UOA has lots of pieces :)
[19:36] <dobey> s/lots of/too many/
[19:37] <ssweeny> this looks promising: ONLINE_ACCOUNTS_PLUGIN_DIR_BASE = share/accounts/qml-plugins
[19:37] <ssweeny> kenvandine, is a qml UI mandatory for the phone? on the desktop it seems like a webkit window just opens to the login page for a service
[19:38] <kenvandine> ssweeny, yes
[19:38] <dobey> ssweeny: depends on what you're doing. if it's generic oauth you can just use the generic oauth thing
[19:38] <kenvandine> yeah, i think most of them need it though
[19:38] <dobey> almost nobody does oauth exactly as specified
[19:38] <ssweeny> well the plugin i just dropped on the phone doesn't load the webkit
[19:38] <kenvandine> so it is in ubuntu-system-settings-online-accounts
[19:38] <ssweeny> good to know... :)
[19:39] <kenvandine> src/qml/constants.js.in:var qmlPluginPath = \"$${ONLINE_ACCOUNTS_PLUGIN_DIR}/\"
[19:39] <successus> salud
[19:39] <nik90> kenvandine: can you help me debug the 3rd party store selection
[19:40] <kenvandine> and then the qmlPluginPath is used in the various qml pages
[19:40] <kenvandine> source: qmlPluginPath + providerId + "/Main.qml"
[19:40] <kenvandine> so tricky to make it a list... :/
[19:40] <kenvandine> nik90, sure
[19:40] <ssweeny> kenvandine, cool, thanks
[19:40] <kenvandine> nik90, what are you trying to do?
[19:40] <nik90> kenvandine: here is the onclicked function of my import button http://paste.ubuntu.com/6383901/
[19:41] <nik90> kenvandine: I installed your exporter app which is a 3rd party source for pictures
[19:41] <nik90> kenvandine: and I created my own importer app which should show your app as on option
[19:42] <kenvandine>                         var peer = ContentHub.knownSourcesForType(ContentType.Pictures)
[19:42] <kenvandine> that returns a list
[19:42] <kenvandine> so you could make that peers
[19:42] <kenvandine> and then use that as a model
[19:42] <kenvandine> with a listview or something
[19:42] <kenvandine>  list<ContentPeer>
[19:43] <cwayne> ssweeny: that's gonna be a problem that it looks for a dir called providerID
[19:43] <nik90> kenvandine: ah okay. and then present that as a dialog where the user can select from
[19:43] <kenvandine> yup
[19:43] <t1mp> sil2100: you asked about the toolbar before. This MR landed: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/ap-toolbar-open/+merge/194122
[19:43] <kenvandine> nik90, and ContentPeer has name and id
[19:43] <cwayne> ssweeny: for the qml-plugins, i think we might have to actually write a script to do some more processing
[19:43] <kenvandine> name is the friendly display name, as it is in the .desktop file
[19:44] <kenvandine> and id is the appId
[19:44] <nik90> okay
[19:44] <mardy> ssweeny, cwayne: hi
[19:44] <nik90> let me create a dialog and other stuff and see how it goes
[19:44] <kenvandine> nik90, have you installed the exporter?
[19:44] <kenvandine> mardy!
[19:44] <mardy> kenvandine:  :-)
[19:45] <nik90> kenvandine: yes I installed the exporter
[19:45] <kenvandine> nik90, cool
[19:45] <kenvandine> mardy, having a good holiday?
[19:45] <ssweeny> mardy, hi
[19:45] <mardy> kenvandine: at home, with wife's parents looking after the baby, and me hacking on Mappero :-)
[19:45] <kenvandine> mardy, it isn't a holiday if you look at IRC :)
[19:45] <kenvandine> yay mappero!
[19:45]  * kenvandine is sooooo excited
[19:46] <mardy> kenvandine: I'm implementing search -- but I'm afraid it won't be ready by the weekend :-(
[19:46]  * ssweeny used mappero on his n900
[19:46] <ssweeny> pretty sweet app
[19:46] <cwayne> mardy: heya!
[19:47] <kenvandine> mardy, is mappero a later revision of the mapping app that was on the 770 ?
[19:47] <mardy> ssweeny, cwayne: so, I think you are looking at the correct place: u-s-s-o-a is the component which needs to be extended to support ~/.local/share, for the QML plugins
[19:47] <kenvandine> i think it was the 770 i had
[19:47] <kenvandine> mardy, that's a pain though... it's defined in javascript and used inline in qml
[19:47]  * ssweeny had the n810 and n900
[19:47] <sil2100> t1mp: AWESOME! Thanks :)
[19:47] <mardy> kenvandine: yes. It was maemo-mapper initially, then for the N900 I renamed it to Mappero and ported it to clutter; and finally now to Qt/QML :-)
[19:47] <ssweeny> me owning two devices was enough for nokia to abandon the platform
[19:48] <kenvandine> cool, i used it all the time on mine
[19:48] <kenvandine> i think it was the 770 i had, maybe 810
[19:48] <kenvandine> it was the 2nd maemo mid
[19:49] <kenvandine> ah, it was the 810
[19:49] <kenvandine> nice device :)
[19:49] <mardy> kenvandine: yeah, that one had the GPS, IIRC
[19:49] <kenvandine> yeah
[19:54] <mardy> cwayne: this probably doesn't affect you, but FYI: https://bugs.launchpad.net/ubuntu/+source/click/+bug/1245826
[19:54] <cwayne> mardy: ah, this does affect me :)
[19:57] <cwayne> mardy: sorry, I hadn't realized you were already working on the click hooks, I'd've just waited for you if i'd known :)
[19:58] <mardy> cwayne: no problem, I only ot started with .application hooks, so I'll make use of your work :-)
[19:58] <mardy> cwayne: but I think I'll let u-s-s-o-a install them
[19:58] <mardy> (we have already tons of projects for OA :-) )
[19:58] <kenvandine> mardy, could we have come up with a longer name for that?  i type that way too often :)
[19:59] <cwayne> mardy: makes sense to me :)
[19:59] <cwayne> mardy: have you heard anything on how theyw ant to distribute account-plugins though?
[19:59] <cwayne> like if it's a standalone click, apps would need to depend on it, but there's no dependencies with click
[20:00] <mardy> cwayne: no; OTOH the dependency is not hard, so I wonder if it could be handled by u-s-s-o-a
[20:01] <mardy> cwayne: that is, the app requests a "fitbit" account to u-s-s-o-a, which doesn't know anything about it
[20:01] <cwayne> hmm
[20:01] <mardy> cwayne: then maybe u-s-s-o-a could query the software center, and suggest the user to install the plugin
[20:01] <mardy> (just thinking out loud)
[20:02] <nik90> kenvandine: When I do property list<ContentPeer> peers the app does not even start
[20:02] <cwayne> hm, that seems like a good idea
[20:02] <nik90> I keep getting virtual void ContentHubPlugin::registerTypes(const char*)
[20:02] <nik90> file:///home/krnekhelesh/Documents/Clock/ContentHub/ContentHub.qml:19 created by hub
[20:02] <nik90> error
[20:03] <kenvandine> nik90, yeah, you have to treat it as an object
[20:03] <kenvandine> not ContentPeer
[20:03] <kenvandine> it can't be instantiated that way
[20:04] <kenvandine>             var peers = ContentHub.knownSourcesForType(ContentType.Pictures);
[20:04] <kenvandine> nik90, and it'll do the right thing
[20:05] <nik90> ah okay
[20:05] <mardy> kenvandine: hey, speaking of content HUB, what do you think about using it as a backend to implement https://qt.gitorious.org/qt/qtbase/source/becdfa6fabb80d35e430a13835f01a6ff3f6cc73:src/gui/kernel/qplatformdialoghelper.h#L215 ?
[20:06] <mardy> kenvandine: that is, the file dialog :-)
[20:06] <mardy> kenvandine: then any Qt app which wants to just open files wouldn't need to be ported but it could still benefit from it
[20:06] <kenvandine> mardy, you mean a generic file dialog?
[20:06] <nik90> kenvandine: do you know how to pass data to a dialog? I created a dialog which has a listview with a model. The model should be contentPeer model. But I do not know how to pass that to it
[20:06] <kenvandine> tricky... because it'll only have access to files it owns
[20:06] <nik90> kenvandine: I open the dialog using PopUtils.open(dialog)
[20:07] <kenvandine> in the dialog, you should be able to use dialog.peers ?
[20:07] <mardy> kenvandine: I mean, a QPA plugin which implements the Qt FileDialog APIs, which uses the content HUB
[20:08] <kenvandine> mardy, would it first let you select an app that has files, then browse that apps files?
[20:08] <mardy> kenvandine: I'm thinking of a dialog where on the left side you don't have the tree of folders to browse, but a list of content providers, and on the right the files they can offer
[20:08] <mardy> kenvandine: exactly :-)
[20:09] <kenvandine> it's certainly possible
[20:09] <kenvandine> but
[20:09] <kenvandine> the design expects each app to provide the picker
[20:09] <kenvandine> so each content provider would display it's own UI
[20:09] <kenvandine> which wouldn't work for this use case
[20:10] <kenvandine> unless... we make the registration richer
[20:10] <mardy> kenvandine: mmm... is it a QML item which can be embedded?
[20:10] <kenvandine> no... it's a separate process
[20:10] <kenvandine> which will be embedded eventually
[20:10] <kenvandine> but not in this sense
[20:10] <davmor2> Guy my phone has locked up on me again.  I've grabbed syslog and ~/.cache/upstart/unity8.log is there anything else useful I can get while it is in this state?
[20:10] <kenvandine> the shell would treat the source and destination apps as the same
[20:11] <davmor2> ogra_, pmcgowan, kgunn: ^
[20:11] <kenvandine> mardy, perhaps each source that registers could also define it's Content::Store
[20:11] <mardy> kenvandine: well, I don't really care if the file browser is in the same window as the application list, I mostly care about the client API
[20:11] <kenvandine> and the hub could provide a generic component that lets you list content in those stores
[20:11] <mardy> that would be nice
[20:12] <kenvandine> not sure how jdstrand would feel about that though :)
[20:12] <kenvandine> that would allow apps to be able to discover files from other apps
[20:12] <mardy> kenvandine: but not really needed, right? I mean, the "dialog" could be implemented by first picking the application provider, and then browsing its files, in two steps
[20:13] <jdstrand> that the current plan aiui
[20:13] <kenvandine> mardy, the current implementation could do a simpler version
[20:13] <kenvandine> list of all sources
[20:13] <kenvandine> and selecting a source
[20:13] <jdstrand> if we diverge from that, it needs discussion. in general, iiuc, that would be an info leak in our confinement
[20:13] <kenvandine> would use that source in picker mode to display
[20:13] <mardy> kenvandine: yep, that's fine
[20:13] <pmcgowan> davmor2, did you look at top to see if anything is out of control
[20:14] <successus> salud, hasta otro rato
[20:14] <kenvandine> that could be done today
[20:14] <kenvandine> one app at a time, not just listing all photos from all sources
[20:14] <davmor2> pmcgowan: just did everything seems to be at 0.0 0.0 except for top it's like everything died but the phone is still running adb
[20:15] <mardy> kenvandine: I just wonder though: the Qt API has namefilters ("*.jpg", for instance) and mimetype filters (though these seem to be used only in the QWidget API)
[20:15] <kgunn> davmor2: sorry was in mid chat...just caught up
[20:15] <pmcgowan> davmor2, ok, I had a lockup the other day and init was spinning so sounds different
[20:15] <kgunn> davmor2: i would've suggested same thing as pat
[20:15] <mardy> kenvandine: it might not be trivial to map these to the Content HUB types, right?
[20:15] <kenvandine> nope
[20:15] <kenvandine> it handles a well defined set of types
[20:15] <kenvandine> not file types
[20:15] <kenvandine> however
[20:15] <kgunn> davmor2: is the kernel still up and running ?
[20:16] <kenvandine> we might want to extend that to Content.Category and Content.Type
[20:16] <kgunn> davmor2: wonder if its in a state where top reporting is just "not correct"
[20:16] <mardy> kenvandine: I'd really like to have mime types :-)
[20:16] <kenvandine> but still doesn't really map to that
[20:16] <davmor2> pmcgowan: ah so it looks like stuff is up and top was slacking
[20:17] <kenvandine> i guess what we have is mime types, but as defined by the source provider instead of file extension
[20:17] <davmor2> http://ubuntuone.com/2PNtvAd4Carkw65Tc6XRON
[20:17] <mardy> kenvandine: at least, mappero can read/write GPX files, I don't think we want to make a new contentHUB release everytime that there's a new strange filetype to support?
[20:17] <kenvandine> so gallery-app handles content of type ContentType.Pictures
[20:17] <kenvandine> not *.jpg
[20:17] <jdstrand> I think mime probably needs to be handled when thinking about converged. I'm not up on the discussions
[20:17] <mardy> kenvandine: the problem is that it's an enum, not a string
[20:17] <kenvandine> yeah, but read/writing those files within the app doesn't matter
[20:17] <kenvandine> mardy, that is intentional
[20:18] <kenvandine> it was decided to make that an explicit list of well known types
[20:18] <kenvandine> at least for now
[20:18] <pmcgowan> kenvandine, I should loop mardy in on that other thread I started
[20:18] <kgunn> davmor2: be patient and let it run for a bit
[20:18] <kenvandine> pmcgowan, yeah, he probably has valuable input
[20:18] <mardy> kenvandine: but what if my app can handle JPG but not GIF?
[20:18] <kenvandine> mardy, why would you need the hub for your gpx files?
[20:18] <kgunn> davmor2: otherwise it look like indicators running at fractional % :)
[20:19] <mardy> kenvandine: someone will write a great GPX analyzer :-)
[20:19] <kenvandine> mardy, that is why we might need to extend it to Content.Category and Content.Type
[20:19] <mardy> kenvandine: or a sport tracker
[20:19] <kenvandine> where category would be photo
[20:19] <kenvandine> and type could be jpb
[20:19] <kenvandine> jpg
[20:19] <kenvandine> etc
[20:20] <mardy> kenvandine: you sure we are not reinventing mime types? image/jpeg sounds terribly close :-)
[20:20] <kenvandine> maybe we really use just be using categories
[20:20] <kenvandine> and relying on existing known mimetypes
[20:20] <kenvandine> yeah, that is what i wanted to avoid
[20:20] <kenvandine> but image != photo
[20:20] <kenvandine> from a users pov
[20:21] <mardy> kenvandine: right
[20:21] <kenvandine> the hard part comes in when you deal with an app that handles "video" but really can only handle ogg
[20:22] <kenvandine> we have no way of dealing with that
[20:22] <davmor2> kgunn, pmcgowan: so the system is alive and kicking just really low numbers in top.  The occasional item bubbles up and back down again
[20:22] <nik90> kenvandine: http://paste.ubuntu.com/6384128/
[20:22] <mardy> kenvandine: so, maybe Content.Category could be an enum which defines the source type (image vs photo, that is screenshot-tool vs camera-app), and Content.Type could be a real mime-type
[20:22] <nik90> kenvandine: the dialog gets peers I think, but then the data is not received by the listview
[20:22] <kgunn> davmor2: you on galaxy nexus?
[20:23] <kgunn> or n4 ?
[20:23] <davmor2> kgunn: maguro g nexus
[20:24] <mardy> kenvandine: so, Content.Category = Video, Content.Type = "video/ogg"
[20:24] <kenvandine> nik90, since this is just a js object... you might need to do something silly like peers.get(i, name)
[20:24] <kenvandine> and recreate the model
[20:24] <kenvandine> which is clearly too complicated
[20:24] <kenvandine> i think there was something painful about that
[20:25] <kenvandine> mardy, exactly
[20:25] <kenvandine> mardy, but... that makes it more complicated to give the user a list of sources
[20:25] <kgunn> davmor2: so, basically, the screen has not changed at all ?
[20:25] <kgunn> and doesn't respond to touch ?
[20:25] <kgunn> curious....does the clock change ?
[20:26] <kgunn> davmor2: just thinking...it could be input wonky...not necessarily the renderer
[20:26] <kgunn> ...assuming you can see the clock in the panel
[20:26] <kenvandine> mardy, maybe an app just needs to list all the mimetypes it supports and the importing app can list the mimetypes it supports
[20:26] <kenvandine> and we find potential sources
[20:26] <kenvandine> that's complicated too...
[20:26] <kenvandine> mardy, no good answer!
[20:27] <kenvandine> mardy, maybe mimetypes could be a filter provided by the app that wants the content
[20:27] <kenvandine> so they pick a source that provides videos
[20:27] <davmor2> kgunn: no the phone has basiaclly suspended so black screen and hitting power button doesn't wake it
[20:28] <kenvandine> and then the picking app can filter it's pick mode to files that match the mimetypes in a filter
[20:28] <kgunn> davmor2: ah...
[20:28] <davmor2> kgunn: let me try ringing it
[20:28] <kgunn> davmor2: good one
[20:29] <mardy> kenvandine: maybe the content HUB can ask the picking app to perform this filtering, before allowing it to be used as a source
[20:29] <mardy> kenvandine: I mean, just in case the number of supported files is 0
[20:29] <kenvandine> no... that would be painful
[20:29] <kenvandine> because it has to start the app
[20:29] <mardy> kenvandine: right
[20:30] <davmor2> kgunn: oh that's weird the land line was ringing like it had connected to the mobile but no response from the mobile
[20:32] <kgunn> ricmm: or racarr either of you know a potential wakeup signal you can send from the shell to a galaxy nexus ?
[20:32] <kgunn> davmor2: has the top list changed ? (like when you called it ?)...are there other processess running other than indicators ?
[20:33] <kgunn> davmor2: just curious, can you adb push/pull to the device...?
[20:34] <davmor2> kgunn: so when the call initially connects I get a huge change in top loads pop up,  1 second latter they all disappear again
[20:34] <davmor2> kgunn: yeap I can adb
[20:35] <kgunn> davmor2: is unity8 in the list ?
[20:36] <davmor2> kgunn: when I call it is but again only for a second
[20:37]  * nik90 will just wait for content hub to land in the sdk. Still a bit rough around the edges.
[20:44] <kgunn> davmor2: so i'm tempted to say...try relaunching unity8...at least that might squawk if its already alive
[20:45] <kenvandine> nik90, i have some code somewhere that has a working list of peers
[20:45] <kgunn> davmor2: do you know how?
[20:45] <nik90> kenvandine: that will help
[20:45] <nik90> kenvandine: I tried every way I can think of..the model is still not recognised
[20:45] <davmor2> kgunn: restart unity8 ?
[20:46] <kgunn> davmor2: well...just go into the shell of the device, then "sudo -u phablet -i"
[20:46] <kenvandine> http://bazaar.launchpad.net/~phablet-team/content-hub/trunk/view/head:/examples/import-qml/import.qml
[20:46] <kgunn> davmor2: then "start unity8"
[20:46] <kenvandine> nik90, ^^
[20:46] <kgunn> and see what happens
[20:47] <kgunn> davmor2: if it says job already running...then we know....otherwise, if it starts up & ui comes back...we know unity8 bailed somehow
[20:47] <davmor2> kgunn: start: Job is already running: unity8
[20:48] <kgunn> davmor2: ok
[20:48] <kgunn> thinking...
[20:48] <kenvandine> nik90, so what you were missing was modelData.name
[20:48] <kenvandine> since the item in the list is an object
[20:49] <nik90> kenvandine: it is hard to figure this out lookinng at the docs
[20:49] <davmor2> kgunn: the only thing I've had any joy with to get the system back up was pulling the battery putting it back in and powering it back on
[20:49] <kgunn> rsalveti: do you happen to know of a shell command to trip powerd on the device to an on state ? (davmor2 has a device where unity8 is running, but he's got a blank screen)
[20:49] <kenvandine> i'm assuming that is why modelData is needed, I've never really known why that is needed
[20:50] <kenvandine> nik90, yeah... i'll make sure there is examples in the docs that use knownSourcesForType
[20:50] <kgunn> davmor2: i hesitate to say do that...its nice to have a live but hung phone...just not sure what more poke points are available to us
[20:51] <kgunn> davmor2: i would grab all the logs of it...and check for crash log just in case (figuring you've already done that tho?)
[20:51] <robotfuel> davmor2: did you try sudo initctl restart powerd
[20:51] <kgunn> davmor2: ....mmm, can you just do start (not restart powerd?)
[20:52] <kenvandine> nik90, would it be clearer if you could define peers as a list<ContentPeer> ?
[20:52] <kgunn> hmmm...that's not what its called i guess....
[20:52] <davmor2> kgunn: so I grabbed syslog and ~/.cache/upstart/unity8 I'll have a look at crash now.
[20:52] <kgunn> i just tried on my device
[20:52] <kenvandine> nik90, instead of letting it just dynamically do it for you?
[20:52] <kgunn> davmor2: you might gran all the other logs out of upstart also...(if time stamps make sense)
[20:52] <nik90> kenvandine: yeah..then retrievingn data of the model will be similar to the contentItems
[20:52] <kenvandine> right
[20:53] <kgunn> davmor2: did you have many apps open ?
[20:53] <nik90> which should make it easier
[20:53] <davmor2> kgunn: start: Unknown job: powerd
[20:53] <davmor2> kgunn: no apps open at the time
[20:53] <kgunn> davmor2: yeah me too on powerd
[20:53] <davmor2> let me grab the rest of the logs
[20:54] <kgunn> davmor2: oh duh...you have to sudo initctl start powerd
[20:54] <kgunn> like robotfuel said
[20:54] <kgunn> try that
[20:55] <davmor2> kgunn: initctl: Job is already running: powerd
[20:55] <robotfuel> davmor2: when I want to turn on the screen I use restart
[20:55] <kgunn> davmor2: ok...hmmm....now try the restart like robotfuel said
[20:56] <kgunn> davmor2: if it all comes back, then that might point to race in powerd/unity8/mir menage a trois
[20:59] <davmor2> kgunn: powerd start/running, process 2454 and still dead
[21:00] <kgunn> davmor2: hmmm....ok....now it might be worthy to try, stop unity8....make sure its happy, then start unity8
[21:02] <davmor2> kgunn: stopped unity8,  restarted powerd, started unity 8 now I have a working device again
[21:03] <davmor2> kgunn: stopping unity8 and restarting by itself didn't do anything
[21:04] <kgunn> davmor2: hmmm.....did you drop it ? :)
[21:04] <davmor2> kgunn: nope
[21:04] <kgunn> davmor2: restarting unity8 would restart mir as well....it would pretty much restart everything but the kernel
[21:05] <kgunn> davmor2: so that's the other potential problem area...kernel maybe ?
[21:05] <robotfuel> davmor2: you are not out of disk space? I did that once and had the same issue.
[21:05] <davmor2> kgunn: it's only been happening on and off since image 13 and it is completely random as to when it happens
[21:06] <kgunn> robotfuel: interesting...is image13 on the hairy edge for mem use ?
[21:06] <kenvandine> nik90, i have a branch that makes ContentPeer creatable in the QML bindings
[21:06] <robotfuel> kgunn: no I filled it up with apps and log files.
[21:06] <davmor2> robotfuel: kgunn:  Not that I'm aware of how ever there is a boat load of logs just copied over let me see how much space that was using
[21:07] <davmor2> 1.8 MB so not that much
[21:08] <robotfuel> davmor2: df -h shows a lot of space available?
[21:08] <davmor2> robotfuel, kgunn: http://paste.ubuntu.com/6384362/
[21:09] <nik90> kenvandine: okay, I am trying to get the peers working
[21:10] <nik90> that should finish my importer app before I jump to the exporter
[21:10] <davmor2> robotfuel, kgunn: only thing anywhere near is lib/module/
[21:12] <kgunn> davmor2: if you got all the logs...and we've tried all this stuff...i guess just type "reboot"
[21:12] <davmor2> kgunn: will do
[21:13] <davmor2> I need to get off I'll file a bug over the weekend
[21:20] <kenvandine> nik90, this branch should make it easier, and i included a QML example for the docs https://code.launchpad.net/~ken-vandine/content-hub/creatable_content_peer/+merge/194597
[21:26] <cgregan> achiang: ping
[21:27] <cgregan> Hello all. Wondering if there are any SDK devs on the channel atm?
[21:30] <cgregan> pmcgowan: you might be a good resource
[21:31] <pmcgowan> cgregan, you are desperate
[21:31] <cgregan> hehe
[21:31] <cgregan> pmcgowan: we are using the SDK for driver testing and used toolbar for button on the bottom of our window
[21:32] <cgregan> pmcgowan: now what used to be a static button area hides
[21:32] <cgregan> pmcgowan: => problem
[21:32] <pmcgowan> ah
[21:33] <cgregan> pmcgowan: is there a better object to use here, or do we need a "lock" switch on the that toolbar?
[21:33] <pmcgowan> cgregan, I think its the latter, trying to look at the docs
[21:33] <cgregan> thanks
[21:36] <pmcgowan> cgregan, there is a locked property
[21:36] <pmcgowan> also and opened prop
[21:36] <cgregan> pmcgowan: \o/
[21:36] <pmcgowan> cgregan, on toolnaritems
[21:37] <cgregan> pmcgowan: can you send me that spec so we can call it with the locked prop?
[21:37] <pmcgowan> looking here:
[21:37] <pmcgowan> http://developer.ubuntu.com/api/devel/ubuntu-13.10/qml/ui-toolkit/qml-ubuntu-components0-toolbaritems.html
[21:37] <pmcgowan> I assume the are writeable
[21:38] <pmcgowan> example uses them
[21:39] <cgregan> pmcgowan: awesome! Thanks
[21:39] <pmcgowan> np
[21:54] <t1mp> cgregan: is the problem that your toolbar is locked, but still (automatially) hides?
[21:54] <t1mp> cgregan: that is this bug https://bugs.launchpad.net/ubuntu-rssreader-app/+bug/1248759
[21:54] <t1mp> cgregan: I fixed it in UITK trunk. It will take a bit of time before it arrives in the UITK package.
[21:55] <cgregan> t1mp: yes..that is the problem
[21:55] <cgregan> t1mp: how much time?
[21:57] <t1mp> cgregan: I am not sure. We should have daily releases, but at the moment we switched to manual because some important changes (to autopilot tests) have to land.
[21:57] <kenvandine> i think robru is working on getting that landed
[21:58] <cgregan> ok..but by the end of next week?
[21:58] <t1mp> I'm not sure if the daily releases are active again. If yes, it should be there tomorrow. If not, I would say Monday.
[21:58] <t1mp> cgregan: yes
[21:58] <cgregan> awesome news t1mp...made my weekend! Beer'o clock!
[21:58] <robru> t1mp, cgregan: i intend to release this within an hour. just fighting an infrastructure issue at the moment
[21:59] <cgregan> cool...MOnday is perfect
[21:59] <cgregan> thanks
[21:59] <kenvandine> robru, remember all that infrastructure is going down for the weekend
[21:59] <robru> ugh
[21:59] <robru> manual merge time!
[21:59]  * kenvandine isn't sure what time
[22:00] <t1mp> kenvandine: why?
[22:01] <kenvandine> migrating data centers
[22:01] <t1mp> ah nevermind, it is in some email :)
[22:01] <t1mp> kenvandine: ah, yes.
[22:02] <kenvandine> looks like there were some delays, so might start later than expected
[22:04] <robru> kenvandine, k, building & will publish asap. hopefully this gets in
[22:04] <kenvandine> i guess it won't go down until tomorrow, i re-read the mail
[22:04] <robru> ah
[22:04] <t1mp> robru: thanks for merging https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/locked-toolbar-autohide/+merge/194565
[22:04] <kenvandine> or rather sometimeafte 2400 UTC tonight :)
[22:04] <robru> t1mp, no worries. jenkins was being fussy
[22:14] <botfap> hi all, can anyone help getting trusty touch on to a nexus 7 3g?
[22:23] <fginther> sergiusens, updated https://bugs.launchpad.net/phablet-tools/+bug/1249162
[22:23] <fginther> sergiusens, restarting adbd on the phone helped
[22:25] <sergiusens> fginther, great, thanks; then it is surely adbd that's breaking
[22:47] <fginther> sergiusens, what are your thoughts on using "system-image-cli --build 0 --verbose" as a lightweight flash?
[22:48] <kenvandine> fginther, that's basically what i do on a regular basis
[22:48] <fginther> kenvandine, ever see any issues that might break automation?
[22:49] <kenvandine> nope
[22:49] <kenvandine> system-image-cli  -c trusty-proposed -b 0 -v
[22:49] <kenvandine> i do that a few times a week
[22:49] <kenvandine> it's never failed
[22:50] <kenvandine> just to cleanup anything i've done myself
[22:50] <kenvandine> installing debs, etc
[22:53] <fginther> kenvandine, reverting installed debs is the primary reason to flash, so good to hear
[22:54] <random_> everyone afk?
[22:55] <Minilodon> no...was just peaking in to see if anyone had attempted to build for the nexus 5
[22:56] <random_> do I need a USB drive to Install to a nexus 10 7 .
[22:59] <sergiusens> fginther, you won't get a clean build that way though
[23:00] <sergiusens> fginther, as in all prev installed click packages and a populated home et.al.
[23:01] <sergiusens> fginther, we are doing something similar with phablet-flash already; also take into account that you will need to download the full image everytime instead of once
[23:01] <sergiusens> fginther, what I can do is try and add support for delta upgrades with a --bootstrap attached to phablet-flash
[23:02] <sergiusens> fginther, last but not least; given what we discussed about adb, this won't get you out of the bug
[23:02] <fginther> sergiusens, people have been asking me about that as a workaround, I needed to know more
[23:03] <fginther> thanks
[23:03] <fginther> sergiusens, I'd rather get the problem with adbd resolved.
[23:04] <sergiusens> fginther, well phablet-flash creates the same ubuntu_commands as system-image cli or one fairly similar if you don't do a --bootstrap
[23:05] <sergiusens> and just adds a wipe/format data if you choose to wipe
[23:10] <fginther> sergiusens, ah I see, it still has to the same recovery-reboot (or whatever it's doing to get the little android guy to do it's thing)