[06:58] <sil2100> didrocks: morning! Robert fixed the webapps stack yesterday, but as there is a lot of packaging changes, well, I'll have to ask you to ACK them ;p
[07:00] <sil2100> didrocks: I browsed them through yesterday and mostly they look so fine, although it's easy to miss something with 20 diffs
[07:00] <sil2100> They're usually standard packaging changes
[07:04] <didrocks> sil2100: hey! sure, having a look :)
[07:05] <didrocks> sil2100: FYI, I'm building the mir stack right now for the unity-notification new component and will publish it
[07:06] <sil2100> didrocks: ah, it's in mir?
[07:07] <sil2100> didrocks: is it tightly bound with the MIR server?
[07:07] <didrocks> sil2100: as there is no API/ABI stability between all unity8 components, I think it's easier for now to put everything in the mir stack
[07:15] <MacSlow> Saviq, hey there
[07:16] <sil2100> Makes some sense, yes
[07:20] <didrocks> sil2100: +1 from me on the webapps (reviewed them all), I'll have to NEW some, I pulled the configuration on the archive admin side
[07:21] <didrocks> Saviq: unity-notifications NEWed in universe, under dailies now (without the integration tests running though)
[07:23] <sil2100> \o/
[07:24] <sil2100> Publishing!
[07:25] <didrocks> sil2100: once that done, tell me if we should reenable the indicator-clients
[07:25] <didrocks> sil2100: https://code.launchpad.net/~mathieu-tl/cupstream2distro-config/indicators-client_enable/+merge/170105
[07:28] <sil2100> didrocks: let's do that - let me just check that chewie thing
[07:29] <didrocks> sil2100: I think you already did see the numerous FTBFS, right?
[07:30] <sil2100> didrocks: yes, at first glance it looked like the usual issue with slow machines and any <-> all
[07:31] <didrocks> sil2100: yeah, proposed stroke :( see http://irclogs.ubuntu.com/2013/06/20/%23ubuntu-devel.html#t05:49
[07:31] <didrocks> sil2100: if we kind of see that again, we'll remove it I guess
[07:31] <didrocks> sil2100: so rebuilding in the ppa the failures and relaunching the stacks with "foo", right?
[07:31] <sil2100> didrocks: no other sane choice!
[07:32] <sil2100> Doing
[07:32] <didrocks> sil2100: unfortunately… :/
[07:33] <sil2100> didrocks: huh, actually I see that libusermetrics just failed normally! Symbols problem (as we already noticed yesterday)
[07:34] <sil2100> didrocks: I think it would be much less troublesome to have the symbols file removed for this lib
[07:35] <sil2100> didrocks: will poke Pete again once he's up
[07:35] <didrocks> sil2100: depends, it passed once, so it means that it doesn't change depending on the archs, right?
[07:36] <didrocks> sil2100: so it just means that the ABI changed and it wasn't updated, right?
[07:36] <didrocks> (as it should have been updated)
[07:38] <sil2100> True, just hope it won't change with the toolchain change or something
[07:39] <sil2100> I'm always weary when I see those mangled symbols, there's not much guarantee on those ;)
[07:39] <didrocks> sil2100: yeah, let's try and get upstream to update it
[07:39] <didrocks> sil2100: if we see there are too much changes and they don't garantee their ABI, we can remove it for now
[07:42] <sil2100> k
[07:43] <sil2100> Not sure if it makes sense to re-run the indicators stack with this not fixed - I'll prepare a merge for modifying those symbols for upstream to review
[07:44] <sil2100> So that it's faster once pete pops up
[07:45] <didrocks> sil2100: agreed! just ensure you are making him aware of the change so that this doesn't happen everyday. Would be nice to know also why the upstream merger didn't block him first? (if it was a direct push to trunk, that should cease)
[07:46] <sil2100> hm, true
[07:46] <sil2100> Let me ask QA about that
[07:57] <sil2100> eh
[07:57] <sil2100> I just tried using bzr lp-propose on my system
[07:57] <sil2100> RuntimeError: the sip module implements API v9.0 to v9.2 but the PyQt4.QtCore module requires API v10.0
[07:57] <sil2100> -_-
[07:57] <sil2100> from PyKDE4.kdeui import KWallet
[07:58] <pstolowski> heyas, for everyone who flashed their phone with latest image and lost everything but apps in the dash - the culprit is new stack that just landed, and I forgot about two new dependencies. for now you need to apt-get unity-scope-home demo-assets-scopes and reboot. in the meantime I'll fix the deps
[08:04] <tsdgeos> Saviq: why does build -s still branch hud and unity, doesn't the saucy ones work for us?
[08:06] <tsdgeos> and we need for HW for running qmluitests
[08:06] <tsdgeos> the queue is empty and my job has been waiting for 5min already
[08:07] <tsdgeos> MacSlow: http://paste.kde.org/~tsdgeos/778466/ ?
[08:12] <tsdgeos> pstolowski: installing unity-scope-home wants to kill unity-lens-shopping on the desktop, is that ok?
[08:12] <pstolowski> tsdgeos: are you on R? it's ok, shoppins lens is gone
[08:12] <tsdgeos> i'm on saucy
[08:13] <pstolowski> tsdgeos: ok, anyway, that's fine
[08:15] <pstolowski> sil2100: ping
[08:19] <tsdgeos> mmrazik: been waiting for 20 minutes for "next available executor on ps-saucy-server-amd64-1", is that normal?
[08:20] <pstolowski> didrocks: : do you know if recommended packages are installed on the phablet image? it's seems they aren't judging from the missing home scope?
[08:21] <didrocks> pstolowski: I heard they disabled the recommends installation by default
[08:21] <pstolowski> didrocks: in that case I'll move it to required
[08:21] <didrocks> pstolowski: no, please don't
[08:21] <didrocks> pstolowski: they should fix their image
[08:21] <didrocks> or list the scopes themselves
[08:21] <didrocks> but we don't want to force people to have them installed them
[08:22] <didrocks> so recommends is the right relation
[08:22] <pstolowski> didrocks: and what about demo-assets-scopes?
[08:23] <didrocks> pstolowski: hum, is that needed for all images?
[08:24] <didrocks> pstolowski: if it's not, they should just be directly listed in their seed
[08:25] <pstolowski> didrocks: demo-assets-music and demo-assets-videos provides the scopes already and they are installed; demo-assets-scopes provides just a gsettings schema override and it's needed for them
[08:26] <mhr3> pstolowski, theoretically it's not needed, you could just as well change the schema yourself
[08:26] <pstolowski> didrocks: ok, I see, it seems in both cases it needs to be handled by their seed, as demo-assets-scopes depends on demo-assets-music and demo-assets-videos
[08:28] <didrocks> pstolowski: yeah, so directly in their seed
[08:28] <pstolowski> didrocks: thanks
[08:28] <didrocks> pstolowski: like for the netbook edition image at the time, this is how we list the delta
[08:28] <didrocks> pstolowski: yw :)
[08:28] <pstolowski> didrocks: do you know who is the right person to ping for this?
[08:28] <didrocks> pstolowski: thanks for asking before modifying :)
[08:29] <didrocks> pstolowski: I guess sergiusens, ogra or rsalveti
[08:29] <didrocks> (I think sergio is on holidays though)
[08:29] <pstolowski> didrocks: cool, ta
[08:29] <didrocks> yw :)
[08:33] <mmrazik> tsdgeos: let me check
[08:33] <mmrazik> but I fear it is :-/
[08:33] <mmrazik> well.. it should not
[08:33] <mmrazik> tsdgeos: which job?
[08:34] <mmrazik> might be misconfiguration of the job
[08:34] <mmrazik> tsdgeos: looks like the job wants a specific node and if something else is running there it doesn't pick up one of the others which are available
[08:34] <tsdgeos> mmrazik: ah, i see, i just did a rebuild of my last job
[08:35] <tsdgeos> mmrazik: unity-phablet-qmluitests-saucy
[08:35] <tsdgeos> job #135
[08:35] <mmrazik> tsdgeos: yeah... it requires that specific node which is not ok
[08:35] <mmrazik> tsdgeos: but it might be there for purpose
[08:35] <mmrazik> mzanetti: ^^
[08:36] <mmrazik> mzanetti: do you know why unity-phablet-qmluitests-saucy requires ps-saucy-server-amd64-1 node?
[08:36] <mmrazik> mzanetti: is it because of the changes you are doing to the VMs?
[08:36] <mmrazik> and which are not deployed to all of them?
[08:36] <mzanetti> mmrazik: yes... I'm on it to prepare something to speed up tsdgeos
[08:37] <mmrazik> tsdgeos: so its expected and fix is on the way
[08:37] <mmrazik> srry for the inconvenience in the meantime :/
[08:40] <mzanetti> mmrazik: just stumbled over this again: #mmrazik: adding libqt5quick5 libqt5qml5 to workaround some dependency issues
[08:41] <mzanetti> mmrazik: can we get rid of this or is it still needed?
[08:41] <pete-woods> guys, do I need to be on Saucy to build unity8 now? libunity is complaining about a lack of dee>=1.2.5
[08:42] <mzanetti> pete-woods: not exactly sure if it should still work with raring
[08:42] <mzanetti> pete-woods: we're all on saucy
[08:42] <pete-woods> okay
[08:42] <Saviq> morning
[08:42] <pete-woods> guess I should upgrade then
[08:42] <mzanetti> hey Saviq
[08:43]  * Saviq flashes his phone and crosses his fingers
[08:43] <pete-woods> mzanetti: thanks for the infos :)
[08:43] <mmrazik> mzanetti: I think it can be removed
[08:44]  * mzanetti whishes best luck to saviq :D
[08:44] <tsdgeos> Saviq: there's 2 packages you need, but looks ok
[08:44] <tsdgeos> Saviq: any clue about http://paste.kde.org/~tsdgeos/778466/ ?
[08:44] <tsdgeos> desktop-wise
[08:44] <Saviq> tsdgeos, qtdeclarative5-unity-notifications-plugin
[08:44] <Saviq> from desktop-deps
[08:45] <tsdgeos> grrrrrr
[08:45] <Saviq> mzanetti, uh, we forgot to add the demo scopes to the image <facepalm>
[08:45] <mzanetti> Saviq: yeah... I've seen the mail from Pawel
[08:45] <tsdgeos> Saviq: https://launchpad.net/~phablet-team/+archive/desktop-deps/+packages contains no such plugin?
[08:48] <Saviq> tsdgeos, ah, /my bad
[08:48] <Saviq> tsdgeos, was waiting for the old version to go away from there
[08:48] <Saviq> and didn't copy it back from ppa:phablet-team
[08:48]  * Saviq does now
[08:49] <Saviq> tsdgeos, actually https://launchpad.net/ubuntu/saucy/+source/unity-notifications/0.1.0daily13.06.20-0ubuntu1
[08:49] <Saviq> didrocks, thanks! ↑
[08:49] <Saviq> lol, /me loves the changelog
[08:50] <seb128> Saviq, good work on getting those in the archive!
[08:50] <tsdgeos> Saviq: does that mean that should be on the regular archive?
[08:50]  * tsdgeos has nothing either
[08:50] <Saviq> seb128, next stop - unity8
[08:51] <Saviq> seb128, about ↑↑
[08:51] <Saviq> seb128, I didn't yet get the unity update from yesterday either...
[08:51] <Saviq> seb128, using the global mirror
[08:52] <Cimi> mzanetti, ping dude :)
[08:52] <Saviq> tsdgeos, just grab the package and install manually in the mean time
[08:52] <Saviq> https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4729845
[08:52] <seb128> Saviq, hum, that's weird ... dpkg -l | grep unity?
[08:52] <Saviq> seb128, yeah, daily13.06.07 for the most part
[08:53]  * Saviq tries disabling apt-cacher-ng again
[08:53] <mzanetti> Cimi: pong dude :D
[08:54] <Cimi> mzanetti, I have an issue with tests
[08:54] <Cimi> mzanetti, and I am confused :)
[08:54] <mzanetti> so far nothing new :P
[08:54] <Cimi> mzanetti, have a look at calendar and calendarmenu tests (cc Saviq )
[08:54] <Cimi> mzanetti, ahahaha :P
[08:54] <mzanetti> Cimi: ok, I'll have a look
[08:54] <Cimi> mzanetti, so sometimes the date objects are undefined
[08:55] <Cimi> the ones that arrive at the calendar component
[08:55] <Cimi> I have a race
[08:55] <nic-doffay> Saviq, had a chance to look at that branch yet?
[08:56] <Saviq> nic-doffay, not yet, got to sleep at 4am, making sure the new stuff lands
[08:57] <nic-doffay> Saviq, sheesh
[08:57] <nic-doffay> You feeling ok today?
[08:57] <Saviq> nic-doffay, yeah, nothing new ;)
[09:00] <Cimi> mzanetti, QWARN  : qmltestrunner::Calendar::test_maximumDate(row 0) file:///home/cimi/Development/indicators-client/system-components/SystemComponents/SettingsComponents/Calendar/Calendar.qml:161: TypeError: Cannot call method 'getDate' of undefined
[09:01] <Cimi> (which might be my bug...)
[09:05] <Saviq> seb128, got the updates, not sure where it got stuck before (I tried disabling the cache yesterday, too)
[09:06] <seb128> Saviq, ok, weird, it migrated to the archive yesterday, I got it around 6pm european time
[09:08] <mzanetti> Cimi: hmm... can't reproduce here
[09:08] <Cimi> mzanetti, can be a race in my calendar code
[09:08] <tsdgeos> lol, there's somethig weird with the greeter slideme to the left animation
[09:08] <Cimi> mzanetti, happens sometimes
[09:08] <mzanetti> Cimi: you didn't push yet?
[09:08] <tsdgeos> it's cool, but if you try to slide left after the hint has happened
[09:08] <tsdgeos> it won't work :D
[09:08] <Cimi> no no it has always been
[09:08] <Cimi> mzanetti, ^
[09:09] <mzanetti> tsdgeos: hmm... noted... thanks
[09:09] <mzanetti> Cimi: can you describe it a bit more?
[09:10] <mzanetti> Cimi: where should I look for it?
[09:10] <Cimi> mzanetti, I have to run make testCalendar until I have the warning
[09:10] <mzanetti> tsdgeos: your vm for debugging should be ready soon... Jenkins reverted it on half way so I had to start again
[09:11] <tsdgeos> ok, thanks
[09:11] <mzanetti> Cimi: I've ran them now 20 times in a row. no failure
[09:12] <mzanetti> can you pastebin the output you get when it fails?
[09:12] <mzanetti> Cimi: ^
[09:12] <Cimi> mzanetti, I did
[09:12] <Cimi> ^^
[09:12] <Cimi> before
[09:12] <mzanetti> Cimi: oh... thats all :/
[09:12] <Cimi> yeah
[09:12] <Cimi> text: dayStart.getDate()
[09:16] <mzanetti> Cimi: does the test fail because of this?
[09:16] <mzanetti> Cimi: looks "just" like a warning to me
[09:17] <Cimi> mzanetti, no it doesn't fail
[09:17] <Cimi> mzanetti, but it's still a warning
[09:17] <mzanetti> Cimi: yeah. just want to understand whats happening
[09:18] <mzanetti> Cimi: the only thing I could imagine is this (note that's just a guess):
[09:18] <mzanetti> Cimi: dayStart is the result of quite a long chain of properties
[09:18] <Cimi> yes indeed
[09:18] <mzanetti> Cimi: and also quite some functions in that chain
[09:18] <Cimi> we change the dates, the model repopulates
[09:19] <Cimi> thus the dates are still in creation
[09:19] <Cimi> but the test proceeds
[09:19] <mzanetti> so it might postpone recalculating all the chain in the event loop and in the meantime the test tries to read it already
[09:19] <mzanetti> next time the event loop runs its successfully there and everything is fine. but the warning is already printed
[09:20] <mzanetti> Cimi: try adding some debug prints. eg: onDayStartChanged: print("dayStart now updated:", dayStart)
[09:20] <mzanetti> and try to figure the order of the properties changing and the message getting printed
[09:20] <Cimi> mzanetti, that might fix the race :)
[09:20] <mzanetti> Cimi: maybe, or maybe make it worse
[09:21] <Cimi> mzanetti, I could add a trycompare somewhere
[09:21] <mzanetti> Cimi: yeah... try with a initTestCase() { wait(0) }
[09:21] <Cimi> mzanetti, mm no
[09:22] <mzanetti> :D
[09:22] <Cimi> mzanetti, I believe it's caused by the change to maximumDate in the test
[09:22] <Cimi> mzanetti, which is what triggers the whole chain
[09:22] <Cimi> I might add wait(0) after changing that
[09:22] <mzanetti> ah... yeah... as I said, I can't reproduce so I don't know when this happens
[09:23] <Cimi> nope
[09:23] <Cimi> still happens
[09:23] <mzanetti> Cimi: please paste the whole output
[09:24] <Cimi> mzanetti, http://paste.ubuntu.com/5783077/
[09:25] <Cimi> they're random :(
[09:25] <Cimi> now line 120 too
[09:25] <Cimi> actually it's the same
[09:25] <Cimi> it's dayStart
[09:26] <tsdgeos> argggg
[09:27] <tsdgeos> can we make ./run_on_device not run if make fails
[09:27] <tsdgeos> i've been wondering for like 5 minutes why all my debugs did not show up
[09:30] <Cimi> mzanetti, don't think there's a way to make the code wait for a valid dayStart
[09:32] <mzanetti> Cimi: well... there is one thing:
[09:33] <mzanetti> Cimi:  instead of doing model: __diffMonths(min, max) it would be better to create a ListModel {}
[09:33] <mzanetti> Cimi: and onMinMaxMonthChanged: do some listModel.append() and listModel.remove()
[09:33] <Cimi> mmm ok
[09:33] <mzanetti> Cimi:  that has the advantage that not everything needs to be recreated
[09:34] <Cimi> I see, valid point mzanetti
[09:39] <mzanetti> tsdgeos: ready!
[09:41] <rsalveti> pstolowski: didrocks: I updated the seeds already to include demo-assets-scopes and home by default
[09:41] <didrocks> rsalveti: excellent, thanks!
[09:41] <rsalveti> any other lens/scopes you want there?
[09:41] <pstolowski> rsalveti: great, thanks
[09:41] <rsalveti> saucy-19 should have them
[09:41] <rsalveti> still downloading to see
[09:42] <pstolowski> rsalveti: no, that's all for now
[09:42] <rsalveti> pstolowski: great
[09:43] <didrocks> Saviq: yw ;)
[09:45] <Saviq> rsalveti, dude, don't you sleep?
[09:46] <Saviq> rsalveti, and thanks lots!
[09:46] <rsalveti> Saviq: haha, just woke up
[09:46] <rsalveti> fixed the image and went to bed
[09:46] <Saviq> rsalveti, yeah, saucy-19 is good
[09:46] <rsalveti> awesome
[09:46] <Saviq> rsalveti, and the removed packages - nice :D
[09:46] <didrocks> rsalveti: I think you should make a launchpad team for little/never sleeper, I know infinity will be there as well :p
[09:46] <rsalveti> Saviq: yeah, long list :-)
[09:47] <rsalveti> hahah, infinity is a bot
[09:47] <Saviq> rsalveti, Pat will be happy - 12MBs saved ;d
[09:47] <rsalveti> Saviq: haha, yeah :-)
[09:48] <didrocks> Saviq: -nux -notify-osd ?
[09:48] <Saviq> didrocks, not -nux, unity-core still uses it
[09:48] <Saviq> didrocks, http://paste.ubuntu.com/5782399/
[09:49] <Saviq> don't ask me what pulled aspell in ;)
[09:49] <didrocks> Saviq: ahah :) nice!
[09:51] <seb128> Saviq, how come gstreamer is droping off?
[09:51] <Saviq> seb128, don't ask me :D
[09:51] <seb128> ok ;-)
[09:51] <Saviq> seb128, we only really dropped notify-osd :D
[09:51] <seb128> the e-d-s stack is dropping off because the people lens got dropped I think
[09:52] <seb128> I'm surprised gtk is not dropped though
[09:53] <seb128> since dconf-tools/editor got dropped which is what was brining gtk in last time I checked
[09:53] <seb128> need to check again
[09:53] <Saviq> right, a lot went away due to people lens going away, true
[09:55] <rsalveti> seb128: gstreamer is back with latest image
[09:55] <seb128> rsalveti, great ;-)
[09:55] <Saviq> rsalveti, hmm any idea "????????????    no permissions"?
[09:56] <rsalveti> Saviq: what?
[09:56] <rsalveti> where?
[09:56] <Saviq> rsalveti, maguro via usb
[09:56] <Saviq> rsalveti, I can actually connect via TCP
[09:56] <rsalveti> weird, is this with the unflipped image?
[09:56] <Saviq> rsalveti, yeah, just flashed saucy-19
[09:56] <Saviq> rsalveti, and it's the same in recovery
[09:57] <Saviq> rsalveti, can I manually flash -b?
[09:57] <rsalveti> weird, try killing adbd from your host
[09:57] <rsalveti> sure, it's just quite a few steps to reproduce that manually
[09:58] <rsalveti> usually when I get some weird behaviors I just kill adbd in my host and try adb again
[09:58] <rsalveti> yeah, 19 is looking good
[09:58] <Saviq> rsalveti, no adbd running locally
[09:58] <rsalveti> just missing people lens :-(
[09:59] <Saviq> rsalveti, just adb fork-server
[09:59] <rsalveti> sorry, should just be adb
[09:59] <rsalveti> yeah
[09:59] <Saviq> yeah, killed
[09:59] <mzanetti> vesar: hey man
[09:59] <rsalveti> you'll noticed it'll be started again when you use adb again
[09:59] <Saviq> rsalveti, yeah, doesn't help :/
[09:59] <mzanetti> veebers: I made the height of the foldingArea configurable by a parameter and shrinked it by itemSize/4. seems to work fine and still looks good
[10:00] <mzanetti> vesar: that was for you ^^
[10:00] <rsalveti> Saviq: make no sense, so to reproduce -b manually
[10:00] <rsalveti> download both the boot and system img
[10:00] <rsalveti> fastboot format system
[10:00] <rsalveti> fastboot format userdata
[10:01] <rsalveti> fastboot flash boot <boot.img>
[10:01] <rsalveti> fastboot flash system <system.img>
[10:01] <rsalveti> boot to recovery
[10:01] <rsalveti> and adb push ubuntu-touch.zip /sdcard/autodeploy.zip
[10:01] <rsalveti> adb reboot recovery
[10:03] <mhr3> Saviq, had the "no permissions" error as well, unplugged usb, replugged and it worked again
[10:04] <vesar> mzanetti, so you got it working! great. can we see it somehow?
[10:04] <rsalveti> got to love these weird adb errors
[10:05] <mzanetti> vesar: pushed it to the branch
[10:06] <mzanetti> vesar: noticed performance issues on the galaxy nexus tho :(
[10:06] <mzanetti> vesar: however, I have a handful of ideas how to improve it. so no big deal right now. but still, needs to be done still
[10:07] <vesar> mzanetti, oh, due to the new changes you needed made?  Good to hear you  have fix ideas already:)
[10:08] <mzanetti> vesar: no... I think its a bit much to calculate the angle, the offset, the opacity and the brightness during the movement. but lots of that code is duplicate and can be merged if being a bit clever
[10:08] <vesar> mzanetti, hey looks good to me man! I'll have a chat with Martin to hear what he thinks about it.
[10:08] <mzanetti> vesar: also the brightness and the angle can be merged into one shader effect instead the 2 it is now
[10:08] <mzanetti> vesar: I'm starting to look at the shader soon. its really about time for me to learn that stuff
[10:09] <vesar> mzanetti, damn I should learn that stuff myself as well. Well this is a good opportunity for you to get a grip on that!
[10:15] <mzanetti> tsdgeos: I can't reproduce the greeter with the unlock hint
[10:15] <mzanetti> tsdgeos: what exactly do you do?
[10:16] <mzanetti> tsdgeos: I mean the issue with the unlock hint
[10:16] <tsdgeos> mzanetti: on the phone, press with the finger close to the right edge not not "close enough"
[10:16] <tsdgeos> and don't release
[10:16] <tsdgeos> you'll see the hint
[10:16] <tsdgeos> now without releasing drag left
[10:16] <tsdgeos> doesn't unlock
[10:16] <mzanetti> tsdgeos: ah... I think thats intentional
[10:16] <tsdgeos> really?
[10:16] <tsdgeos> ok :D
[10:17]  * tsdgeos goes back to his lvwph issues
[10:17] <mzanetti> tsdgeos: mterry noticed the same with the left side
[10:17] <mzanetti> katie: hi
[10:20] <mzanetti> Saviq: just flashed the latest image... installed apps don't show icons
[10:24] <katie> mzanetti, hello
[10:25] <mzanetti> hi katie. I just had an idea regarding the edge hinting. but I'll bring it up in the meeting today. sorry for the noise
[10:25] <katie> mzanetti, I wanted to ask you about that.. has the right edge landed yet?
[10:25] <mzanetti> katie: yes
[10:26] <katie> mzanetti, I still haven't been able to see it
[10:26] <mzanetti> katie: since a while already. We just didn't manage to make a new release of unity8 in the past 2 weeks
[10:26] <mzanetti> katie: but if you flash your phone now you'll have it. also the new infographics colors
[10:26] <katie> mzanetti, great thanks
[10:36] <dednick> Saviq: ping
[10:36] <Saviq> dednick, otp
[10:39] <Saviq> dednick, what's up?
[10:39] <dednick> Saviq: hi. where does Unity.Notifications come from?
[10:39] <Saviq> mzanetti, yeah, reboot - we need to track that down yet, not sure where it happens
[10:39] <mzanetti> Saviq: ok.
[10:40] <Saviq> dednick, qtdeclarative5-unity-notifications-plugin
[10:40] <Saviq> dednick, it should be in distro soon
[10:40] <dednick> need to put that in build file?
[10:40] <Saviq> dednick, yeah
[10:40] <Saviq> dednick, I need to do cleanup
[10:40] <dednick> ok
[10:41] <Saviq> dednick, if it's not in distro yet - you can get the binary from https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4729845
[10:41] <Saviq> mhr3, yeah, unplug worked before
[10:42] <Saviq> mhr3, now it doesn't
[10:42] <dednick> Saviq: it seems to be there
[10:42] <Saviq> dednick, cool
[10:44] <Saviq> dednick, did you get anywhere with the indicator bugs?
[10:46] <dednick> Saviq: working on it now
[10:46] <Saviq> dednick, thanks
[10:47] <dednick> Saviq: haven't got anywhere yet though
[10:50] <tsdgeos> wooooo, unity survived a run of my test suite
[10:51] <tsdgeos> without crashing!
[10:52] <dednick> lol
[10:52] <dednick> success
[10:53] <dandrader> tsdgeos, that's why I'm using xubuntu
[10:53] <tsdgeos> dandrader: dist-upgrade
[10:53] <tsdgeos> seems last update gave some stability
[10:54] <sil2100> pete-woods: hi!
[10:55] <sil2100> pete-woods: do you have a moment?
[10:55] <pete-woods> sil2100: sure
[10:55] <pete-woods> is hud broken again? :p
[10:55] <ricotz> Trevinho, hi
[10:57] <ricotz> Trevinho, i have taken a short look at the last bamf commits and http://bazaar.launchpad.net/~unity-team/bamf/trunk/revision/557 doesnt preserve abi while adding "void (*icon_changed)                (BamfView *view, gchar* icon);" in between
[10:57] <sil2100> pete-woods: nooo ;)
[10:57] <sil2100> pete-woods: this time something different!
[10:57] <Saviq> MacSlow, about the notify-osd icons, I was thinking... shouldn't we really include them with the icon theme?
[10:58] <MacSlow> Saviq, that would be the easiest solution...
[10:58] <sil2100> pete-woods: we encountered a strange error with libusermetrics - during our daily-release job, it failed to build in our PPA because of a symbol that has been removed but not removed from the .symbols file
[10:58] <Saviq> MacSlow, I think it's the right solution, too, as we want for themes to be able to provide icons as well
[10:58] <sil2100> pete-woods: (i.e. take a look at https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4729129 )
[10:58] <MacSlow> Saviq, we didn't do that with the old notify-osd because of license issues
[10:59] <MacSlow> I think...
[10:59] <sil2100> pete-woods: the thing is, I prepared a merge, but it seems that that symbol was not removed in the end
[10:59] <Saviq> MacSlow, ah interesting, would those issues be valid still?
[10:59] <sil2100> pete-woods: https://code.launchpad.net/~sil2100/libusermetrics/update_symbols_file/+merge/170547
[10:59] <MacSlow> Saviq, but it's so long ago... I'm not 100% sure
[10:59] <sil2100> pete-woods: please notice the CI failure, which says that the .symbols file modification is wrong
[11:00] <sil2100> pete-woods: what I wanted to ask... in the end, that symbol is still in the library, yes?
[11:00] <Saviq> MacSlow, ok, let's say we expect them to be in the theme then
[11:00] <Saviq> MacSlow, or full paths, obviously
[11:01] <MacSlow> Saviq, in-theme and full paths are technicaly the easiest solution
[11:01] <Saviq> MacSlow, your work on the gicon provider isn't lost, I'd say, it'd still be nice to support it
[11:03] <Saviq> MacSlow, just that I don't think we should use it in the shell
[11:04] <Esokrates> smspillaz
[11:05] <MacSlow> the igcon-provider patch is something I still want to polish a bit (also due to your suggestions)
[11:06] <MacSlow> Saviq, but the snap-decision expansion is first... need to undust the old branch and make it work/apply with current trunk... and get the its test working with Chris timer-patch
[11:08] <arsson> just install unity updates for saucy and after that cant login unity desktop?
[11:10] <sil2100> pete-woods: ?
[11:11] <Saviq> MacSlow, sure
[11:11] <nic-doffay> Saviq, flashed the phone things are looking good. Glad to see the infographics colours landed too!
[11:11] <pete-woods> sil2100: sorry, just looking at it now
[11:11] <Saviq> nic-doffay, yup, the prevailing adjective for the new colors was "sexy" ;)
[11:12] <pete-woods> sil2100: I'm pretty sure I haven't changed the ABI
[11:12] <nic-doffay> Saviq, haha! That's a word I could get used to regarding work!
[11:13] <sil2100> pete-woods: it looks like it didn't change at least, but I wonder why our PPA would suddenly think otherwise - could something in the environment remove that method? Like, the lack of something installed?
[11:13] <sil2100> (during build)
[11:13] <pete-woods> sil2100: I supposed it's possible
[11:14] <MacSlow> nic-doffay, with saucy image did you put on the phone?
[11:14] <MacSlow> nic-doffay, 16 or newer?
[11:14] <Saviq> MacSlow, saucy-19
[11:14] <Saviq> MacSlow, 16? that's ages old! :D
[11:14] <MacSlow> Saviq, oh... I'm behind again :)
[11:14] <Saviq> MacSlow, saucy-19 has notifications and smart scopes
[11:14] <tvoss> greyback, ping
[11:14] <Saviq> +qml
[11:14] <MacSlow> sweet!
[11:14] <nic-doffay> MacSlow, the new colours?
[11:16] <greyback> tvoss: pong
[11:16] <MacSlow> nic-doffay, will reflash my phone again today
[11:16] <pete-woods> sil2100: I'm trying building the package under saucy (on my own machine) for the first time now - will see what it says
[11:17] <sil2100> pete-woods: thanks!
[11:17] <mzanetti> vesar: \o/
[11:18] <mzanetti> vesar: written my own BrightnessEffect and its way faster than using the plain Qt's BrightnessContrast effect
[11:18] <mzanetti> vesar: though its cheeting. but looks the same so I guess still good :)
[11:21] <nic-doffay> mzanetti, how did you do it?
[11:25] <tsdgeos> doubles suck
[11:25] <tsdgeos>    Actual   (lvwph->m_minYExtent): -2,84217e-14
[11:25] <tsdgeos>    Expected (0.): 0
[11:26] <tsdgeos> come on!
[11:27] <tvoss> tsdgeos, why is the number negative, though?
[11:28] <mzanetti> nic-doffay: a black ubuntu shape on top that changes the opacity
[11:28] <tsdgeos> tvoss: don't know to be honest
[11:28] <tsdgeos> if jsut happens if i test what i don't to test anyway
[11:29] <tsdgeos> so it's ok :D
[11:29] <tsdgeos> was a random complaint
[11:29]  * tvoss says lalala
[11:31] <Saviq> seb128, is there a mk-build-deps for runtime deps? i.e. if I want to install runtime deps based on debian/control?
[11:32] <pete-woods> sil2100: so on my freshly installed saucy build it seems to pass
[11:36] <sil2100> pete-woods: it might have something to do with the toolchain...
[11:36] <sil2100> pete-woods: the only usage of 'release' is in the std::unique_ptr usage, no where else
[11:36] <sil2100> What is this symbol anyway?
[11:37] <pete-woods> sil2100: that's a good question, actually
[11:38] <seb128> Saviq, not that I know about...
[11:38] <Saviq> seb128, thanks
[11:39] <seb128> Saviq, rename Depends in Build-Depends and use ml-build-deps? :p
[11:39] <Saviq> seb128, ;D
[11:39] <seb128> Saviq, the issue is that most Depends are generated at build time, like shlib:Depends is make from looking at the linker
[11:39] <Saviq> seb128, right
[11:42] <sil2100> pete-woods: that's why actually it's not recommended to use .symbols files for C++ libraries ;)
[11:42] <sil2100> pete-woods: not only there is name mangling, there's always that big 'WTF is this symbol'!
[11:43] <pete-woods> sil2100: fair enough, I just didn't want my ABI to break
[11:43] <seb128> pete-woods, talk to xnox about symbols and cpp
[11:43] <xnox> pete-woods: hola =) how are you?
[11:44] <xnox> sil2100: there is a way to c++filt symbols and make them look very nice actually.
[11:44]  * xnox looks up an example.
[11:44] <pete-woods> xnox: good apart from strange symbols I don't know why I have in my ABI
[11:44] <xnox> don't export them =)
[11:44] <seb128> pete-woods, https://wiki.ubuntu.com/DailyRelease/FAQ#I.27m_exposing_a_new_C.2BAC8-C.2B-.2B-_symbols_in_my_library.2C_it_seems_that_some_packaging_changes_are_needed.2BICY-
[11:45] <sil2100> \o/
[11:45] <xnox> pete-woods: sil2100: albeit didrocks still did not merge my updated symbols file for phatform-api. But if you look at the merge proposal: https://code.launchpad.net/~xnox/platform-api/hybris-symbols/+merge/158878
[11:45] <sil2100> seb128: ooh! There was no info like this before in this section!
[11:45] <xnox> you will notice that they are all very sensible
[11:46] <seb128> sil2100, xnox added them
[11:46] <sil2100> seb128: someone had to add it recently !
[11:46] <xnox> and that's c
[11:46] <xnox> let me look up C++
[11:46] <xnox> pete-woods: sil2100: C++ example of sensible symbols https://code.launchpad.net/~xnox/qtubuntu-sensors/symbols/+merge/158884
[11:46] <sil2100> Woha, c++filt feels awesome
[11:46] <pete-woods> xnox : it still seems like the problem here is me exporting the wrong symbols, though
[11:47] <xnox> pete-woods: hmm... where abouts? Can I take a look?
[11:47] <sil2100> pete-woods: thanks for the examples! Looks much better
[11:47] <sil2100> I mean, xnox
[11:47] <pete-woods> :)
[11:47] <sil2100> xnox: https://code.launchpad.net/libusermetrics
[11:48] <sil2100> xnox: here, you can see the symbols file, which seems to be pre-generated, and it has some strange symbols as well, probably from std:: ?
[11:48] <xnox> sil2100: it _looks_ better, but doesn't catch everything. (as some compiler/arch specific information is removed) hence something like abi-compiance-checker is still needed.
[11:48] <sil2100> xnox: http://bazaar.launchpad.net/~indicator-applet-developers/libusermetrics/trunk/revision/75
[11:49] <sil2100> xnox: we've been trying to figure out the source of the "_ZNSt16_Sp_counted_baseILN9__gnu_cxx12_Lock_policyE2EE10_M_releaseEv@Base 1.0.1" symbol
[11:49] <sil2100> Looks like std::unique_ptr
[11:49] <xnox> sil2100: so I run that diff through the pipe of | c++filt and it gives me this: http://paste.ubuntu.com/5783396/
[11:49] <xnox> i9ndeed std:: leaked.
[11:50] <sil2100> pete-woods: ^
[11:50] <xnox> c++filt    is awesome =)
[11:50] <sil2100> c++filt is one thing I need to write down NOW
[11:50] <pete-woods> sil2100: yeah, that's what it looks like, and yes!! how awesome is c++filt
[11:50] <xnox> surprisingly the cryptic c++ symbol naming is  deterministic translation ;-)
[11:52] <didrocks> xnox: yeah, it dropped low on my list with all the transitions, I needed to make some research again on c++filt because IIRC, our latest attempt failed
[11:52] <didrocks> xnox: but as CI seems to be happy, we can give it a try
[11:53] <xnox> didrocks: you might want to retrigger CI again, since it has been a while....
[11:53] <didrocks> xnox: yeah, let's try that :)
[11:53] <didrocks> xnox: and the merge
[11:53] <didrocks> that will enable me to look at the symbols updates if any
[11:55] <pete-woods> so this is almost certainly one of the offending files: http://pastebin.ubuntu.com/5783416/
[11:56] <pete-woods> how do I stop if from exporting the shared_ptr stuff?
[11:59] <didrocks> xnox: platform-api is exposing a C and C++ API, right? That's why we have mangled and unmangled generated symbols?
[11:59] <didrocks> like:
[11:59] <didrocks> + _ZN6ubuntu6hybris6BridgeD2Ev@Base 0.18.1daily13.06.18-0ubuntu1
[12:00] <didrocks> and + u_application_description_destroy@Base 0.18.1daily13.06.18-0ubuntu1
[12:00] <xnox> think so.
[12:00] <xnox> but will double check in a minute.
[12:00] <xnox> pete-woods: but the typedef MetictPtr is part of the API, thus someone else can use UserMetricsInput::MetricPtr type? no?
[12:01] <dandrader> MacSlow|lunch, "Shell.qml:31:1: module 'Unity.Notifications' is not installed". Where can I get it. I just run ./build --setup and ./build --clean
[12:01] <pete-woods> xnox: that's true, yes
[12:01] <pete-woods> didrocks: yes, the library has both C++ and C symbols
[12:01] <MacSlow|lunch> dandrader, that's the backend your missing... lp:unity-notifications
[12:01] <didrocks> pete-woods: thanks ;)
[12:02] <MacSlow|lunch> dandrader, that's happening locally on you development-machine?
[12:02] <dandrader> MacSlow|lunch,  yes
[12:02] <dandrader> MacSlow|lunch, shouldn't the build script have sorted that out?
[12:02] <didrocks> xnox: I just wonder, so dh_makeshlibs is unmangling them to make the diff, right? why is that not part of it?
[12:02] <MacSlow|lunch> dandrader, get lp:unity-notifications compile and install
[12:02] <MacSlow|lunch> dandrader, but there should be a ppa for this by now... ehm...
[12:03] <pete-woods> xnox: I have no problem with it exporting those symbols, as long as it does it consistently
[12:03] <xnox> didrocks: good point. i wish we could ask it to spit out demangled diffs via an option or something like that.
[12:03] <MacSlow|lunch> dandrader, https://launchpad.net/ubuntu/saucy/+source/unity-notifications
[12:04] <xnox> pete-woods: right, I think any changes in api/abi of the std:: will leak into your exported api/abi, thus you're api/abi is a bit screwed i think.
[12:04] <didrocks> xnox: yeah, as it seems the magic is just in :) Let me first refresh the FAQ with more steps by steps process (refreshing with the platform-api and will hand you over a branch to merge in yours)
[12:04] <pete-woods> xnox: surely if the C++ ABI has changed, all bets are off anyway?
[12:04] <pete-woods> and it's time to recompile
[12:05] <dandrader> MacSlow|lunch, is that the package name? qtdeclarative5-unity-notifications-plugin
[12:05] <xnox> pete-woods: well, is std::shard_ptr part of c++11 api/abi which is not yet stable by gcc? thus it will change until 4.9.0 is released.
[12:06] <xnox> pete-woods: the other option is to mark that symbol (optional) thus it may or may not be present depending on which gcc is used.
[12:06] <xnox> pete-woods: and indeed if C++ ABI changes most bets are off =) (well anything that uses the bits that changed backwards incompatible)
[12:06] <dandrader> MacSlow|lunch, so my apt-cache shows it. I think it's just a matter of having the build script installing it as well
[12:06] <MacSlow|lunch> dandrader, yes... sudo apt-get install qt... and so on
[12:07] <MacSlow|lunch> dandrader, I'll look into that
[12:08] <MacSlow|lunch> dandrader, I'll provide a patch for that
[12:09]  * greyback moving to workspace, back in 30
[12:12] <dandrader> MacSlow|lunch, thanks. Please also take the opportunity to break that insanely long line. With a patch like this one: http://paste.ubuntu.com/5783454/
[12:15] <dandrader> and sorting them alphabetically would also be nice...
[12:20] <didrocks> sil2100: xnox: ok, updated the FAQ: https://wiki.ubuntu.com/DailyRelease/FAQ#I.27m_exposing_a_new_C.2BAC8-C.2B-.2B-_symbols_in_my_library.2C_it_seems_that_some_packaging_changes_are_needed.2BICY-, tell me if the commands make sense
[12:22] <xnox> didrocks: yes, much better. I don't think I understood the special meaning of the 0replaceme
[12:23] <pete-woods> sil2100: I have proposed this MR: https://code.launchpad.net/~pete-woods/libusermetrics/update-symbols-file/+merge/170596 - using the optional tag to handle different versions of GCC / STL that we are building upon
[12:23] <pete-woods> sil2100: will update to that nice symbols format using c++filt in a separate MR
[12:23] <xnox> dandrader: if that package is in the archive already, surely one can simply do $ sudo apt-get build-dep mypackage
[12:23] <didrocks> xnox: see the section before the one you added. It's just to update to the version which is uploaded when daily releasing
[12:23] <xnox> which will apt-get install all of the build-dependencies.
[12:24] <xnox> didrocks: hehe =)
[12:24] <MacSlow|lunch> dandrader, btw.. you missed one line break ("libdee-qt5-dev libpulse-dev \") https://code.launchpad.net/~macslow/unity/build-script-update-for-notifications/+merge/170597
[12:26] <dandrader> MacSlow|lunch, right
[12:27] <didrocks> xnox: do you mind pulling ~didrocks/platform-api/refresh-symbols in your branch? I'll merge it then
[12:28] <didrocks> xnox: no need for a MP, as I resynced to trunk and so the diff you will see against your branch will be horrible, just look at the last commit :)
[12:31] <didrocks> xnox: your qtubuntu-sensors are still up to date! approving that one meanwhile :)
[12:32] <didrocks> xnox: ah no, you didn't change 0.5.1 by 0replaceme ;)
[12:33] <didrocks> (should be the same for the initial platform-api I think, all with 0replaceme)
[12:34] <Saviq> MacSlow, uh, sorry, I cleaned up the build scripts in https://code.launchpad.net/~saviq/unity/8.clean-scripts
[12:35] <MacSlow> Saviq, nah... no harm done
[12:35] <Saviq> MacSlow, will MP in a sec, just doing what you did for the installs
[12:35] <MacSlow> Saviq, I can mark my branch as obsolete
[12:36] <Saviq> MacSlow, or just resubmit with mine as the source
[12:43] <MacSlow> Saviq, looking at it, I don't think my branch would add any value to 8.clean-scripts
[12:45] <Saviq> MacSlow, ah, you deleted?
[12:45] <xnox> didrocks: yeap, it all should be 0replaceme.
[12:46] <didrocks> xnox: want me to change those (in both?)
[12:46] <xnox> didrocks: please do =) i don't think i have those branches around locally any more.
[12:48] <didrocks> xnox: hum, should I then MP them directly against trunk and you approve? :)
[12:48] <didrocks> (if you can't pull my branches to yours)
[12:48] <xnox> didrocks: that works as well.
[12:54] <didrocks> xnox: https://code.launchpad.net/~didrocks/qtubuntu-sensors/symbols/+merge/170603 and https://code.launchpad.net/~didrocks/platform-api/refresh-symbols/+merge/170601
[12:54] <MacSlow> Saviq, yup
[12:55] <xnox> didrocks: lgtm
[12:56] <didrocks> xnox: thanks a lot! I globally approve then! And thanks to have found a great solution for C++ symbols, multiple people (I included) stroke to find a scalable and reliable solution in the past years :)
[12:56] <didrocks> seems there is one now!
[12:57] <dandrader> Saviq, does your 8.clean-scripts add the missing qtdeclarative5-unity-notifications-plugin?
[12:59] <Saviq> dandrader, yes
[12:59] <Saviq> dandrader, MacSlow removed his MP
[12:59] <dandrader> yeah, I noticed
[13:00] <Saviq> dandrader, it's using a virtual dep of "unity-notifications-impl-1"
[13:00] <Saviq> dandrader, that the plugin has in its Provides:
[13:00] <MacSlow> Saviq, dandrader: but... yeah... isn't that sufficient
[13:00] <sil2100> pete-woods: once you have the nice symbols MR, give me a sign
[13:00] <Saviq> MacSlow, yeah, everything's fine
[13:00] <dandrader> Saviq, ah, that explains why I didn't find qtdeclarative5-unity-notifications-plugin there
[13:01] <Saviq> dandrader, yeah, it's a scheme we came up with to be able to replace each of the backends if we need/want to
[13:01] <Saviq> dandrader, as long as it implements the interface that's (to be) defined in lp:unity-api
[13:02] <Saviq> dandrader, it should work, so it has a Provides: unity-notifications-impl, unity-notifications-impl-1
[13:02] <Saviq> where 1 is the API version
[13:02] <Saviq> and the shell depends on *an* implementation of that API at a certain version
[13:03] <dandrader> Saviq, what command do I use to check a purely virtual pacakge? apt-cache show doesn't do it
[13:03] <Saviq> dandrader, apt-get install will select it
[13:04] <Saviq> dandrader, and apt-cache search, too
[13:05] <dandrader> Saviq, I mean, checking what packages unity-notifications-impl-1 brings in without having to install it. But just figured out that it's just a matter of apt-geti install it and then replying "no" to the install doesn't take place
[13:06] <Saviq> dandrader, `apt-cache search` came up with a list for me, too
[13:07] <dandrader> hmm, interesting
[13:20] <Saviq> seb128, do you know what runs optipng on the images during sbuild/PPA builds/distro builds?
[13:21] <Saviq> seb128, and if there is a way to preprocess them and add to a whitelist?
[13:21] <seb128> Saviq, pkgbinarymangler
[13:21] <seb128> I think
[13:21] <seb128> let me check
[13:21] <Saviq> it takes like 40 minutes to optipng on a panda
[13:21] <Saviq> where the build itself takes ~20...
[13:21] <Saviq> (for unity8, that is)
[13:22] <seb128> Saviq, you can set NO_PNG_PKG_MANGLE to skip it apparently
[13:22] <seb128> https://launchpad.net/ubuntu/+source/pkgbinarymangler/89
[13:22] <Saviq> seb128, yay!
[13:22] <Saviq> seb128, hmm I won't be able to set that on LP builders, though, will I ;)
[13:23] <seb128> Saviq, you should be able to export it in debian/rules
[13:23] <Saviq> seb128, mhm, checking
[13:24] <seb128> Saviq, just add a export NO_PNG_PKG_MANGLE=1 in the rules
[13:24] <seb128> before the actual build rules
[13:24] <Saviq> seb128, yup, thanks
[13:34] <Saviq> nic-doffay, dednick Cimi standup
[13:34] <Cimi> Saviq, coming
[13:46] <dobey> Saviq: hrmm, i have the desktop-deps ppa, and just did an apt-get update, on raring, but that notifications-plugin package isn't found
[13:46] <Saviq> dobey, we're working on saucy for a long time now
[13:46] <Saviq> dobey, raring is no longer supported
[13:47] <Saviq> (for unity8 development, that is)
[13:47] <Saviq> dobey, we've just switched to saucy's smart scopes, so that's simply where you need to be
[13:49] <Saviq> dobey, sorry if that wasn't clear before
[13:49] <Saviq> dobey, where would you have expected to find such info?
[13:51] <dobey> Saviq: well, when i was asking for info/help about all the problems i've been having with getting it running at all over the past 3-4 days, it would have been nice if someone had said "oh, btw, you must do this on saucy"
[13:52] <dednick> Saviq: fix for https://bugs.launchpad.net/touch-preview-images/+bug/1191822 is in indicators-client. Should i just put it into the new indicators branch?
[13:52] <Saviq> dednick, yeah, link the bug to your branch
[13:53] <dobey> aside from that, i don't know where one would "expect" to find such info. or where i'm supposed to look for it.
[13:53] <Saviq> dobey, it's a bad assumption from our side indeed
[13:53] <Saviq> dobey, CODING mentions that only 13.10 is supported
[13:53] <Saviq> dobey, I'll add a check in ./build
[13:54] <Saviq> dednick, and change the bug to "In Progress"
[13:54] <Saviq> dobey, as mhr3 said yesterday, it was a very bad moment to get involved, sorry we didn't make it easier for you
[13:56] <tsdgeos> dandrader: typo at https://code.launchpad.net/~dandrader/unity/8_bottomBarDDA/+merge/170296
[13:56] <tsdgeos> otherwise looks cool :-)
[13:56] <Saviq> dobey, with any further issues you have, just ping me directly, ok?
[13:57] <mfisch> dpm: hey, I just woke up but it looks like the doc is mostly done, do you need a review?
[13:57] <dobey> Saviq: well, originally i was told to use lp:unity/phablet, which was apparently just flat out wrong; and the way ./build and ./build_unity and all that stuff works is just awful. :-/
[13:57] <Saviq> dobey, not anymore
[13:57] <dpm> mfisch, that'd be awesome, yes!
[13:57] <Saviq> dobey, we're building against distro since yesterday
[13:58] <Saviq> dobey, so build_unity is gone, build is simplified a lot
[13:58] <dpm> mfisch, we still don't have the scope template in Qt Creator, so I might have to adjust the initial setup section a bit before publishing it.
[13:58] <Saviq> dobey, we just had to maintain a delta between distro and our backends until yesterday
[13:59] <Saviq> dobey, which was obviously painful, and we tried to help with that with the scripts
[13:59] <Saviq> obviously sometimes we failed
[14:02] <seb128> mardy, around?
[14:06] <Saviq> mterry, hi, you might've noticed I assigned you a simple bug with the greeter - it animates when unlocking after a timeout
[14:06] <Saviq> mterry, ultimately it should listen to powerd for the suspend signal on DBus
[14:06] <Saviq> mterry, but for now it should be enough if the behaviour is only enabled when it's on screen
[14:07] <mterry> Saviq, yup I saw it go by.  I'm busy with this mir stuff right now, but it's in my TODO
[14:07] <Saviq> mterry, thanks
[14:07] <dandrader> tsdgeos, fixed. thanks for the review
[14:13] <tsdgeos> dandrader: how dos the dda work? you don't need to define the width? it just takes all the screen? or?
[14:14] <dandrader> tsdgeos, works essentially like the DraggingArea, but with gesture recognition and touch events instead of mouse ones
[14:14] <dandrader> tsdgeos, the main thing in this patch
[14:14] <dandrader> tsdgeos, is that BottomBar doesn't really use the features Revealer provides
[14:14] <tsdgeos> dandrader: true
[14:14] <dandrader> on top of a plain DraggingArea or DirectionalDragArea
[14:15] <tsdgeos> oh, i was asking more about the 12	- width: shell.width
[14:15] <tsdgeos> 13	- height: shell.height
[14:15] <tsdgeos> you removed
[14:15] <tsdgeos> but i see the Shell already fills those
[14:15] <dandrader> tsdgeos, yes
[14:16] <dandrader> tsdgeos, removed use of "shell" context var from within BottomBar
[14:16] <dandrader> for the sake of modularity
[14:17] <tsdgeos> good stuff
[14:17] <tsdgeos> that code is frankly not the best
[14:17] <tsdgeos> result of the mutation the hud lived in how its invokation works
[14:27] <mhall119> Saviq: I upgraded to saucy-18 phablet build last night and got a regression: https://bugs.launchpad.net/unity/+bug/1192816
[14:27] <mhall119> Saviq: is this something we can put an autopilot test on so it doesn't happen again?
[14:30] <Saviq> mhall119, saucy-18 was all kinds of broken, but this bug is on our radar, will be tested from both the shell side and the app management side
[14:30] <Saviq> mhall119, atm we can't do autopilot for it 'cause it's only happened on the device
[14:30] <mhall119> it's not that bad, it's about as usable as any other release for me
[14:31] <mhall119> ah, ok
[14:31] <Saviq> mhall119, no scopes, though, other than Home?
[14:31] <Saviq> mhall119, saucy-19 followed with the needed scopes
[14:31] <mhall119> well there was that, yeah, but easily fixed by installing them after upgrade
[14:31] <Saviq> mhall119, yeah
[14:31] <Saviq> mhall119, but we're working on running autopilot tests on the device atm
[14:31] <mhall119> cool
[14:31] <Saviq> mhall119, and yeah, it's a perfectly valid test case when that happens
[14:32] <tsdgeos> Saviq: https://code.launchpad.net/~saviq/unity/8.optimize-pngs/+merge/170619 is just a optipng run?
[14:33] <Saviq> tsdgeos, and the debian/rules addition
[14:33] <Saviq> tsdgeos, so that it isn't run on every PPA / distro build
[14:33] <Saviq> tsdgeos, it took 40 mins on a pandaboard yesterday...
[14:33] <Saviq> tsdgeos, after maybe 20 mins of actual build time...
[14:34] <olli> Saviq, with the smartscopes in the build now, what changes are user visible?
[14:34] <olli> Saviq, and just to be clear, which build do I want?
[14:34] <tsdgeos> Saviq: sure sure, but i mean it's "just" optipng, right?
[14:34] <tsdgeos> Saviq: and from now on we have to remember to run optipng manually
[14:34] <Saviq> olli, saucy-19
[14:35] <Saviq> olli, and the only visibile change is the lack of People lens (which was the target, really)
[14:35] <Saviq> tsdgeos, yeah
[14:35] <tsdgeos> Saviq: oki, approving
[14:35] <olli> Saviq, k, thx
[14:35]  * olli flashes
[14:35] <Saviq> olli, if you install additional master scopes (social, for example) - they'll show up
[14:35] <tsdgeos> Saviq: or want me to wait for ricardo?
[14:35] <Saviq> olli, and we're working on enabling the full smart scope experience
[14:35] <Saviq> tsdgeos, no, it was mostly an FYI for Ricardo
[14:36] <tsdgeos> ok
[14:36] <Saviq> tsdgeos, 'cause we both were waiting for it yesterday
[14:36] <olli> Saviq, what does that mean?
[14:36] <tsdgeos> hehe
[14:36] <Saviq> olli, reaching the smart scope server when searching in Home, for example
[14:36] <Saviq> the most apparent change that will come from that
[14:37] <Saviq> and generally showing results from smart scopes in Home, which isn't happening yet
[14:37] <olli> ok
[14:37] <olli> Saviq, thx
[14:37] <olli> good job getting all that stuff in team!
[14:38] <mfisch> dpm: doc looks good, I made a few small additions
[14:41] <dpm> mfisch, perfect, thanks!
[14:45] <dednick> *sigh* now i got to go find a simcard so i can send myself messages :(
[14:51] <tsdgeos> Saviq: also from the switch to this smart scopes thing, now we open in music instead of home but the dashbar thinks we are in home
[14:51] <tsdgeos> at least on the pc
[14:51] <tsdgeos> seems phone is different
[14:51] <tsdgeos> no clue why
[14:58] <Saviq> tsdgeos, yeah, dashbar is broken with many scopes
[15:09] <mzanetti> mterry: meeting?
[15:13] <tsdgeos> any clue of what failed here? https://jenkins.qa.ubuntu.com/job/unity-8.0-saucy-armhf-ci/160/console
[15:13] <tsdgeos> cmake crashed?
[15:17] <Saviq> tsdgeos, something like this
[15:17] <Saviq> tsdgeos, can't find cmake's exit codes
[15:17] <mfisch> dpm: do you need any more assistance from me on this?
[15:17] <tsdgeos> Saviq: i'd say signal 11 means sigsev
[15:18] <tsdgeos> not cmake exited with value 11
[15:18] <Saviq> tsdgeos, right
[15:18] <tsdgeos> scared when cmake segfaults :D
[15:18] <tsdgeos> anyway restarted the job
[15:18] <tsdgeos> Saviq: anyone looking at the build/run script fixes?
[15:18] <mterry> mzanetti, oh shoot
[15:18] <tsdgeos> want me to?
[15:18] <mterry> mzanetti, I didn't set a timer
[15:20] <dpm> mfish, that was all, thanks. Thanks a lot for writing the code, that was awesome
[15:21] <mfisch> dpm: no problem
[15:25] <mzanetti> Saviq: re  [mterry] Make sure the greeter doesn't animate when locking: TODO
[15:25] <mzanetti> Saviq: I kinda think it should animate, but not only on wakeup
[15:26] <Saviq> mzanetti, well, that's a short term solution
[15:26] <Saviq> mzanetti, but since locking == blank screen usually
[15:26] <Saviq> mzanetti, at least for phone it will be
[15:27] <Saviq> mzanetti, then it doesn't have to animate (I mean the fact that it gets in from the side to cover the screen, not necessarily infographics)
[15:27] <mzanetti> Saviq: not forever I'd say
[15:27] <mzanetti> Saviq: so far for all phones I had quite fast there were apps/plugins/hacks whatever that can keep the display lit
[15:27] <mzanetti> anyways... just thinking loud
[15:28] <Saviq> mzanetti, sure, hence Make the greeter listen to the DBus signal from powerd for locking instead of the power key: TODO
[15:28] <mzanetti> ah
[15:28] <Saviq> mzanetti, but if display lit, also not locked, no?
[15:28] <mzanetti> Saviq: hmm... I think I can lock my N9 without turning off the screen
[15:29] <mzanetti> Saviq: yeah... been using that quite often when I need a clock for giving a talk or such things
[15:30] <Saviq> mzanetti, N9 is different
[15:30] <Saviq> mzanetti, the clock _is_ display off
[15:30] <Saviq> effectively
[15:30] <mzanetti> Saviq: talking about my N950 actually
[15:30] <Saviq> mzanetti, maybe that's different, again, it doesn't have OLED, right?
[15:31] <Saviq> mzanetti, N9, when you lock, goes to a black screen + clock
[15:31] <mzanetti> just always referring as N9 to it. that one doesn't have the awesome clock with screen off feature
[15:31] <Saviq> when you then cover the proximity sensor
[15:31] <Saviq> that's when it shuts completely off
[15:31] <mzanetti> yeah... the N950 can't do that
[15:31] <Saviq> mzanetti, and anyway - sure, suspend != lock
[15:31] <Saviq> mzanetti, not always, but for our purpose, right now, it does
[15:32] <mzanetti> though Nokia did some freaky stuff to show the clock also on non-oled display
[15:32] <mzanetti> I think they showed it 1/4 frame rate and only some interlacing and all that kind of stuff
[15:32] <mzanetti> on the N8 for example
[15:32] <Saviq> mhm
[15:35] <Cimi> mzanetti,  http://paste.ubuntu.com/5784026/
[15:36] <Cimi> mmm no
[15:36] <Cimi> there's an error :)
[15:36] <Saviq> Cimi, so, that means you have unlimited months? :)
[15:37] <Cimi> Saviq, what? :D
[15:37] <Saviq> Cimi, that you're not limited to 12 months back and forward?
[15:37] <Cimi> http://paste.ubuntu.com/5784036/
[15:37] <Cimi> Saviq, it wasn't even before
[15:37] <Cimi> this one is correct, forgot to use the variable of the model
[15:37] <Saviq> Cimi, it was, you could reach the beginning of the calendar ;)
[15:38] <Cimi> *rule of the model
[15:38] <Saviq> and the end, too!
[15:38] <Cimi> Saviq, yes, even now
[15:38] <Cimi> Saviq, we don't want unlimited
[15:38] <Cimi> there's minimum and maximum date
[15:38] <mzanetti> pff... are we mayas?
[15:38] <Cimi> Saviq, maybe I can set year 0
[15:38] <Cimi> mzanetti, http://paste.ubuntu.com/5784036/
[15:39] <Cimi> mzanetti, tests pass... so I believe it's correct xD
[15:39] <mzanetti> Cimi: yep, I've seen... didn't really review the code right now. but the general idea is good, yes
[15:39] <Saviq> Cimi, even so, it shouldn't be a limitation of the component, IMO
[15:39] <Saviq> Cimi, a property - yes, not a limitation
[15:40] <Cimi> Saviq, just let's set the default to a huge value
[15:40] <Saviq> Cimi, no, that's not a solution
[15:40] <Saviq> Cimi, 'cause that will mean the ListModel is huge
[15:40] <Saviq> Cimi, when you only ever need three items - current, previous and next
[15:40] <Cimi> Saviq, how do I create an infinite listmodel then ? :D
[15:40] <Saviq> Cimi, you don't use a listmodel ;)
[15:41] <Saviq> Cimi, but even in a listmodel, you remove and add items as needed
[15:41] <Cimi> mmmm
[15:41] <mzanetti> Saviq: interesting... why not, or what else?
[15:41] <Cimi> yes it's interesting
[15:42] <Saviq> mzanetti, a page stack that you push / pop pages from the back and end
[15:42] <Saviq> beginning and end
[15:42] <Saviq> mzanetti, I'm not saying that's currently supported
[15:43] <mzanetti> Saviq: a page stack? would never have thought of that
[15:43] <Saviq> mzanetti, but when you only need 3 items at any given time (you can only swipe one at a time)
[15:43] <Saviq> mzanetti, you only need the current, the next and the previous ones
[15:43] <mzanetti> Saviq: yeah... not saying he should keep all the delegates all the time
[15:43] <Saviq> mzanetti, and then you can skip to an arbitrary one (obviously without the intermediate ones)
[15:44] <Saviq> mzanetti, delegates, no delegates, the ListModel shouldn't have the whole set all the time
[15:44] <Saviq> mzanetti, it's not a ListView, really
[15:44] <Saviq> mzanetti, it's just a page that you replace with another one
[15:44] <Saviq> and IMO it shouldn't be limited by default
[15:44] <Saviq> mzanetti, obviously the thing that's slightly tricky here is the header
[15:45] <mzanetti> Saviq: right... I understand what you mean
[15:45] <Saviq> mzanetti, which is a list view
[15:45] <mzanetti> not using any Model/View at all
[15:45] <Saviq> but it should be enough to keep a buffer of 10 items at the head and at the tail
[15:45] <Saviq> and then add / remove as needed
[15:46] <mzanetti> Cimi: after this I tend to agree with saviq. Sorry I didn't think of this earlier
[15:46] <mzanetti> dammit...
[15:46] <Cimi> xD
[15:46]  * mzanetti feels bad for wasting Cimis time
[15:46]  * Saviq feels good for not being useless
[15:46]  * Cimi feels completely useless xD
[15:47] <Saviq> Cimi, mzanetti look / think at the desktop - you can jump in years, not only in months
[15:47] <Saviq> a Model/View approach will never scale to that
[15:47] <Saviq> (I mean the desktop indicator for date/time)
[15:48] <mzanetti> Saviq: I'm fully convinced by now
[15:48] <mzanetti> Saviq: well, I guess the listview could keep up if done right... but thats not the point
[15:48] <Saviq> mzanetti, yeah, it's abusing the ListView anyway
[15:48] <Saviq> mzanetti, for no reason
[15:48] <Saviq> mzanetti, as I said, it might require abusing for the tabs
[15:49] <Saviq> mzanetti, but that's smaller, but still only *some* items should be in the model at any given time
[15:49]  * Saviq just flashed manta to maguro
[15:49]  * Saviq feels useless again
[15:49] <mzanetti> lol... upgrade your device
[15:49] <mzanetti> hardware wise
[15:49] <mzanetti> anyways... I'm off for food for a while
[15:50] <Saviq> o/
[16:00] <Cimi> Saviq, I need to scroll though, so I create a flickable?
[16:01] <Cimi> Saviq, doing the same with a model and a listview no? (removing or automatically adding months)
[16:01] <Saviq> Cimi, I wasn't saying I have all the answers, but talk to timp whether this kind of approach (flicking between pages) is possible with a PageStack in the SDK
[16:02] <Saviq> Cimi, and/or whether they've plans for something like this
[16:02] <Saviq> Cimi, a ListView to/from which model's you add/remove items is a potential way
[16:02] <Saviq> Cimi, but you need to see if it actually works fine when you remove and add items
[16:03] <Saviq> rsalveti, maguro adb fixed, thanks!
[16:05] <pete-woods> sil2100: https://code.launchpad.net/~pete-woods/libusermetrics/nice-symbols-format/+merge/170656 - here's the MR for making the symbols files "nice"
[16:05] <Saviq> rsalveti, and no, no CPU hogging here
[16:05] <olli> Saviq, ping
[16:05] <Saviq> olli, pong
[16:05] <sil2100> pete-woods: checking! Thanks!
[16:06] <olli> Saviq, what's the status of Mir/U8 landing in the image
[16:06] <olli> are we going with a separate image for a while as I offered in my mail
[16:06] <olli> or is it going straight in?
[16:06] <Saviq> olli, I thought that wasn't a question :)
[16:06] <olli> asac just mentioned that chicken thought it was already there
[16:06] <Saviq> olli, I'd be glad to have the transition period
[16:07] <olli> Saviq, ok
[16:07] <olli> and there is no Mir anywhere near the saucy image atm
[16:07] <Saviq> no
[16:07] <olli> k
[16:08]  * greyback rebooting into saucy, hoping his wifi doesn't behave as badly as in last saucy install
[16:11] <sil2100> pete-woods: hmmm
[16:11] <pete-woods> sil2100: I noticed we have lots of duplicate lines now..
[16:12] <pete-woods> sil2100: which seems a little strange to me..
[16:12] <sil2100> pete-woods: I commented on the bug, as I tried test building a package, strange things!
[16:13] <mterry> racarr, you around?  I've got some backtraces from that mir crash
[16:13] <pete-woods> sil2100: clearly I'm not using that optional tag correctly
[16:17] <pete-woods> sil2100: trying something else (actually read the manpage this time)
[16:24] <pete-woods> sil2100: I've pushed a change to that branch, by the way - I'm glad you're able to reproduce it locally
[16:24] <pete-woods> sil2100: would you be able to send me a diff of the package changes you made, just so I can rule those out here?
[16:25] <sil2100> Ok ;) One moment
[16:26] <dednick_> larsu: i've just assigned https://bugs.launchpad.net/indicator-messages/+bug/1191820 to you. crirical bug which seems to be in indicator-messages
[16:26] <larsu> dednick_: thanks, I'll have a look
[16:26] <dednick_> larsu: i think new messages are crashing the indicator process.
[16:27] <sil2100> pete-woods: http://paste.ubuntu.com/5784187/ <- the changelog mod is not necessary to include into trunk, as it's only needed when you want to build it locally (normally the daily-release bot will add it himself)
[16:27] <sil2100> pete-woods: it's needed when you want to build locally as otherwise it will try and use the already-released libusermetrics tarball, which is not that we want
[16:27] <larsu> dednick_: do you have a stack trace? I don't have a phone with an up-to-date image right now
[16:28] <sil2100> pete-woods: we need a new tarball generated
[16:30] <pete-woods> sil2100: yes, that is something that is not in my current skillset - how are you supposed to do it?
[16:31] <pete-woods> sil2100: those are the same changes that I have made, btw, so at least we're on the same page here
[16:33] <sil2100> That's all that's needed - we're using .bzr-builddeb's split mode, so whenever he sees an UNRELEASED new upstream version, it tries to create a tarball from the source then
[16:33] <sil2100> So all is fine
[16:42] <olli> Saviq, asac says the image is already in place
[16:42] <olli> are you aware of that?
[16:42] <asac> ok it again was bad notes. it will land today
[16:44] <sil2100> bschaefer: ping!
[16:44] <bschaefer> sil2100, pong
[16:44] <sil2100> bschaefer: hi! How's unity with new X going?
[16:45] <bschaefer> sil2100, you mean that edge barriers branch?
[16:45] <bschaefer> sil2100, theres a ppa if wish to try :)
[16:45] <bschaefer> sil2100, that has the new 1.14 server and my unity branch: https://launchpad.net/~canonical-x/+archive/x-staging
[16:45] <sil2100> bschaefer: awesome - could you show it? When's the ETA for that?
[16:45] <sil2100> Thanks :)
[16:46] <bschaefer> sil2100, its done as far as I know, getting more people to manually test is about all :)
[16:46] <bschaefer> sil2100, everything works, and it compiles (more important part)
[16:46] <bschaefer> sil2100, if there are bugs that arise from it in the edge barriers bit we can surly fix them
[16:46] <sil2100> mlankhorst, bschaefer: sooo, when can we get this merged in and new X in -proposed? ;)
[16:46] <bschaefer> sil2100, I don't see why not :)
[16:47] <sil2100> Since we have to do this transition someday!
[16:47] <sil2100> bschaefer: how long is that in the PPA already?
[16:47] <bschaefer> sil2100, hmm it was merged with trunk a couple weeks ago...but its been done since march :)
[16:47] <didrocks> sil2100: did you see the discussion we had this morning (I pinged you as well), it seems that we didn't get the traditional public call for testing with each drivers
[16:47] <didrocks> so this is needed before taking any action :)
[16:48] <sil2100> Ok, so should we contact baloons about that?
[16:49] <bschaefer> cool, well ill try to get Trevinho to test that ppa out, hes been busy :)
[16:49] <bschaefer> but so far everything works (and the new xinput stuff is a lot nicer!)
[16:50] <didrocks> sil2100: I think, but mlankhorst would know more about the procedure :)
[16:51] <sil2100> Ok ;)
[16:51] <bschaefer> sil2100, also, any news on the 12.04 SRU from the XIM stuff? I thought I just saw the first SRU go through...
[16:51] <bschaefer> sil2100, also...shouldn't we look at getting the SRU branches merged before doing the SRU?
[16:52] <bschaefer> (as they are sitting on approved atm)
[16:55] <sil2100> bschaefer: oh! Probably there's no auto-merger, forgot about that!
[16:56] <bschaefer> sil2100, let me get you links to the branchs :)
[16:56] <bschaefer> https://code.launchpad.net/~brandontschaefer/unity/xim-focus-fix-12-04-SRU/+merge/163619
[16:56] <bschaefer> https://code.launchpad.net/~brandontschaefer/nux/xim-support-12-04-SRU/+merge/163577
[17:00] <pete-woods> sil2100: I think that latest change might make the build happy - somehow I managed to replicate the original fail on my machine
[17:04] <sil2100> pete-woods: let me take a look
[17:05] <sil2100> bschaefer: ok, I'll bookmark those and maybe before we get those in, I'll double check with the SRU guys that they're fine with it
[17:05] <sil2100> Since right now no one seems to be around
[17:05] <bschaefer> sil2100, alright, thanks!
[17:10] <Saviq> olli, asac, image with Mir?
[17:11] <olli> Saviq, yep
[17:11] <Saviq> olli, awesome
[17:11] <olli> concerned you don't know about it
[17:11] <olli> asac ^
[17:11] <olli> kgunn, ^
[17:12] <Saviq> olli, yeah, sorry, I wasn't involved with the developments on that side recently, but now that sscopes and notify is out the door, that's definitely our focus
[17:13] <sil2100> pete-woods: tested the branch - I would say it's ok now! It's spitting out a warning about the symbol, but does not crash thanks to the 'optional' parameter
[17:13] <sil2100> \o/
[17:13] <pete-woods> :D
[17:15] <sil2100> pete-woods: approved!
[17:15] <sil2100> See you tomorrow
[17:15] <pete-woods> sil2100: woot!
[17:15] <pete-woods> cheers
[17:17] <asac> Saviq:it was in my notes as "done", but it turned out it will be there by ricmm later today
[17:17] <asac> so check in #ubuntu-touch
[17:17] <Saviq> asac, ah, that's where the cool kids hang out, eh?
[17:18] <asac> right
[17:18] <asac> thats the central hub where the future of the next world dominating linux OS is forged
[17:24] <larsu> kenvandine: https://code.launchpad.net/~larsu/+junk/gsettings-qt
[17:24] <larsu> kenvandine: hm, should have been in #ubuntu-desktop really
[18:09] <seb128> mterry, just for info I'm looking at the gtk breakage on arm, unity-greeter failed on what is a gtk issue
[18:09] <seb128> mterry, but gtk/glib didn't change, seems like a toolchain problem :/
[18:10] <mterry> seb128, I was looking at that too, and was thinking one solution for the greeter is just to ignore warnings during the test.
[18:11] <mterry> seb128, is this a widespread problem?
[18:11] <seb128> mterry, feel free to go for it, gtk rebuilds hit the same issue though and I think there is a chance gresource or gtk animations are broken on armhf atm
[18:11] <mterry> hmm
[18:11] <seb128> not that we had lot of gtk on arm users, but that need to be addressed
[18:11] <seb128> I had no luck today fighting it
[18:11] <seb128> I can reproduce on a porter box
[18:12] <seb128> and neither gtk or glib changed between the working gtk and the archive rebuild which has the issue
[18:12] <seb128> very few packages changed
[18:12] <seb128> but that include gcc and binutils
[18:12] <seb128> doing a test build on my tablet now to see if I can nail dow which one is causing it
[18:46] <ricotz> Trevinho, hi
[18:47] <ricotz> looks like the root cause is that "get_application_for_desktop_file" doesnt provide a just started non-favorite application, although that this call gets triggered is a bug on my side
[19:49] <mfisch> mhall119: ping
[19:51] <mhall119> mfisch: in an irc meeting, but pong
[20:39] <mfisch> mhall119: sorry just ping me when your meeting is over
[20:56] <mhall119> mfisch: ok, I'm available now
[20:56] <mfisch> mhall119: I'll ask you in #utouch
[22:04] <kgunn> robert_ancell: so i uninstalled libboost-system1.49.0 just in case....
[22:04] <kgunn> and now in the unity-system-compositor.log
[22:04] <kgunn> it says /usr/bin/unity-system-compositor: error while loading shared libraries: libboost_system.so.1.49.0: cannot open shared object file: No such file or directory
[22:05] <kgunn> does your log say anything?
[22:05] <robert_ancell> kgunn, can you run ldd /usr/bin/unity-system-compositor|grep boost
[22:06] <robert_ancell> kgunn, I'm just switching from the staging ppa to the u-s-c ppa so I have the same behaviour
[22:07] <kgunn> https://pastebin.canonical.com/93190/
[22:07] <kgunn> robert_ancell: ^
[22:08] <robert_ancell> kgunn, I do not get that - if appears your package is linking against the wrong version of libboost_program_options
[22:08] <robert_ancell> kgunn, this wasn't compiled locally?
[22:08] <kgunn> nope
[22:09] <robert_ancell> kgunn, for comparison (http://paste.ubuntu.com/5785143/)
[22:09] <kgunn> installed via apt-get dist-upgrade today....saw specifically unity-system-compositor
[22:09] <robert_ancell> kgunn, md5sum /usr/bin/unity-system-compositor?
[22:09] <robert_ancell> mine is 9d84498719daba4462161090260c690b
[22:09] <kgunn> stays freaky
[22:09] <kgunn> 9d84498719daba4462161090260c690b  /usr/bin/unity-system-compositor
[22:10] <robert_ancell> wft
[22:10] <robert_ancell> wtf
[22:10] <robert_ancell> kgunn, dpkg -s libboost-program-options1.53.0?
[22:11] <robert_ancell> kgunn, and dpkg -s libmirserver0
[22:12] <kgunn> dpkg for libboost-program-options1.53   ->   https://pastebin.canonical.com/93192/
[22:13] <kgunn> for libmirserver0   ->   https://pastebin.canonical.com/93193/
[22:14]  * robert_ancell scratches his head
[22:15] <robert_ancell> kgunn, so to confirm. unity-system-compositor only resolved symbols if you have libboost_program_options 1.49 installed?
[22:15] <mhr3> Trevinho, ping?
[22:15] <kgunn> seems so
[22:15] <robert_ancell> kgunn, does ldd /usr/bin/mir_demo_server show the demo server also linking against the old libboost?
[22:16] <robert_ancell> kgunn, or ldd /usr/lib/libmirserver.so.0.0.4?
[22:17] <robert_ancell> something in the chain must be compiled against an old version
[22:19] <kgunn> i don't even have mir_demo_server....and libmirserver.so.0.0.4 is showing libboost 1.53
[22:19] <kgunn> hmmmm
[22:19] <kgunn> can i purge and redo dist-upgrade
[22:19] <kgunn> robert_ancell:
[22:20] <robert_ancell> kgunn, I tried ppa-purge just before but I couldn't get it to work. The hack is to manually check the versions you have installed against https://launchpad.net/~mir-team/+archive/system-compositor-testing
[22:21] <robert_ancell> kgunn, mir_demo_server is in mir-demos - it would be good to check you can run a stand-alone mir server from a VT