[00:15] <imgbot> [00:15] <imgbot> [00:17] <Saviq> robru, could you please delete indicator-datetime, indicator-sound and unity-greeter-session-broadcast packages from silo 004?
[00:24] <Saviq> robru, and please reconfigure when that done, I just kicked a rebuild of unity8 and removed MPs that we don't need any more
[00:24] <Saviq> /sleep
[00:25] <robru> Saviq, sure thing
[05:12] <Mirv> testing infra shows signs of coming back to life, but not quite
[05:23] <Mirv> so um we should have 258 built for the unity8, I guess by cron if no-one around?
[05:28] <Mirv> wow, the silo 004 is still a bit amiss
[05:29] <Mirv> trying reconfigure + build/watch again
[05:45] <Mirv> doing direct magic on the silo project files
[05:52] <Mirv> direct magic helped, silo 004 now looking good again (after lightdm rebuild finishes)
[07:09] <didrocks> Mirv: hey! mind looking at the 2 failures (unity8 and messaging-app) on the dashboard? My guess is really a missing dep…
[07:13] <Mirv> so, unity8 resolved leaves messaging-app and dialer-app having some weirdness still (visible also on #256 test results, but locally both passed)
[07:15] <Mirv> let me check with manually updated image with the new unity8
[07:20] <imgbot> [07:31] <didrocks> Mirv: probably a missing dep, you think?
[07:46] <Mirv> well, I do have messaging app failing also locally after dist-upgrade. but my Unity8 also has an ugly grey background, which is probably not what robert saw when testing it.
[07:51] <ogra_> hmm, i wonder who turned off the 3am build
[07:52] <vila> Mirv: so messaging app failures are reproduced locally ?
[07:52] <Mirv> vila: "maybe", I'm not sure if my phone is good after dist-upgrade
[07:53] <Mirv> vila: but the thing is that they're now also shown for #256 while I have logs clearly showing how they passed locally.
[07:53] <vila> Mirv: ok, I'm looking at the unity8 ones for now, have ruled out python-autopilot-trace as a missing dep despite the warning as I found the same message in a passing run
[07:54] <Mirv> vila: unity8 is possibly resolved by this new #258 build, see robert's message on the mailing list
[07:54] <vila> ha, let me check mail
[07:54] <Mirv> vila: #257 was started too early so it lost packages, so it's broken
[07:54] <vila> O_o
[07:58] <vila> Mirv: argh, arck, stopping #257 analysis :-/
[07:59] <Mirv> I wonder if the "tested and ready for release" content-hub is causing some problems too, meaning if it's related to my grey background in unity8
[07:59] <Mirv> but maybe when #258 is upgradeable to it will be seen if it continues to be so
[08:11] <didrocks> vila: yeah, right now, I would say just keep an eye on the messaging-app one, I'll fetch info on the dialer-app (and ignore unity8)
[08:13] <ogra_> didrocks, i thought asac wanted to have both rolled back ?
[08:14] <didrocks> both?
[08:14] <didrocks> I don't know
[08:14] <didrocks> is it latest producing that?
[08:14] <ogra_> he talked about it last night
[08:14] <ogra_> dialer and messaging fail local tests too and there was no fix uploaded
[08:14] <ogra_> at least i dont see anything on -changes
[08:15] <didrocks> ogra_: and rollbacking make them pass?
[08:15] <didrocks> this analyze was done?
[08:15] <ogra_> (this went back and forth a little last night, i havent read the backlog, gimme a bit)
[08:15] <ogra_> didrocks, i dont think so, no
[08:16] <didrocks> ogra_: so why rollbacking? :p
[08:16] <ogra_> didrocks, aha ... seems dialer and messaging are tested as click apps
[08:17] <didrocks> ah, but I would expect all dialer test to fail
[08:17] <didrocks> not only a few
[08:17] <didrocks> right?
[08:17] <didrocks> vila: can you get that fixed? ^
[08:17] <ogra_> hmm, so what i grok from the backlog between plars and bfiller is that there is some issue with the testing
[08:18]  * didrocks finishes his download + flashing
[08:18] <didrocks> just to ensure that testing "the proper way" pass
[08:19] <ogra_> last thing i see on the topic: <boiko> plars: bfiller: found the reason for the failures, cooking a fix here
[08:19] <didrocks> hum
[08:19] <didrocks> so not only that?
[08:19] <ogra_> now he doesnt say where he fixes it :)
[08:20] <didrocks> ok, upgraded to the broken image done
[08:20] <didrocks> broken as in == no scope :p
[08:21] <didrocks> at least, the tests are starting here
[08:21] <didrocks> (messaging-app)
[08:21] <Mirv> nope, my grey background problem isn't fixed by reverting the content-hub landing, so it's more likely the scopes/unity8 landing in general
[08:21] <Mirv> or then just my cats again and no-one else has the problem
[08:21] <didrocks> Mirv: I don't have a grey background
[08:21] <sil2100> Mirv: ;)
[08:21] <didrocks> Mirv: just the default background
[08:21] <didrocks> and no scope
[08:21] <Mirv> bah
[08:21] <didrocks> which was to expect
[08:22] <ogra_> you should probably borrow Mirv's cats then
[08:22] <ogra_> seems to be needed to reproduce :)
[08:22] <didrocks> ogra_: I have enough regressions on my own to not add more :)
[08:23] <Mirv> good riddance regarding the white one, at least temporarily. he's quite demanding.
[08:23] <ogra_> heh
[08:23] <didrocks> Mirv: he likes camera as well, you should just do some videos
[08:23] <vila> didrocks: no idea what you're talking about O_o will read backlog
[08:24] <vila> ogra_: the discussion happened here ?
[08:24] <ogra_> vila, yep
[08:24] <ogra_> vila, 22:23 -> 23:03
[08:24] <sil2100> I wonder where the problems from dialer-app and messaging-app come from, I don't remember publishing those two recently
[08:25] <ogra_> (say my timestamps in xchat)
[08:25] <sil2100> And they're still .deb's
[08:25] <ogra_> sil2100, they changed on saturday
[08:25] <sil2100> Oh, that's why I don't know about it
[08:25] <sil2100> ;)
[08:25] <Mirv> UDS video appearances are enough :)
[08:25] <ogra_> i think robru published them
[08:26] <ogra_> (he did a few silo flushes over the weekend ... which would have been fine had we had tests :P )
[08:26] <Mirv> sil2100: the weird thing is that I've logs here running dialer-app+messaging-app tests successfully on #256 yesterday, but they are now visible in the #256 results too
[08:26] <ogra_> (and one image build per flush)
[08:27] <didrocks> ogra_: landing line 40
[08:27] <didrocks> Mirv: sil2100: ^
[08:27] <didrocks> but wasn't tested
[08:28] <Mirv> oh, right, I assigned that silo in the morning and kicked the build
[08:28] <Mirv> anyone with slightly more functional device could maybe test the silo
[08:30] <didrocks> there is clearly another issue
[08:30] <didrocks> all messaging-app tests are passing here
[08:31] <sil2100> didrocks: without changes from this silo?
[08:31] <didrocks> ouai, but running it as a deb package
[08:31] <ogra_> well they clearly dont on the infra
[08:32] <didrocks> yeah
[08:32] <didrocks> vila: so, it seems that messaging-app is configured as a click package
[08:32] <didrocks> instead of a deb
[08:32] <ogra_> and i know that t1mp pointed me to it yesterday as well
[08:32] <didrocks> maybe that was fixed and next image will have the configuration ok
[08:32] <didrocks> meanwhile, I'm installing the silo
[08:32] <ogra_> even before we had test results
[08:32] <didrocks> and will test
[08:33] <ogra_> didrocks, the backlog doesnt look like someone put any other effort into it
[08:33] <ogra_> (from CI side)
[08:33] <didrocks> vila: can you check that?
[08:33] <ogra_> (would really help if you could see the backlog :P )
[08:35] <didrocks> ogra_: well, I can go to irclogs.ubuntu.com
[08:36] <didrocks> but I guess that we understand what we know
[08:37] <sil2100> didrocks: hm, but from what I see on the testconfig.py, messaging and dialer are still configured to run as deb tests
[08:37] <popey> bah
[08:37] <didrocks> sil2100: do you see any recent change for that?
[08:37] <popey> broken unity. updated in between robru and ogra mails ☹
[08:37] <ogra_> yay timing
[08:37] <popey> http://popey.com/~alan/phablet/device-2014-03-25-083743.png
[08:38] <sil2100> http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/jenkins/testconfig.py <- no, just this python-gi addition
[08:38] <popey> clean aesthetic
[08:38] <didrocks> popey: my build should be finished quite soon
[08:38] <ogra_> popey, shiny
[08:39]  * popey files a bug in case anyone else happens to update
[08:39] <didrocks> hum
[08:39] <ogra_> sil2100, right, python-gi is installed by default (new autopilot-touch dep since the last upload)
[08:39] <didrocks> so, why do we have all failures on the infra?
[08:39] <didrocks> I don't reproduce that locally
[08:40] <imgbot> [08:40] <imgbot> [08:40] <didrocks> popey: upgrade ^
[08:40] <popey> will do
[08:40] <popey> how?
[08:40] <popey> system-image-cli ?
[08:40] <Mirv> popey: like suggested on another channel, it's this new "clean style"
[08:40] <didrocks> popey: yeah, the -cli
[08:40] <popey> ☻
[08:41]  * didrocks does the same
[08:41]  * Mirv wonders why -cli?
[08:41] <popey> cant start apps
[08:41] <didrocks> Mirv: well, no way to launch system settings
[08:41] <ogra_> http://paste.ubuntu.com/7150216/
[08:41] <didrocks> or
[08:41] <popey> actually you can
[08:41] <didrocks> su - phablet
[08:42] <didrocks> phablet@ubuntu-phablet:~$ upstart-app-launch ubuntu-system-settings
[08:42] <Mirv> oh, right, luckily I've such a good #257 in use that I can launch it (just ignore my and my device...)
[08:42] <ogra_> thats the complete log from the last dislaer-app test
[08:42] <popey> pull down on the indicator, battery, battery settings, back, updates ☻
[08:42] <didrocks> popey: my way is easier!
[08:42]  * didrocks clicks now install and restart
[08:42] <vila> didrocks, ogra_, Mirv, sil2100: So, from the backlog, it seems boiko worked on https://bugs.launchpad.net/messaging-app/+bug/1296826 made https://code.launchpad.net/~boiko/messaging-app/fix_autopilot_tests/+merge/212526 and https://code.launchpad.net/~boiko/dialer-app/fix_autopilot_tests/+merge/212527
[08:43] <didrocks> vila: yeah, we do have it in silo8
[08:43] <vila> but the later haven't even landed (they are approved but not merged)
[08:43] <didrocks> vila: so, you are sure that messaging-app isn't configured as a click app?
[08:43] <didrocks> vila: I'm just puzzled to not reproduce the messaging-app "all failures" locally
[08:43] <didrocks> so just want to ensure this is enough
[08:44] <didrocks> popey: upgrade works FYI
[08:44] <didrocks> I have scopes
[08:44] <didrocks> and yeah, I can see the first bug you'll open
[08:44] <ogra_> see the log
[08:44] <ogra_>  phablet-test-run -o /var/lib/jenkins/slaves/mako-11/workspace/trusty-touch-mako-smoke-daily/clientlogs/dialer_app -a /var/crash -a /home/phablet/.cache/upstart -v dialer_app
[08:44] <vila> didrocks: no I'm not, I can't connect that click/deb thing to the MPs
[08:44] <popey> bug 1297141
[08:44] <ogra_> that is what is run
[08:45] <ogra_> 02:20:32.131 INFO _launcher:230 - Launching process: ['/usr/bin/dialer-app', '-testability', '--desktop_file_hint=/usr/share/applications/dialer-app.desktop']
[08:45] <didrocks> vila: ok, ogra_'s log confirm it's not click
[08:45] <sil2100> ogra_: the config says that it installs the -autopilot packages, so I would say it doesn't use it as click - but let me see the infra logs
[08:45] <vila> popey: right, sounds like the way to go, but can someone explain me ... ack
[08:45] <didrocks> vila: let's hope the ofono-phonesim fix is what's failing on the infra reliably
[08:46] <didrocks> I don't understand why it's not here :/
[08:47] <ogra_> didrocks, do you have a SIM ?
[08:47] <didrocks> ogra_: not as of now, I remove it when running the AP tests on purpose
[08:48] <ogra_> the lab devices have a SIM card
[08:48] <didrocks> (and it's always a PITA to remove on the N4 as you know :p)
[08:48] <didrocks> oh
[08:48] <sil2100> oh?
[08:48] <didrocks> so the fake phonesim is starting too late
[08:48] <didrocks> and the hack could work?
[08:48] <didrocks> (stopping the service, restarting…)
[08:48] <didrocks> that would explain
[08:48] <ogra_> (which was causing the difference between tablet and phone tests in the past when we had ofono-phonesim issues)
[08:48] <didrocks> osomon is testing the silo btw as of now
[08:49] <ogra_> i think it just doesnt know which to chose
[08:49] <didrocks> yeah
[08:49] <didrocks> so "last wins"
[08:49] <didrocks> and it can be racy
[08:49] <didrocks> I guess
[08:49] <ogra_> if one wins at all (by looking at the fixes)
[08:49] <didrocks> yeah
[08:49] <didrocks> ok, so, we should be good
[08:49]  * didrocks crosses fingers
[08:49] <ogra_> try reproducing with a SIM inserted :)
[08:50] <didrocks> let's getting that published once tested and kicking an image
[08:51] <sil2100> didrocks: should we just wait for osomon's testing, or you want us to double-test that before landing?
[08:51] <didrocks> sil2100: if you have time, if you can do the double testing that ogra_ suggested
[08:51] <didrocks> with a sim card in
[08:52] <sil2100> ACK
[09:01] <ToyKeeper> Hello, r258... my, how you've changed since r257.
[09:02] <sil2100> ;)
[09:02] <ToyKeeper> Looks like I picked the wrong time to start...  got most of the way through image 257's tests (despite the complete lack of any scopes), and then 258 popped out.
[09:03] <didrocks> ToyKeeper: hey! always read the phone ML :)
[09:04] <ToyKeeper> I often try to get things done before I deal with email, to avoid spending the whole day on unplanned tangents.  :(
[09:05] <ToyKeeper> didrocks: Did you see bug 1292420?  It seems like something we should probably care about.
[09:06] <sil2100> ToyKeeper: was that working before?
[09:06] <ToyKeeper> I think it was working like 30+ images ago...
[09:06] <ToyKeeper> ... but I'm not 100% sure.
[09:06] <didrocks> ToyKeeper: too old and too many uncertainities to consider that as a blocker, just something to put high on the trusty release list
[09:07] <ToyKeeper> Okay, good to know.
[09:08] <ToyKeeper> Anyone else had issues with the screen getting stuck on after making/receiving calls?
[09:16] <didrocks> I just had one app getting stuck
[09:16] <didrocks> (webbrowser)
[09:16] <didrocks> try switching to another app
[09:16] <didrocks> to see if only the app is stuck or everything
[09:21] <ToyKeeper> ... what.  Image 258 insists that I don't even have cell service, and refuses to make calls.
[09:22] <seb128> thostr_, your indicator-sound landing from yesterday is creating issues on desktop, https://bugs.launchpad.net/ubuntu/+source/indicator-sound/+bug/1297034
[09:23] <seb128> thostr_, can you join #ubuntu-devel?
[09:29] <ogra_> gah
[09:30] <ToyKeeper> On the odd chance that my local cell provider is actually down, I think I'll wait until morning to continue.
[09:30]  * ogra_ forgot to re-set his DSL reconnect ... i might drop off the meeting after a few (for a few) minutes
[09:30] <sil2100> didrocks: be ther in a minute
[09:30] <didrocks> sil2100: coming soon as well
[09:36] <sil2100> GRRR
[09:52] <ogra_> didrocks, indicator-sound revert just landed
[09:52] <seb128> ogra_, does it matter to touch?
[09:53] <seb128> ogra_, it should be a noop for you, since you don't consider recommends
[09:53] <ogra_> seb128, yeah, it should, but didrocks was watching for the landing
[09:55] <sil2100> ok, now suddenly google hangouts crashed, that's something new
[09:55] <didrocks> sil2100: we just finished
[10:00] <Mirv> sil2100: FWIW I updated to 5.2.4.0-1 of google-talkplugin which still seems to work ok
[10:02] <sil2100> Mirv: maybe I need to upgrade as well, need to check on which version I am now on
[10:03] <ogra_> sil2100, did you verify dialer/messaging break with a SIM ?
[10:14] <ogra_> bah
[10:14] <ogra_> no wallapaer on the new unity/scopes
[10:15] <ogra_> seb128, is someone working on that ?
[10:16] <ogra_> wow, and the UI is a lot slower when scrolling
[10:16] <seb128> ogra_, no wallpaper on scopes?
[10:16] <seb128> ogra_, no idea, I've nothing to do with unity8/scopes
[10:16] <ogra_> seb128, yeah, we got new scopes and a new unity ...
[10:17] <seb128> try asking Saviq or mhr4?
[10:17] <Saviq> ogra_, that's by design
[10:17] <ogra_> seb128, now there is no wallpaper anymore, i guess they look for it in a different place now
[10:17] <ogra_> Saviq, you mean i cant adjust the wallapaer anymore ?
[10:17] <Saviq> ogra_, no, wallpaper is only meant to bleed-through below the header
[10:17] <ogra_> *wallpaper
[10:17] <ogra_> well, it doesnt do that either
[10:18] <Saviq> ogra_, yes, "meant to"
[10:18] <Saviq> ogra_, that's not implemented yet
[10:18] <Saviq> ogra_, check out https://drive.google.com/a/canonical.com/?tab=co#folders/0B-a_7E3tDxOgTTBybG1TZG9GWnM
[10:18] <ogra_> and if thats by design, the wallapaer selection in system-settings needs updating
[10:18] <Saviq> indeed
[10:18] <ogra_> .oO( why cant i type "wallpaper" today)
[10:18] <Saviq> ogra_, what's more scopes are meant to (not implemented yet either) choose a different background for themselves now, too
[10:18] <ogra_> thats why i pinges seb128
[10:18] <ogra_> *pinged :)
[10:19] <Saviq> ogra_, it was devised that people's wallpapers are too busy for a generic background
[10:19] <ogra_> sigh, who decided the back button needs to sit on the farthes reachable place on the screen ?
[10:19] <ogra_> i cant use my phone one handed anymore
[10:19] <Saviq> ogra_, you need bigger hands
[10:20] <ogra_> a twice as long thumb
[10:20] <ogra_> thats annoying
[10:20] <ogra_> really takes the fun out of using it
[10:20] <Saviq> ogra_, talk to Dani, he's doing the header UX
[10:20] <Saviq> ogra_, FWIW, this came out of user testing
[10:20] <ogra_> Saviq, well, putting it *in* the header at all is my issue
[10:21] <didrocks> Saviq: ogra_: when you tell "no wallpaper", you mean, you only have the origami effect?
[10:21] <didrocks> with the lines?
[10:21] <ogra_> i cant reach it if i hold the phone normal
[10:21] <Saviq> didrocks, yes
[10:21] <Saviq> ogra_, and is coming to apps, too
[10:21] <didrocks> ok, for me that was a wallpaper :p
[10:21] <Saviq> didrocks, in a sense, it is!
[10:21] <ogra_> didrocks, i'm using my own wallpaper (this is my main phone) since day one
[10:21] <ogra_> didrocks, now i cant anymore
[10:21] <didrocks> ogra_: even my desktop has the default!
[10:21] <Saviq> it's even made of... paper!
[10:21] <didrocks> heh
[10:21] <ogra_> didrocks, i found the default phone one awful when we started :)
[10:22]  * sil2100 had a nice manga wallpaper on his phone always
[10:22] <ogra_> (i did even set it before we had an app for this ... excitingly seb128 workd so well that the app just picked it up since :) )
[10:22] <didrocks> sil2100: btw, osomon confirmed the AP test pass
[10:22] <didrocks> sil2100: same for you?
[10:23] <ogra_> but that back button placement really wants me stop to use it
[10:23] <ogra_> thats the most pointless place to put such an important function
[10:23] <ogra_> i can only go back by using the other hand or hold the phone in a way that makes it awkward to reach the bottom parts
[10:23] <sil2100> didrocks: dialer app was ok on the first run, running messaging app - I also ran the tests without the PPA on my phone and confirmed that they're failing as on the infra ;/
[10:24] <Mirv> Saviq: ogra_: oh, it's (the background) by design, that explains it
[10:24] <didrocks> ogra_: stop to want to go backward, go forward! </non constructive joke ;)>
[10:24] <didrocks> sil2100: "grreat"! :)
[10:24] <ogra_> didrocks, heh
[10:25] <seb128> didrocks, ogra_, Saviq: I didn't update my device yet, I'm unsure what the wallpaper is but it's not likely something for me or settings: p
[10:25] <didrocks> seb128: agreed
[10:25] <ogra_> seb128, you cant set it anymore ... so the setting is pointless
[10:25] <didrocks> we should just be prepared to see rant/bugs opened
[10:25] <didrocks> Saviq: do we have this statement from design in a bug report?
[10:25] <seb128> ogra_, oh ok, "great"
[10:25] <didrocks> ogra_: it changes in the greeter, still?
[10:25] <Saviq> didrocks, yes
[10:25] <sil2100> No more girls as my background, OUTRAGE
[10:25] <ogra_> yes
[10:25] <ogra_> greeter is fine
[10:25] <ogra_> wallpaper isnt
[10:26] <seb128> shrug
[10:26] <Mirv> it's even more confusing because the background setting in settings shows "how it'd look like" in the old style
[10:26] <didrocks> Saviq: do you have it off hand so that we can direct people to it?
[10:26] <seb128> Saviq, since when do you know about that?
[10:26] <Saviq> didrocks, https://drive.google.com/a/canonical.com/?tab=co#folders/0B-a_7E3tDxOgTTBybG1TZG9GWnM
[10:26] <sil2100> When I was buying my first phone ever as a teen my only requirement was 'being able to set a custom wallpaper' ;)
[10:26] <didrocks> seb128: it was in the mwc image, but I thought personnaly they wanted to use another wallpaper by then
[10:26] <Saviq> didrocks, scopes are now meant to be able to set their own backgrounds, too
[10:26] <didrocks> Saviq: that a public folder?
[10:26] <Saviq> didrocks, I think so
[10:26] <didrocks> ok, will share on the ML if needed
[10:26]  * Saviq tries
[10:27] <Mirv> sil2100: when I was telling my parents what phone to purchase me as a present, the only requirement was to "be able to send SMS in addition to receiving them" (Nokia 1611 - awesomeness!) :D
[10:27] <Saviq> didrocks, no, it's actually not
[10:27] <didrocks> Saviq: yeah, tried an inconito mode
[10:27] <sil2100> ;p
[10:27] <didrocks> Saviq: can we have design stating that in a bug report?
[10:27] <didrocks> Saviq: that needs to be documented publicly
[10:27] <Saviq> didrocks, the images themselves are public, though
[10:28] <didrocks> Saviq: well, it's not a document
[10:28] <Saviq> seb128, well... since I saw the first designs...
[10:28] <ogra_> oh
[10:28] <ogra_> since when do we have a mediaplayer icon !
[10:28] <Saviq> didrocks, no, I just meant that the folder isn't, but the images themselves are
[10:28] <ogra_> lol
[10:28] <ogra_> which just gets me an error message
[10:28] <didrocks> Saviq: yeah, not something we can point people directly to
[10:28] <Saviq> ogra_, we had it before... looks like a bug in click scope
[10:29] <Saviq> ogra_, shouldn't be there
[10:29] <seb128> Saviq, which is?
[10:29] <ogra_> yeah, seems it cant start if it doesnt get a file handed over
[10:29] <ogra_> would need a file selection dialog or so
[10:29] <seb128> Saviq, it would have been nice if somebody told us earlier, rather than letting us do work that can be thrown away now
[10:29] <seb128> oh well
[10:29] <Saviq> seb128, well, no, not that far back
[10:29] <Saviq> seb128, separate wallpapers were there in the image already when I first saw those
[10:30] <seb128> Saviq, we did the new design changes like in january
[10:30] <ogra_> Saviq, wehn do we get the new app switcher UI ?
[10:30] <seb128> but a part of that still works for the greeter image selection I guess
[10:30] <Saviq> ogra_, it's coming
[10:30]  * ogra_ thought that would land alongside
[10:30] <Saviq> seb128, but it's not like it's a new thing for us that we get complete redesigns every month or so
[10:30] <ogra_> k
[10:30] <Saviq> ogra_, no, we wanted it separate on purpose
[10:30] <seb128> Saviq, it's new for me, I'm used to live in my desktop world ;-)
[10:31] <ogra_> ah
[10:31] <Saviq> seb128, slacker
[10:31] <seb128> lol
[10:31] <ogra_> the UI is really stuttery for me wiht a few apps open
[10:31] <Saviq> ogra_, yeah we need to work on performance
[10:31] <Saviq> ogra_, open apps don't really matter
[10:31] <ogra_> flicking through the app sand going back to the home scope the UI doesnt react for a while
[10:32] <ogra_> *through the apps
[10:32] <Saviq> ogra_, the items got much more flexible, so complex
[10:32] <Saviq> we need to do some tricks there
[10:32] <ogra_> yeah, understood
[10:39] <sil2100> oh, mhr3 reverted his ABI change
[10:39] <sil2100> He's back to soname 3
[10:40] <seb128> he got downgraded
[10:41] <mhr3> sil2100, yea, introduced lots of regressions :P
[10:41] <sil2100> ;)
[10:42] <sil2100> didrocks: publishing the dialer and messaging silo
[10:42] <didrocks> reverted even \o/
[10:42] <didrocks> sil2100: sweet!
[10:45] <imgbot> [10:45] <didrocks> urgh
[10:45] <didrocks> who started it?
[10:45] <didrocks> ogra_: ^ ?
[10:46] <sil2100> o_O
[10:46] <sil2100> Why would we need a new image now?
[10:46] <didrocks> we'll loose time as well on the testing side
[10:47] <sil2100> Image build shtaph!
[10:49] <ogra_> you cant stop them
[10:49] <ogra_> and i told you that infinity had a build queued
[10:49] <sil2100> didrocks: you think I can publish a thumbnailer landing?
[10:49] <didrocks> ogra_: but for Touch as well?
[10:49] <ogra_> which would get us one unconditional build
[10:49] <ogra_> didrocks, for touch what as well ?
[10:50] <didrocks> ogra_: a build queued for Touch? Not sure why infinity care about that one (or he rebuilds all flavors?)
[10:51] <cjwatson> bulk rebuild for beta
[10:51] <ogra_> didrocks, he turned off the cron job and had added all builds to his rebuild queue script it seems
[10:51] <ogra_> i turned cron back on, but cant remove his queue stuff :)
[10:51] <cjwatson> it was probably accidental but can't really be stopped now
[10:51] <ogra_> right
[10:52]  * sil2100 sighs
[10:53] <didrocks> ok
[10:53] <didrocks> I hope that this will be more fine grained in the future
[10:54] <ogra_> didrocks, well, he just wanted to be nice and make sure we still get a build even with disabled cron
[10:54] <didrocks> ogra_: yeah, but not sure why cron was disabled in the first place
[10:54] <ogra_> (not knowing about our way to build images on demand)
[10:54] <didrocks> as we are not part of beta freeze
[10:54] <cjwatson> again by accident
[10:54] <ogra_> didrocks, cron was generally disabled for all build
[10:54] <ogra_> s
[10:54] <didrocks> ok
[10:54] <ogra_> by accident
[10:55] <cjwatson> I'm going to move it out to a separate block so that it's more obvious
[10:55] <ogra_> and the queue script obviously was already holding a build for touch
[10:55] <didrocks> thanks cjwatson
[10:56] <ogra_> (and i warned you in the meeting so you wouldnt be surprised) :)
[10:56] <didrocks> doanac: hey, once you are around, I hear that parallel testing is mostly there, is it already the case?
[10:56] <didrocks> ogra_: didn't hear that part, only heared about the cron
[10:56] <ogra_> was the same setence ;)
[10:56] <ogra_> (or one before/after that one)
[10:56] <didrocks> that's maybe why I only got half of it :p
[11:03] <didrocks> ogra_: it's not Friday!
[11:03] <didrocks> :)
[11:03] <ogra_> no, on fridays i dont want to smash my phone to the wall :P
[11:06] <seb128> https://errors.ubuntu.com/problem/bb53c1417a42ad5963219f4244f7447f7f5b1042 suggests there is a new issue/segfault with the most recent telephony-service update
[11:07] <seb128> (not sure if that's known, I'm reviewing e.u.c and noticed it so mentioned it in case)
[11:07] <davmor2> Morning all
[11:07] <didrocks> seb128: mind mentionning that to #ubuntu-touch? (I guess osomon)
[11:08] <seb128> didrocks, did it, https://launchpad.net/ubuntu/+source/telephony-service/0.1+14.04.20140319-0ubuntu1 has mterry and boiko listed but neither of them seems to be online atm
[11:09] <didrocks> seb128: yeah, that's why I looked at all, and seems the apps team is doing most of the work there, hence the suggestion
[11:09] <seb128> didrocks, I'm going to ping boiko&mterry as well later
[11:09] <didrocks> seb128: thx!
[11:09] <seb128> yw!
[11:10] <didrocks> davmor2: so, once you get a nice stacktrace, you are going to hand it over to Saviq and mhr3?
[11:10] <didrocks> Saviq: mhr3: FYI, we can get some crashes in unity8 with new scopes by expanding some content (especially the big ones)
[11:10] <davmor2> didrocks: once I can break it again it seems to be refusing to now
[11:11] <davmor2> popey, didrocks: music preview in scope is playing up when the screen shuts down
[11:11] <didrocks> davmor2: playing up as?
[11:13] <davmor2> didrocks: randomly pausing just trying it from applications where it seems fine
[11:13] <popey> seems fine in scope here
[11:13] <didrocks> davmor2: can't that be the same infamous Qt issue?
[11:14] <davmor2> popey: select a long track, play it in the scope and let the device sleep and it pauses more erratically as it goes
[11:14] <popey> not here
[11:14] <popey> been playing a long track for a while now
[11:15] <davmor2> popey: hmmm okay
[11:15] <popey> run top, whats running?
[11:15] <popey> is it busy?
[11:15] <davmor2> popey: yeah could be I'll try that
[11:18] <davmor2> didrocks: could well be but it is more that it pauses for a second or two then carries on so I assume popey is correct and some process suddenly uses all the cpu for a second or something
[11:19] <didrocks> yeah
[11:22] <davmor2> popey: connect the device with music on to your computer, Open rhythmbox select nexus 4 from devices do all the tracks show up as unknown?
[11:24] <popey> yes
[11:24] <davmor2> okay so with top running no hang up
[11:25]  * davmor2 glares at ogra_ and cyphermox 
[11:27] <davmor2> trying music playback again with the lead removed
[11:27] <ogra_> hmm ?
[11:27]  * ogra_ only touches upstart jobs for mtp ... 
[11:28] <davmor2> ogra_: I know but I still blame you ;) It's the way to get things fixed :D   I did glare a cyphermox too to be fair :)
[11:28] <ogra_> :)
[11:30] <davmor2> popey, didrocks: okay so the music is playing perfectly now which is what I expected so I can only assume it was some random spike in use \o/ one less issue :) Now to try and get the damn thing to crash unity8 again :)
[11:31] <didrocks> yeah
[11:32] <didrocks> ok, so let me append an image build now
[11:32] <didrocks> done
[11:36] <Saviq> davmor2, didrocks, if you have steps to repro, that'd be greate
[11:36] <Saviq> -e
[11:37] <didrocks> Saviq: seems that om26er got it as well
[11:37] <didrocks> Saviq: basically, play with the expander on the app scope (for suggestion)
[11:37] <didrocks> I got it once
[11:38] <didrocks> and second time
[11:38] <didrocks> Saviq: want the crash file?
[11:39]  * didrocks wants for apport to finish the collect
[11:40] <Saviq> didrocks, I think we need to up the kill timeout again to make apport work... at least that's what I was seeing for exit crashes, although those might be special
[11:40] <didrocks> Saviq: it's collecting for more than a minute here
[11:41] <didrocks> so, we should be fine
[11:41] <didrocks> 40M until now
[11:41] <didrocks> ah done
[11:41] <didrocks> 43M!
[11:41]  * didrocks adb pull
[11:42] <didrocks> Saviq: a bug with the crash attached is fine? (better for you to retrace manually I guess?)
[11:44] <Saviq> didrocks, well, letting apport do its thing still good
[11:44] <didrocks> Saviq: let's do both :p
[11:44] <Saviq> didrocks, but you can submit somewhere for me to retrace
[11:44] <Saviq> didrocks, exactly :)
[11:44] <didrocks> Saviq: manual bug: https://bugs.launchpad.net/unity8/+bug/1297236
[11:44] <didrocks> Saviq: now, let me report the other one
[11:53] <didrocks> Saviq: bug #1297240 (let's wait for apport to retrace it)
[11:53] <didrocks> I added you to it
[11:53] <didrocks>  QQuickWindowPrivate::polishItems()
[11:53] <didrocks> too much polish! :)
[11:53] <Saviq> didrocks, mhr3 saw that, too
[11:54] <ogra_> want more french ?
[11:54] <didrocks> ogra_: I can ensure you you wouldn't have a crash in that case :p
[11:54] <ogra_> lol
[11:54] <ogra_> right
[11:54] <didrocks> only weird characters :p
[11:54] <didrocks> and typos in the UI
[11:54] <didrocks> anyway, time for a run!
[11:55] <ogra_> run forest run !
[11:55] <didrocks> ;)
[11:55] <didrocks> I'll try to avoid the chinese president
[11:55] <didrocks> seems he's in town today and I can sense traffic jams :p
[11:55] <ogra_> oh, you run that far ?
[11:55] <didrocks> ahah
[11:55] <ogra_> ah, he is in town
[12:05] <imgbot> [12:05] <imgbot> [12:05] <ogra_> hmm, that was a pointelss build
[12:05] <davmor2> ogra_: http://people.canonical.com/~ogra/touch-image-stats/259.changes no changes what?
[12:06] <Mirv> so no new messaging/dialerapp
[12:06] <ogra_> davmor2, yeah, no uploads between 258 and 259
[12:07] <Mirv> "47 minutes ago" (messaging/dialer-app, published, according to LP)
[12:07] <ogra_> right
[12:08] <ogra_> plus publisher run ... which can be another 30min
[12:08] <Mirv> yep, and anyway the image build started half an hour before even that
[12:08] <ogra_> but i think didrocks said he queued a build
[12:09] <ogra_> (though that will be in line with beta builds as well i think, might take a while til it gets a free slot now)
[12:09] <ogra_> oh
[12:09] <ogra_> ne build is already running ... why did the bot not pick it up
[12:10] <Mirv> imgbot: feeling tired of all the builds?
[12:15] <cjwatson> There aren't any beta builds happening at the moment
[12:15] <cyphermox> davmor2: popey: I'm already aware of the tracks showing as unknown, but there were higher priorities than mtp tbh
[12:16] <davmor2> cyphermox: pfff who needs to make calls and listen to them on bluetooth ;)
[12:19] <davmor2> cyphermox: is there a bug for it?
[12:21] <cyphermox> I think there is yeah
[12:21] <cyphermox> well, there are only about 10 bugs for mtp so could you please check and file one if there isn't ?
[12:22] <cyphermox> actualy, hold on, I think I have the tab open
[12:22] <cyphermox> no bug for that
[12:22] <cyphermox> please file one ;)
[12:22] <davmor2> cyphermox: will do
[12:23] <cyphermox> thanks
[12:27]  * Chipaca waves
[12:29] <sil2100> Hello o/
[12:34] <sergiusens> didrocks, sil2100 I just added line 44 for Chipaca on the sheet, can we assign a silo?
[12:34]  * sergiusens just got back, may be lacking info
[12:35] <sil2100> sergiusens: sure :) hi!
[12:35] <sil2100> sergiusens: you're in luck, we have some free silos
[12:36] <sergiusens> sil2100, heh, it seems we may finally be able to land the phablet-tools one as well; just waiting for doanac to be here to ask him to trigger a full run for ci
[12:46] <t1mp> asac: hello
[12:47] <Mirv> sil2100: are you querying for packaging ack for the album art service?
[12:49] <sil2100> Mirv: yeah, I'm dealing with it, just got context switched out of that one - pushing it forward now
[12:50] <sil2100> didrocks: hello! Do you have a moment for a packaging ACK? 2 packages need ACKing, looking safe:
[12:50] <Mirv> ok, just curious
[12:50] <sil2100> http://162.213.34.102/job/landing-009-2-publish/34/artifact/packaging_changes_thumbnailer_1.1+14.04.20140324-0ubuntu1.diff
[12:50] <sil2100> http://162.213.34.102/job/landing-009-2-publish/34/artifact/packaging_changes_ubuntu-ui-toolkit_0.1.46+14.04.20140324-0ubuntu1.diff
[12:50] <sil2100> Mirv: thanks for reminding ;)
[12:53] <t1mp> asac: I have an issue with getting *all* app autopilot tests to pass for merge requests in ubuntu-ui-toolkit.
[12:53] <sil2100> ogra_: hello! Can you take a look on those 2 above ^ ? :)
[12:54] <t1mp> asac: I cannot test with image 250, because I cannot apt-get install the autopilot tests for the versions of apps that are in the image (they are not in the archive anymore?)
[12:54] <sil2100> ogra_: I don't like the double-changelog entry for the UITK, but otherwise well, seems ok
[12:54] <t1mp> asac: if I apt-get update first, I get mismatching versions, and apt-get upgrade and/or installing a newer (not promoted) image will give me apps/tests which are failing
[12:55] <t1mp> asac: how would you propose we proceed? I can top-approve MRs where all tests except known-broken ones (in the latest proposed image) pass, and leave a note that some tests were not executed successfully?
[12:55] <t1mp> asac: that way we can proceed to add new features/bugfixes, but it won't give us 100% good test results so there is always a risk of breaking stuff
[12:58] <asac> t1mp: feels like a bug in your test job that it can't test click apps
[12:59] <asac> maybe lets get that fixed? Alternatively, you should be able to do the landing properly by doing those tests manually
[12:59] <asac> (like most other landers do right now)
[12:59] <t1mp> asac: do all apps have click packages?
[12:59] <asac> t1mp: do you have a job that tests your silo?
[12:59] <asac> t1mp: yes all but two
[12:59] <asac> almost all have been moved to click
[13:00] <asac> t1mp: where is your test job? and who did that job?
[13:00] <t1mp> asac: no, our policy is to pass all tests before top-approving an MR. And no MR goes into a silo before it is top-approved
[13:00] <asac> t1mp: oh i think you use a local script that runs them all?
[13:00] <t1mp> asac: I am trying to run the tests locally
[13:00] <t1mp> asac: yes
[13:00] <t1mp> asac: if there is a better way, please educate me
[13:00] <asac> t1mp: yeah, thats not complete then. you can run them all manually once and get kaleo to support that script?
[13:01] <t1mp> asac: I am running kaleo's script to run them all locally
[13:01] <asac> t1mp: well, that script was developed by kaleo for convenience
[13:01] <asac> so he missed click apckages. you can run them without that script. check with sergiusens
[13:01] <asac> and then have the script fixed
[13:01] <asac> sergiusens: ^^
[13:02] <asac> sergiusens: any doc that shows how to run click tests with phablet-test-run?
[13:02] <t1mp> asac: phablet-click-test-setup should install the click tests for all apps?
[13:02] <sergiusens> asac, in the wiki
[13:02] <asac> t1mp: probably with some command argument
[13:02] <asac> t1mp: https://wiki.ubuntu.com/Touch/Testing
[13:03] <sergiusens> https://wiki.ubuntu.com/Touch/Testing#Running_Click_tests
[13:03] <asac> sergiusens: how do we keep the list of commands in sync?
[13:03] <asac> do we have a process or automationm in place?
[13:03] <sergiusens> asac, can you rephrase that question?
[13:03] <asac> sergiusens: i see a list of commands in that section
[13:03] <asac> sergiusens: that gives the impression that if you run those you have run all
[13:03] <sergiusens> asac, we don't change them
[13:04] <sergiusens> I consider that 'backwards' compatibility to never break ci
[13:04] <t1mp> asac, sergiusens thanks
[13:04] <t1mp> sergiusens: is that list of click tests supposed to be complete? a bunch of apps are not in the list there
[13:04] <ogra_> asac, as i mentioned yesterday, ev tasked doanac with developing a phablet-test-run-all script
[13:04] <asac> sergiusens: no. that list has like 10 apps
[13:04] <asac> sergiusens: how do we ensure that we add app 11
[13:04] <sergiusens> t1mp, that's easy to answer
[13:04] <asac> in case we add it to our image
[13:04] <ogra_> to actually have a wrapper that runs everything in one go
[13:05] <sergiusens> t1mp, these are all the preinstalled apps http://people.canonical.com/~ubuntu-archive/click_packages/click_list
[13:05] <popey> asac: sergiusens i dont believe that page should list every app, but just list how to do it for one app and point to the list of installed apps
[13:05] <asac> sergiusens: the way it is presented gives the impression that its all tests (e.g. timp and me had the same question indicates that its a bit confusing) ... I would suggeste to add a comment so that people know that this list might not be up-to-date or complete; and also explain how to find the complete list
[13:06] <t1mp> sergiusens: so I still have a problem with everything that is installed as a deb in the image. If I use (for example) image 250, I need to use apt-get to install the autopilot tests for that right?
[13:06] <t1mp> sergiusens: but the autopilot tests with matching versions for image 250 seem no longer in the archive
[13:06] <asac> popey: right, but the list looks too complete to make that clear is my comment
[13:06] <popey> asac: its a wiki, feel free to edit it
[13:06] <sergiusens> asac, it's a wiki; people change it as they feel it adds clarity; just change it
[13:06] <asac> think just giving one or two examples would be better
[13:06] <t1mp> sergiusens: for example messenger-app, webbrowser, unity8
[13:06] <sergiusens> I didn't even create that wiki
[13:06] <boiko> t1mp: btw, I found the problem yesterday and proposed an MR for that, it is already merged
[13:06] <asac> sergiusens: well, there was an intend by someone to do it like that :)
[13:07] <t1mp> boiko: cool, thanks :) I saw it in the bug report.
[13:07] <asac> so i dont want to just wipe that intend. maybe the intend was to have it always have all commands for copy/paste
[13:07] <t1mp> boiko: do you know if it is in image 259?
[13:07] <sergiusens> t1mp, yes; you need apt-get for the debs
[13:07] <boiko> t1mp: flashing now, let's see
[13:07] <sergiusens> t1mp, I dn't like the tests as deb solution fwiw, I like the click approach I did a lot better; intended for times like these ;-)
[13:08] <ogra_> asac, the initial issue was that the names of the test packages varied a lot in the beginning ... some were ubuntu-$appname-app ... others had an underscore instaed of a dash others again had no ubuntu-
[13:08] <sergiusens> t1mp, for the tests as debs you need to discuss with the ci and qa team
[13:08] <popey> asac: sergiusens added a line to make that clear
[13:08] <asac> sergiusens: we would like to have that click_list for our images
[13:08] <asac> sergiusens: http://people.canonical.com/~ubuntu-archive/click_packages/click_list
[13:08] <ogra_> asac, so initially a detailed list was needed
[13:08] <asac> thats not versionmed
[13:08] <asac> so its probably as of today :)
[13:08] <t1mp> sergiusens: I'm confused who is on which team, and in which channel to find people for that
[13:09] <sergiusens> ogra_, I am already working on phablet-test-run click:com.ubuntu.music and it will run all the tests the manifest sasys it can run (which isn't just autopilot)
[13:09] <t1mp> so I come to this channel for all CI testing and landing questions
[13:09] <ogra_> if the tests now follow a scheme it should be fine to just have the procedure once
[13:09] <sergiusens> t1mp, this is obviously the wrong channel, don't know why people discuss everything here
[13:09] <asac> t1mp: you are in right channel :)
[13:10] <t1mp> haha ;) that explains my confusion
[13:10] <sergiusens> asac, just run 'click list' on the image
[13:10] <asac> t1mp: fginther, ev are your point of contact; doanac is working on an official version of kaleo's script
[13:10] <sergiusens> asac, there's even a test for that that I wrote
[13:10] <asac> sergiusens: why do you think that -ci-eng is not the right question to ask about anything CI and landing related?
[13:10]  * asac confused :)
[13:11] <asac> this is the channel for that purpose :)
[13:11] <ogra_> he said that to confuse you :)
[13:11] <asac> sergiusens: i know. but having that info off the image is also good
[13:11] <sergiusens> asac, was a general comment; I see people discuss architecture; general image testing here as well
[13:11] <ogra_> mission accomplished
[13:11] <sergiusens> and not everyone is here
[13:11] <sergiusens> and shouldn't be
[13:11] <Mirv> t1mp/etc: FYI for the non-click AP tests (ie come from debs) what I do is http://pastebin.ubuntu.com/7151115/
[13:11] <asac> ah
[13:11] <asac> well, thats the normal case for IRC channel
[13:11] <sergiusens> I would much rather prefer this channel to be treated as the ubuntu-release channel
[13:11] <asac> you just wander off topic and then you pull in crowds to the wrong channel. thats why private channels end up talkinga bout public matters etc.
[13:12] <Mirv> the first two separately and then outputting to a log file when executing a script with the rest of the lines
[13:12] <asac> sergiusens: this is the channel about our CI engine and services :)
[13:12] <t1mp> Mirv: I do something similar, but the problem is that for the latest proposed image, apt-get install fails saying those packages are not available (anymore). And running apt-get update first will give me mismatching packages or new packages that are broken
[13:12] <asac> landing is a service by CI
[13:13] <sergiusens> asac, t1mp, ok, so IMO the autopilot as debs, you need to contact either/or thomi, veebers, jfunk, fginther doanac
[13:13] <sergiusens> that's the list I believe
[13:13] <Mirv> t1mp: promoted, I guess? I tend to always test on latest image, promoted or not. but yes, I can see how that could happen.
[13:13] <Mirv> t1mp: generally our non-promoted images are in good enough shape too that one can compare own results to those (seen at http://reports.qa.ubuntu.com/smokeng/trusty/touch/ - usually close to 100%)
[13:13] <t1mp> Mirv: yes promoted (250). In newer images at least messaging-app tests were failing so I couldn't get 100% OK
[13:14] <Mirv> t1mp: yeah you can't get better results than on the dashboard if it's some other component breaking the tests
[13:14] <t1mp> Mirv: yes, *close* to 100%. But is "close" good enough to top-approve an MR? That's the question. It is not difficult to top-approve it, but do we want to run the risk?
[13:15] <Mirv> t1mp: if you see no regressions compared to dashboard results, then it should be (almost) as good as 100%. but whenever dashboard is not 100%, it makes of course trusting one's own results harder in the case of those specific packages having AP errors.
[13:15] <sergiusens> asac, this package-autopilot depends on package (= $Version) means that in the image test that you run, if you don't make sure that the package is not brought in as a dependency update you aren't really testing the latest image
[13:15]  * sergiusens had just thought of that
[13:16] <Mirv> t1mp: at least I've kept the dashboard as the meter of "100%", ie something that needs to be equalled
[13:16] <asac> sergiusens: so you say we still ship a deb to pull in test deopendencies?
[13:16] <asac> i just wondered this morning how that works in click case
[13:17] <asac> e.g. expressing your runtime/test requirements
[13:17] <sergiusens> asac, in click it's much fancier (and in contradiction to the ci train which came later)
[13:17] <sergiusens> asac, when a package lands to trunk, it's revno and branch used to build the click are annotated in the manifest
[13:17] <asac> sergiusens: so how does click express what needs to be put on image for testing?
[13:18] <t1mp> sergiusens, asac ok I try with the click tests and contact one of the people for the autopilot deb tests
[13:18] <t1mp> thansk
[13:18] <t1mp> *thanks
[13:18] <sergiusens> asac, so when you want to pull the tests for any click on any image, it will grab the branch, take the tests and push it into the device
[13:18] <t1mp> thostr_: how did a ubuntu-ui-toolkit MR end up in this silo? https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dFlCc1VzeVZzWmdBZS11WERjdVc3dmc&usp=drive_web#gid=27
[13:18] <t1mp> thostr_: it has not even been reviewed by someone in the SDK team yet
[13:19] <asac> sergiusens: right, but how do you express "i also need mock backend XYZ from archive?"
[13:19] <sergiusens> asac, you don't
[13:19] <didrocks> sil2100: did you ask on another core-dev for a packaging ack?
[13:19] <sergiusens> asac, clicks definition is to have everything in the bundle; no apt get allowed
[13:19] <thostr_> t1mp: are you saying neither james nor Satoris contacted you on this?
[13:20] <sil2100> didrocks: yes, but no answer so far ;) Poked ogra_ and cyphermox, but both seem to be busy!
[13:20] <sergiusens> asac, if you apt-get, you risk modifying the image enough to not be sure you are testing what you want to test
[13:20] <asac> sergiusens: right. so in case of the messaging/dialer app will we ship the mock backends as part of click?
[13:20] <t1mp> thostr_: no, not that I am aware of. There is no approval from an sdk-team member in the MR, and it has not been top-approved yet
[13:21] <sergiusens> asac, I'll leave that to the QA team, I asked too many times for them to architecture their solutions with click in mind
[13:21] <seb128> thostr_, you should avoid listing/landing non reviewed/approved mps as an usual rule (getting them in a silo for testing is fine, landing them is not)
[13:21] <asac> right
[13:21] <asac> we want the checklist to be used :)
[13:21] <seb128> thostr_, it's not the first time your landing have some unreviewed code included
[13:22] <didrocks> sil2100: +1 on thumbnailer
[13:22] <asac> sergiusens: yay thats evil :P
[13:22] <thostr_> seb128: yes, I know :(
[13:22] <sergiusens> asac, I actually really tired of playing catch up and being ignored
[13:22] <sergiusens> asac, that's why the gallery took so long to land; testing
[13:23] <didrocks> sil2100: +1 on toolkit, please remind me to promote the new binary package to main
[13:23] <seb128> didrocks, was it ever discussed to make CI land (in sense of upload to the archive) only if the includes mps are status approved?
[13:23] <didrocks> seb128: it was, but the fundation team has a special workflow which involve to never approving mps
[13:24] <didrocks> seb128: and I don't want to special case on a per project basis
[13:24] <ogra_> sil2100, oh, sorry, yeah. day is full of meetings for me
[13:24] <seb128> didrocks, oh, right, they do release and have 1 mp which is basically "merge the content of the release" back
[13:24] <asac> sergiusens: yeah i can understand. i think the problem is really that this click topic is not really owned by anyone for real
[13:24] <ogra_> didrocks, didnt you say above you had queued a new image build ?
[13:24] <asac> the overall click topic
[13:24] <didrocks> ogra_: I had, why?
[13:24] <asac> maybe foundations should take on driving the COMPLETE story
[13:24] <didrocks> seb128: right
[13:24] <seb128> didrocks, thanks
[13:25] <didrocks> ogra_: the request was ignored?
[13:25] <ogra_> didrocks, because there wasnt any after 259
[13:25] <didrocks> yw :)
[13:25] <didrocks> maybe the webui is buggy?
[13:25] <ogra_> might be
[13:25] <thostr_> sil2100: can you hold back silo 9 from landing (stuck anyway)
[13:25] <didrocks> it was clearly letting me add one and then the ui was "rebuilding)
[13:25] <seb128> didrocks, what about refusing to land but having a checkbox to force? ;-)
[13:25] <didrocks> seb128: enough checkbox
[13:25] <didrocks> or people will always use it
[13:25] <asac> sergiusens: the idea is that we only publish MPs that follow the checklist etc.
[13:25] <didrocks> and will complain about the tool :p
[13:25] <asac> seb128: ^^
[13:26] <asac> sorry sergio, wrong
[13:26] <sil2100> didrocks: thanks!
[13:26] <asac> seb128: landing team is suppposed to take samples and if folks dont follow help upstreams to fix their process
[13:26] <sil2100> thostr_: you want to free up the silo?
[13:26] <asac> and look closer for those components the next few times
[13:26] <seb128> asac, right, the theory and what happens in practice diverge though
[13:27] <asac> seb128: didrocks says his team has no time to quickly look at the MPs to see if they filled out a checklist
[13:27] <seb128> asac, they shouldn't
[13:27] <thostr_> sil2100: no, just wait couple of minutes first... I'll ping you in couple of minutes
[13:27] <asac> seb128: someone needs to control
[13:27] <seb128> they are not there to watch the universe
[13:27] <asac> seb128: well, they are the ones that get a ping, so they are the best to take a look
[13:27] <asac> because they already have a trigger
[13:27] <seb128> or you need to trust people
[13:27] <asac> upstreams shouldn't fail
[13:27] <asac> seb128: right, the reason we implemented it this way was that upstream managers asked for help
[13:27] <seb128> well, they shouldn't get a ping either :p
[13:27] <sil2100> thostr_: ok :)
[13:27] <asac> seb128: they couldnt enforce MP practices
[13:28] <asac> seb128: hence having the opportuntiy of an independent team quickly looking and blocking was perceived as very welcome
[13:28] <seb128> well, the system can enforce those for you (or at least help to catch errors)
[13:28] <seb128> right
[13:28] <seb128> it's welcome for those who get the reviews
[13:28] <seb128> not for those who have to review the universe for others
[13:28] <asac> its not that much imo
[13:28] <sergiusens> asac, the click thing is a ci thing most likely and there are bugs open
[13:29] <seb128> asac, it's probably worth some hours of work a day, when your days are already packaged that's starting being much
[13:29] <asac> seb128: no its a matter of 2 minutes for each landing
[13:29] <asac> seb128: you open aall MPs, quickly eyeball if there is checklist pattern
[13:30] <asac> otherwise bounce it back
[13:30] <asac> seb128: and we dont need to do it for every landing
[13:30] <asac> seb128: just take random sample
[13:30] <asac> like run "should-i-look"
[13:30] <seb128> right
[13:30] <asac> that spits out yes for every 190th
[13:30] <seb128> it's a bit orthogonal
[13:31] <asac> and then if you find something you feed back and ensure that this component doesnt land :)
[13:31] <seb128> it doesn't prevent us to helping people be noticed about potential errors in an automatic way
[13:31] <seb128> e.g "your landing include non approved changes"
[13:31] <seb128> noticed->notified
[13:31] <asac> seb128: well, then lander should really look at his MPs
[13:31] <asac> hence its just putting up that there might be checks
[13:31] <seb128> they should
[13:31] <asac> so they dont go lax
[13:31] <asac> if you know there will never be alcohol checks on streets
[13:31] <asac> you start not caring about that :)
[13:32] <asac> but the fact that you might get busted helps - even if super unlikely
[13:32] <didrocks> thostr_: so, oyu don't want us to publish 009?
[13:32] <didrocks> thostr_: testing was set to yes though and it's blokcing the ui toolkit that we need for unity8
[13:32] <thostr_> didrocks: no, wait until we get the MP also fully approved from sdk guys
[13:32] <asac> seb128: having a convenience tool that opens all URLs in the landing silo so you can quickly skim through would be nice
[13:33] <asac> but more automation we really dont need imo. its about establishing review practies
[13:33] <asac> after all :)
[13:33] <seb128> asac, I think it's the wrong approach here though, people should do the right thing because it's their job and they care about what they are doing, not because they want to avoid public shaming or something :p
[13:33] <didrocks> thostr_: ah, so it's not a rebuild/test issue, can you ensure next time it's approved first so that we are not into that situation?
[13:33] <asac> and the review is manual anyway :)
[13:33] <thostr_> didrocks: if we get a quick review from sdk guys we can go forward
[13:33] <didrocks> ok
[13:33] <asac> seb128: its not about shaming
[13:33] <asac> seb128: its about helping people to not go lax
[13:34] <thostr_> didrocks: I'll... was assuming james/Satoris had it gotten reviewed by now
[13:34] <asac> :)
[13:34] <didrocks> thostr_: no worry, keep us posted
[13:34] <asac> seb128: i assume good faith in anyway. everyone wants to do that if you ask them up front
[13:34] <asac> seb128: but then during battle etc. they might just dismiss their principles
[13:34] <thostr_> didrocks: can you go forward without silo9 for unity8 right now?
[13:34] <asac> and thats where another instance can help by reminding, helping them to not shoot themselves etc.
[13:34] <asac> and saying NO :)
[13:35] <thostr_> didrocks: then we review first and get it in after unity8
[13:35] <seb128> asac, well, I think some of those are honest overlook and the tool could easily ask you for confirmation when it stops an overlook
[13:35] <seb128> when it spots*
[13:35] <asac> seb128: i dont think its really an overlook, its a priority
[13:35] <asac> and then forgetting etc.
[13:35] <didrocks> thostr_: I guess we need both (with the toolkit), but I guess you have ~30 minutes as Saviq is on a crasher that I'm sure he want to sneak a fix in
[13:35] <asac> if you land without having reviewed the MP you are doing something serioyusly wrong
[13:36] <Saviq> didrocks, I can't even repro your crash
[13:36] <asac> if we give them a tool to bless those MPs we send the wrong signal. every MP needs to be carefully reviewed by lander before adding
[13:36] <Saviq> didrocks, so don't hold your breath
[13:36] <didrocks> Saviq: but multiple people can, and the stacktrace doesn't give you any hint?
[13:37] <Saviq> didrocks, not really, everything's in Qt
[13:37] <didrocks> argh
[13:37] <fginther> t1mp, sergiusens, I've read the backlog but may still be missing some context. It sounds like there is a general issue of getting the right version of autopilot packages installed for local testing?
[13:40] <Saviq> didrocks, and please ask the multiple people that you know to report their steps to repro
[13:40] <t1mp> fginther: yes, I think that is the problem
[13:41] <t1mp> fginther: I can get the latest versions of everything, but the latest version does not mean that all tests will pass. I only know that all tests should pass for the latest promoted image
[13:42] <sergiusens> fginther, yeah, it's not possible to get the tests for the latest devel image
[13:42] <t1mp> fginther: I need an image with all the autopilot tests installed on my device for which I know that without changes I get 100% OK on the tests
[13:43] <t1mp> fginther: with that, I can then install the debs from the MR and see if still 100% is OK
[13:43] <fginther> t1mp, so it sounds like what you need is a way to start with the last promoted image, and then install just your MR debs on to it?
[13:43] <t1mp> fginther: no, the promoted image does not include the autopilot tests
[13:44] <t1mp> fginther: so if the last promoted image included all the autopilot tests and all the apps+tests that we need to test, then that would be enough
[13:44] <didrocks> davmor2: popey: can you try to answer Saviq's request?
[13:45] <didrocks> davmor2: popey: on bug #1297240
[13:45] <t1mp> fginther: if you meant the last promoted image including the apps+tests, then the answer is yes :)
[13:45] <davmor2> didrocks: in  a call then Lunch then I can
[13:45] <fginther> t1mp, doesn't phablet-click-test-setup install the tests that match the click packages in the image?
[13:45] <fginther> t1mp, at least for the click apps?
[13:46] <didrocks> ogra_: so repushed the button…
[13:46] <ogra_> good
[13:46] <popey> didrocks: i haven't reproduced that yet
[13:46] <t1mp> fginther: I have to check that, but lets assume it does. Then still I need to get the AP tests for messaging-app, unity8, system-settings, ...
[13:46] <didrocks> ogra_: hum http://ci.ubuntu.com/smokeng/trusty/touch/mako/260:20140325.2:20140304/7371/
[13:46] <didrocks> ogra_: I do see 260
[13:47] <ogra_> weird
[13:47] <didrocks> and the number matches
[13:47] <ogra_> i didnt see it build
[13:47] <didrocks> argh
[13:47] <ogra_> heh
[13:47] <fginther> t1mp, ok, so there is still anything deb based that would be missing...
[13:47] <didrocks> ogra_: anyway to kill the build?
[13:47] <ogra_> didrocks, nope
[13:47] <t1mp> fginther: yes
[13:47] <fginther> t1mp, I'm starting to see the full picture, thanks for filling in my gaps
[13:47] <didrocks> ogra_: so, we are getting a void image again :p
[13:47] <didrocks> I guess that's ok
[13:47] <ogra_> yeah :/
[13:47] <didrocks> as long as 260 is testing
[13:47] <didrocks> which it is now
[13:48] <didrocks> ogra_: can you poke your script?
[13:48] <didrocks> ogra_: I want the image diff!
[13:48] <sil2100> ;p
[13:48] <t1mp> fginther: if I install image 250, I cannot get the matching AP tests for the deb-based packages that are installed. I think they are just no longer available in the archives because there are newer packages already. Is that correct?
[13:48] <didrocks> t1mp: launchpad has them
[13:48] <ogra_> http://people.canonical.com/~ogra/touch-image-stats/260.changes
[13:48] <ogra_> didrocks, ^^^^
[13:48] <didrocks> ogra_: \o/
[13:49] <ogra_> |HELP
[13:49] <imgbot> I am the firendly image watchbot
[13:49] <ogra_> hmm, bot is still there
[13:49] <fginther> t1mp, I suppose a trivial solution would be to archive all of the debs that were installed when image 250 was tested
[13:49] <ogra_> i guess it got confused with the overlapping builds
[13:49] <didrocks> ogra_: if we do a revert of my revert, we will get something like 12.10.2+14.04.20140325.12.10.2+14.04.20140324.is.12.10.2+14.04.20140320-0ubuntu1 like tomorrow :p
[13:49] <ogra_> haha
[13:49] <t1mp> fginther: it is not trivial for me, but if you can do that, as far as I can anticipate now, it would solve the problems
[13:50] <t1mp> didrocks: is there an easy (not time-consuming) way to get them from launchpad?
[13:50] <fginther> t1mp, ok, I'm not saying that's to solution, but it would perhaps provide the raw materials for one
[13:51] <t1mp> fginther: it would at least give us something to work with now so that we can approve MRs with confidence that at least there was a promoted image where the tests passed
[13:51] <davmor2> popey, cyphermox: can I get a confirm on https://bugs.launchpad.net/ubuntu/+source/mtp/+bug/1297301
[13:52] <cyphermox> done
[13:52] <didrocks> t1mp: you need to go to https://launchpad.net/ubuntu/trusty/+source/<packagename>, click on the version you need
[13:52] <popey> davmor2: confirmed
[13:52] <davmor2> popey: ta
[13:52] <didrocks> t1mp: and then, find the maching in arch all "autopilot" package
[13:52] <davmor2> cyphermox: ta
[13:53] <didrocks> davmor2: do we know when this regression started? was this already in latest promoted image?
[13:53] <t1mp> fginther: if it is possible to store the image that was used for testing (including the installed click packages and all tests) that we could re-use for testing then that would make it even easier
[13:53] <t1mp> fginther: so an image that includes all the tests
[13:53] <fginther> t1mp, right, a testing image snapshot
[13:53] <didrocks> t1mp: it's what is planned with the airline FYI
[13:53] <davmor2> didrocks: that has possibly been around for ever but is more noticeable now that mtp is more reliable
[13:54] <t1mp> fginther: exactly, that would be awesome :)
[13:54] <t1mp> didrocks: cool :)
[13:54] <didrocks> davmor2: ah ok :)
[13:54] <t1mp> didrocks: but I'm trying to find a solution for now. We have like 30 MRs pending, and manually searching all the autopilot debs for each package to test sounds like very time-consuming
[13:55] <imgbot> [13:55] <davmor2> ogra_: what happen to 260 /me shakes his fist at imgbot
[13:56] <didrocks> t1mp: the manual way is the only way I know of which is achievable as of today
[13:57] <ogra_> davmor2, the builds kind of overlapped
[13:58] <ogra_> davmor2, 260 is there but the bot missed it
[13:58] <fginther> t1mp, asac, what is the kaleo script that was mentioned?
[13:58] <davmor2> ogra_: you saying that you didn't prep the imgbot for being hammer by 3 images landing at once ;)
[13:58] <t1mp> fginther: https://code.launchpad.net/~fboucault/+junk/ciathome
[13:58] <asac> fginther: a convenience script so that sdk folks can run all AP tests locally/easily
[13:58] <asac> t1mp beat me
[13:58] <asac> :)
[13:59] <ogra_> davmor2, well, it watches for the build command and stores the PID ... if a buiuld runs with the same PID for some reason it doesnt notice that this is a new build ... i guess i need to review that concept
[14:01] <t1mp> fginther: I am using it with these changes: http://pastebin.ubuntu.com/7151299/
[14:01] <thostr_> didrocks: sil2100: invalidate silo 9 for now... obviously no quick way to get the approved from sdk guys
[14:01] <sil2100> thostr_: ok, so freeing up that silo then
[14:01] <fginther> t1mp, thanks
[14:01] <sil2100> thostr_: give us a sign once it can be set-up again
[14:02] <sil2100> thostr_: wait, silo 9?
[14:02] <sil2100> Crap...
[14:02] <sil2100> thostr_: the problem is, when you poked me earlier, I already published silo 9
[14:02] <thostr_> sil2100: yes, that was the one I said I'll clarify the situation for a couple of minutes ago
[14:02] <thostr_> sil2100: yes, but it's stuck
[14:03] <sil2100> thostr_: UITK and thumbnailer are stuck, but mediascanner2 is in -proposed already
[14:04] <sil2100> So hm, it would be a half landing, not sure if we can easily remove something from proposed now
[14:04] <sil2100> didrocks: ^ :(
[14:04] <thostr_> sil2100: half landing here is not good at all
[14:04] <ogra_> asac, didrocks, do we have any plans for the final weeks of the release and how we handle landings yet ?
[14:04] <ogra_> i guess that would deserve a mail with the planning to the ML
[14:05] <didrocks> thostr_: well, if half landing doesn't work, your packaging is there to block mediascanner2 in proposed, right?
[14:05] <ogra_> (how do we go forward while the rest of the world is frozen)
[14:05] <sil2100> cjwatson: hi! So, it seems we have a package version in -proposed that we wouldn't want to enter the release archive - can it be removed somehow easily?
[14:05] <thostr_> didrocks: I'd hope so... let me double check
[14:06] <sergiusens> fginther, t1mp fwiw, you can get the click tests for any released image you want
[14:06] <sergiusens> without going into r/w
[14:07] <fginther> sergiusens, thanks, that what I thought. The missing piece appears to be anything still installed via debs (like unity8-autopilot)
[14:08] <sergiusens> fginther, yeah, the thought was to have that preinstalled
[14:10] <thostr_> didrocks: sil2100: partial landing doesn't screw up anything, so we're safe there
[14:11] <didrocks> thostr_: ok, if we free silo, you need to push the mediascanner branch yourself
[14:11] <thostr_> didrocks: wait, I thought those have fully landed now (except uitk)?
[14:12] <ogra_> sergiusens, the plan was to get rid of autopilot on the image eventually :P
[14:13] <didrocks> thostr_: mediascanner will, not thumbnailer and uitk
[14:13] <didrocks> thostr_: they are both blocked on beta freeze
[14:14] <didrocks> thostr_: so, I can kick thumbnailer and uitk out
[14:14] <didrocks> so that they will never hit the proposed (nor release) pocket
[14:14] <didrocks> do you want that?
[14:15] <thostr_> didrocks: yes, kick those out. everything still save
[14:15] <didrocks> thostr_: ok, so only mediascanner2 will reach the release pocket
[14:15] <thostr_> didrocks: yes, that's ok
[14:15] <sil2100> ok
[14:15] <didrocks> thostr_: you need to push the new trunk manually as we are going to free the silo
[14:16] <sil2100> Phew
[14:16] <didrocks> thostr_: publication gives you a hint where those branches are: http://162.213.34.102/job/landing-009-2-publish/35/console
[14:16] <didrocks> 2014-03-25 13:35:14,486 INFO Pushing mediascanner2 to lp:~ps-jenkins/mediascanner2/trusty-proposed
[14:16] <didrocks> thostr_: so, just bzr pull/bzr push to trunk
[14:16] <sil2100> thostr_: so, we'll do the m&c, you just need to manually push this branch to your trunk
[14:16] <sil2100> thostr_: as we'll do a m&c without merging in the changes to bzr branches
[14:16] <didrocks> and uitk/thumbnailer flushed
[14:17] <sil2100> didrocks: thanks!
[14:17] <sil2100> I do the m&c then
[14:17] <sil2100> didrocks: even after the m&c, the trusty-proposed branches still stay, right?
[14:17] <sil2100> They're not removed in the onlyfreesilo case?
[14:17] <didrocks> sil2100: yep :)
[14:18] <sil2100> Awesome :)
[14:18] <didrocks> we keep it for that
[14:18] <didrocks> and always bzr push --overwrite
[14:20] <thostr_> didrocks: should be done now
[14:20] <didrocks> thostr_: looks perfect!
[14:21] <cjwatson> sil2100: ask on #ubuntu-release if this hasn't been resolved - I'm on vacation this afternoon
[14:21] <sil2100> cjwatson: ok, thanks ;) It's all resolved already
[14:21] <sil2100> Phew \o/
[14:25] <ogra_> cjwatson, didnt you say that 1h ago in -installer already ? go away !
[14:25] <ogra_> :)
[14:26] <Chipaca> does anybody know offhand what could cause dh-exec to not get called? I'm having a package ftb because it's not getting exec'ed
[14:26] <Chipaca> (that is: I have an executable .install file, with #!/usr/bin/dh-exec, with a => pattern, that fails)
[14:26] <didrocks> can't wait to have dialer-app and messaging-app test result: http://ci.ubuntu.com/smokeng/trusty/touch/mako/260:20140325.2:20140304/7371/
[14:26] <didrocks> Chipaca: is it executable?
[14:26] <cjwatson> Chipaca: make sure it's a 3.0 format source otherwise the exec bit isn't preserved
[14:27] <ogra_> didrocks, 261 will most likely step on your toes here
[14:27] <Chipaca> didrocks: yes
[14:27] <Chipaca> cjohnston: hm!
[14:27] <didrocks> ogra_: 260 will finish first
[14:27] <Chipaca> cjwatson: that's the one! :)
[14:27] <cjwatson> ogra_: I wish people would stop policing.  I was just dropping in to ask a question on a non-work channel
[14:27] <ogra_> now thats confidence
[14:27]  * Chipaca hopes and tries
[14:27] <ogra_> cjwatson, sorry
[14:32] <Chipaca> what're "debian native" packages?
[14:32]  * Chipaca looks it up
[14:32] <Chipaca> ok
[14:33] <ogra_> Chipaca, packages that contain the debian/ dir in their source
[14:33] <Chipaca> ogra_: but if i have .bzr-builddeb with split=true, that means the source as far as ubuntu is concerned is non-native, yes?
[14:33] <ogra_> (usually you have an orig.tar.gz with the upstream code plus everyhing thats in the debian/ dir i.e. distro patches ... native packages have that in one tarball)
[14:34] <ogra_> i think thats true, yes
[14:35] <sergiusens> ogra_, no, they are using bzr split deb
[14:35] <sergiusens> Chipaca, you shouldn't be native
[14:35] <ogra_> oh, ok
[14:35] <Chipaca> so 3.0 (quilt) it is
[14:38] <t1mp> thostr_: does this 15:06:23 < sergiusens> fginther, t1mp fwiw, you can get the click tests for any released image you want
[14:38] <t1mp> epaste
[14:39] <t1mp> sergiusens: ^ thanks for the hint. In our testing scripts we are still using some debs where we can use clicks, so for that part of the tests we can solve the testing problems by using click packages instead
[14:39] <t1mp> fginther: do you have an idea for a (temporary?) solution for the deb AP packages?
[14:40] <fginther> t1mp, it's possible to follow the chain from https://launchpad.net/ubuntu/trusty/+source/<source_package> and and get to the deb file that corresponds to a specific version
[14:41] <fginther> t1mp, so assuming you have the version installed for unity8, you can find the deb file for unity8-autopilot
[14:41] <fginther> t1mp, I don't have a better solution then that at the moment
[14:42] <bregma> Chipaca, if you run into grief using 3.0 (quilt) in a bzr split deb (and I expect you will), try using a .maintscript file instead of an executable .install file (see dpkg-maintscript-helper(1))
[14:43] <Chipaca> bregma: so far grief has been due to my rusty packaging skillz, but I'll remember (or rather, my xchatlog will ;-) )
[14:44] <bregma> been there, done that, got the commit logs to prove it
[14:46] <Chipaca> sergiusens: https://code.launchpad.net/~chipaca/ubuntu-push/bring-back-the-good-old-days-of-yore/+merge/212631 plz?
[14:47] <popey> davmor2: can you confirm bug 1297334
[14:47] <popey> ?
[14:47] <davmor2> popey: I'll check
[14:47] <popey> ta
[14:50] <sergiusens> Chipaca, ack
[14:50] <sergiusens> Chipaca, http://162.213.34.102/job/landing-012-1-build/40/console
[14:51] <Chipaca> doom doom doom
[14:52] <Chipaca> :)
[15:17] <cjwatson> Chipaca: you could also chmod +x the .install file in debian/rules
[15:17] <cjwatson> (if you aren't already sorted)
[15:17] <cjwatson> that's a perfectly reasonable old-style belt-and-braces technique
[15:18] <Chipaca> cjwatson: it was +x'ed, but because I'd been asked to remove source/format, it had stopped getting poked at
[15:18] <cjwatson> Chipaca: I mean at build time
[15:18] <Chipaca> ah!
[15:18] <Chipaca> lulz
[15:18] <cjwatson> e.g. right before dh_install
[15:18] <cjwatson> in an override_dh_install rule or whatever
[15:18] <Chipaca> cjwatson: good point. I'm overriding dh_install already, so it'd be trivial to do
[15:20] <imgbot> [15:20] <imgbot> [15:20] <Chipaca> now the package's gotten through the jenkins, what happens?
[15:32] <didrocks> bfiller: boiko: dialer-app at 100% back \o/
[15:32] <sil2100> ;)
[15:32] <didrocks> bfiller: boiko: messaging-app running as we speak
[15:33] <sil2100> I wouldn't expect anything else!
[15:33] <bfiller> didrocks: excellent
[15:35] <boiko> \o/
[15:36] <t1mp> where can I find the scripts that run to generate these test results? http://ci.ubuntu.com/smokeng/trusty/touch/
[15:37] <t1mp> doanac, plars ^
[15:37] <plars> t1mp: you mean to run the tests right?
[15:38] <t1mp> plars: yes
[15:38] <doanac> t1mp: lp:ubuntu-test-cases/touch, https://code.launchpad.net/ubuntu-test-cases/touch
[15:38] <plars> t1mp: ^that :)
[15:38] <t1mp> plars: I think the script that we are using in the sdk team is not optimal ;)
[15:38] <t1mp> doanac: thanks
[15:39] <plars> t1mp: it's pretty easy to use, if you want to use the run-smoke script, it'll do it all in one go and even provision for you
[15:39] <plars> t1mp: or you can run the individual pieces
[15:39] <plars> t1mp: let me know if you need any help getting it up and running
[15:39] <plars> t1mp: the main thing is that you'll need to set NETWORK_FILE to the path of your nm profile
[15:40] <plars> t1mp: that's pretty much the only thing that needs to be customized depending on your environment
[15:40] <t1mp> plars: what I want to do is run the autopilot tests for all the apps  to see if they are OK
[15:41] <t1mp> plars: but after installing the deb's that jenkins created for an MR
[15:41] <t1mp> plars: so basically I want to do the same as these smoke tests, except I want to do it before approving an MR
[15:41] <plars> t1mp: all the apps? or just certain ones?
[15:41] <t1mp> plars: all
[15:41] <plars> t1mp: be aware the full run takes quite a bit of time
[15:41] <plars> ok
[15:41] <t1mp> plars: about 2 hours ;)
[15:41] <t1mp> on my nexus4
[15:41] <t1mp> I think, may be even more, I'm not using a stopwatch :)
[15:42] <t1mp> plars: will the scriipt also install the latest (or promoted?) image and the click packages+autopilot tests?
[15:42] <plars> t1mp: yes
[15:43] <t1mp> plars: oh cool. We were using this https://code.launchpad.net/~fboucault/+junk/ciathome but that has some issues
[15:43] <t1mp> and it installs a bunch of apps with apt-get, while those apps are now clickified ;)
[15:44] <t1mp> kalikiana: ^ there is another script
[15:46] <plars> t1mp: yeah, our stuff only installs packages for non-click autopilot packages. For the click packages it uses phablet-click-test-setup to get the appropriate branches. Also, any packages installed for testing are uninstalled before the next test is started
[15:47] <t1mp> plars: what's the default location of my network manager profile?
[15:48] <t1mp> plars: can the script take my system nm profile to copy to device?
[15:48] <plars> t1mp: /etc/NetworkManager/system-connections is where they are located by default. Copy the one you want it to use to somewhere you can point it easily
[15:48] <didrocks> bfiller: boiko: and messaging-app AP tests passing \o/ well done guys :)
[15:49] <plars> t1mp: well, since this is normally intended to run on a server that's not using a wireless profile, it doesn't try to default to what you are already using
[15:49] <bfiller> didrocks: sorry it was broken in the first place, was unfortunate. boiko thanks for fixing so quickly
[15:50] <kalikiana> plars: is it feasible to take the script and run it locally to test a given merge request?
[15:50] <t1mp> plars: so this should be good: ./provision.sh -n /home/tim/network-manager-conf
[15:50] <didrocks> bfiller: well, it was tricky, with the SIM card in…
[15:51] <t1mp> wow we are at 261 already
[15:51] <t1mp> there is a new image every hour :)
[15:53] <sil2100> ;)
[15:54] <t1mp> plars: is there a way to run *all* tests without passing them all individually via -a ?
[15:57] <plars> t1mp: sure, just set TESTS=all APPS=all
[15:58] <plars> t1mp: for the non-autopilot stuff, you'll also need to install utah on your system from ppa:utah/daily
[15:58] <t1mp> plars: what is the non-autopilot stuff?
[15:58] <t1mp> I thought all apps are tested with autopilot
[15:59] <plars> kalikiana: not really set up to handle merge requests directly, you could possibly do something by either setting it up by hand and just running the tests rather than having it provision, or maybe build your packages in a ppa and have it use that ppa (you'll need to specify the packages to build)
[15:59] <plars> kalikiana: for the mr stuff, fginther is the expert - and he has other scripts that handle those tests
[15:59] <t1mp> plars: testing for an MR is exactly what I am going to try to do with it
[15:59] <plars> t1mp: there are the systemsettle tests, some default smoke tests, security, sdk, and things like that - things that don't have a ui to test with ap
[16:00] <t1mp> plars: when we propose an MR, jenkins runs CI and creates a zip-file with deb packages in it that has the changes of the MR
[16:00] <t1mp> plars: so what I was thinking is to simply install those debs and the run all the AP tests using your script
[16:01] <plars> t1mp: that's the process I mentioned a few lines up, and is different from the smoke tests that run against images
[16:01] <t1mp> plars: the main goal is to verify that the changes in UITK don't break anything in the apps (or at least in the apps autopilot tests)
[16:01] <fginther> t1mp, kalikiana, I'm working on changes to lp:ubuntu-test-cases/touch (or a branch) to specifically test MRs
[16:01] <plars> t1mp: that should work ok, just you'll want to run it with jenkins.sh rather than run-smoke so that it doesn't provision
[16:02] <plars> t1mp: you should provision with provision.sh though (see the help on it)
[16:02] <t1mp> plars: I already provisioned, or waiting for it to finish after flashing the new image
[16:02] <plars> t1mp: fginther has it pretty close I think, but this part was explicitly designed for image testing not mr testing. So fginther is doing some things to make it usable for both
[16:02] <t1mp> fginther: that's great :) how is the progress so far?
[16:03] <t1mp> fginther: tell us if you need anything from us, we are quite eager to get this working
[16:04] <fginther> t1mp, it's close to being usable as a replacement for the tests we're currently running in CI for MPs. It's goals are probably slightly different then your as I wasn't specifically attempting to run all tests
[16:04] <t1mp> plars, fginther one question, are you always using the latest proposed image for testing? it may have (known) failing tests
[16:04] <fginther> but that may be an easy thing to add
[16:04] <fginther> t1mp, yes, the MP testing currently uses the latest proposed.
[16:04] <t1mp> fginther: currently when we propose an MR, I don't think jenkins runs autopilot tests, right?
[16:04] <plars> t1mp: you can specify the image, for image smoke testing we are always testing latest
[16:05] <t1mp> fginther: for UITK, it runs the unit tests only as far as I know
[16:05] <t1mp> plars: ok
[16:05] <fginther> t1mp, if you're referring to ubuntu-ui-toolkit, only the uitk ap tests are executed.
[16:05] <fginther> t1mp, the infrastructure should support adding additional tests as parameters
[16:06] <t1mp> fginther: ok. we need to run all app autopilot tests to be sure nothing breaks
[16:06] <t1mp> fginther: but I don't know if that is feasible on jenkins, that's why we are running it at home on our own devices
[16:10] <fginther> t1mp, I'll look at adding it. My initial concern is that we don't have enough hardware to run all the tests for every MP, but I'll need to collect some data to know for sure
[16:10] <t1mp> ok provision.sh finished
[16:10] <davmor2> didrocks: okay so I don't seem to be able to reproduce the crash so I wonder if it is something that happens only after an update and it is doing the initial data check or something?
[16:10] <t1mp> fginther: true. and running all the tests will make it last at least 90m longer before we get results
[16:12] <t1mp> so I have this MR: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/optIn-newHeader/+merge/208662
[16:12] <t1mp> jenkins CI was succesfull and it created deb packages here: http://jenkins.qa.ubuntu.com/job/ubuntu-ui-toolkit-trusty-armhf-ci/892/artifact/work/output/*zip*/output.zip
[16:12] <t1mp> I'll download&unzip&dpkg -i those on the device
[16:12] <t1mp> what do I run afterwards so that all tests are executed?
[16:14] <t1mp> hmm.. provision.sh didn't install the AP tests for deb packages? for example unity8-autopilot is not there
[16:18]  * t1mp gotta go, bbl
[16:19] <davmor2> didrocks, kenvandine, Saviq: https://bugs.launchpad.net/ubuntu/+source/address-book-app/+bug/1297388  I've assigned it to address book app as ken is saying that content hub has done it's bit.
[16:21] <kenvandine> bfiller, ^^
[16:22] <bfiller> davmor2: this is a dupe, i;ll mark it as such. already fixed in the silo
[16:22] <kgunn> cyphermox: hiya, can i get a reconfig on silo 4 ?
[16:22] <davmor2> bfiller: awesome I just couldn't find a bug for it when I searched :)
[16:23] <bfiller> davmor2: https://bugs.launchpad.net/address-book-app/+bug/1295725
[16:24] <davmor2> bfiller: sweet thanks didrocks ^ use that bug instead :)
[16:25] <kenvandine> bfiller,  my gallery-app branch never got marked as merged, but it looks like it is merged
[16:26] <davmor2> kenvandine: it's just out to blag your head ;)
[16:28] <cyphermox> kgunn: sure
[16:31] <Saviq> cyphermox, can you please kick https://launchpad.net/~ci-train-ppa-service/+archive/landing-008/+build/5848957 for us, the dependency got built in the mean time
[16:42] <sil2100> didrocks: packaging ACK needed! http://162.213.34.102/job/landing-001-2-publish/lastSuccessfulBuild/artifact/packaging_changes_libcolumbus_1.1.0+14.04.20140325-0ubuntu1.diff <- symbols file added, looking nice, using the export symbols map to get rid of symbols leaking from source
[16:42] <didrocks> sil2100: in meetings, please ask another core dev
[16:42] <sil2100> didrocks: ACK
[16:42] <davmor2> didrocks: black screen on media player actually causes a crash https://bugs.launchpad.net/ubuntu/+source/mediaplayer-app/+bug/1297395
[16:42] <Saviq> sil2100, kick the unity8 build above please ↑↑?
[16:42] <sil2100> cyphermox: do you have another free moment for a packaging ACK?
[16:43] <sil2100> Saviq: k
[16:43] <Saviq> sil2100, actually it's building already :D
[16:43] <sil2100> Oh, it's kicked
[16:43] <Saviq> late refresh...
[16:43] <davmor2> didrocks: no public so should be accessible
[16:43] <cyphermox> sil2100: yeah
[16:43] <davmor2> s/no/now
[16:43] <sil2100> cyphermox: http://162.213.34.102/job/landing-001-2-publish/lastSuccessfulBuild/artifact/packaging_changes_libcolumbus_1.1.0+14.04.20140325-0ubuntu1.diff <- as mentioned before, symbols file added, looking nice,using the export symbols map to get rid of symbols leaking from source
[16:44] <cyphermox> sil2100: NAK. It changes a previous changelog entry; see the second hunk for debian/changelog
[16:45] <didrocks> bfiller: would you be available on the landing meeting to discuss those issues? ^
[16:45] <didrocks> davmor2: ^
[16:45] <didrocks> (sorry, in meeting, hard to track for me)
[16:46] <sil2100> cyphermox: hm, right, not sure why the bot did that, we already published something that did something similar, I wonder what's up
[16:46] <plars> sergiusens: ubuntu-device-flash doesn't seem to reboot to the bootloader when you specify --bootloader, which I guess probably also means it won't detect the device type. Is there any reason why?
[16:46] <plars> sergiusens: or am I missing a magic flag to make it do that?
[16:47] <bfiller> didrocks: on a meeting that will conflict so can't make it
[16:48] <bfiller> didrocks: address book app is a regression, not critical imo as you can still set the picture - it just blanks but is actually still there
[16:48] <sil2100> cyphermox: yeah, so, it happened already for unity-scopes-api and it got approved by a core-dev, probably it's not a good thing but anyway
[16:48] <bfiller> didrocks: it's been fixed with line 18 on the sheet which is in testing
[16:48] <bfiller> (along with other fixes)
[16:48] <cyphermox> sil2100: it's very very bad
[16:49] <sil2100> cyphermox: will look into the citrain code why that's happening and poke Didier on the meeting
[16:50] <cyphermox> I'm hungry
[17:00] <didrocks> bfiller: thanks!
[17:03] <didrocks> ogra_: coming?
[17:03] <didrocks> so fast :p
[17:03] <ogra_> no :P
[17:05] <davmor2> ogra_: http://weirdscaryandusualstuff.tumblr.com/post/884597811/finger-stretcher-for-the-aspiring-pianist-hailing
[17:11] <davmor2> didrocks: I hit this too so I don't know if it is linked to yours or caused by the mediaplayer crash https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1297410
[17:12] <ogra_> davmor2, well, i thought more about that "enlargement" SPAM ...
[17:12] <davmor2> ogra_: no real stuff :)
[17:13] <ogra_> davmor2, there are sometimes devices offered that are real stuff too :)
[17:16] <popey> davmor2: Go to Scopes scope, tap grooveshark, tap a song (I chose Get Lucky by Daft Punk), click Play in Grooveshark, click "Play" button in the browser. Note the audio level is much lower than the native music app.
[17:23] <davmor2> sil2100: https://bugs.launchpad.net/address-book-app/+bug/1295725 is the line you are looking for
[17:24] <davmor2> popey: https://bugs.launchpad.net/ubuntu/+source/mediaplayer-app/+bug/1297395
[17:24] <popey> ta
[17:25] <ogra_> popey, while i see you saying grooveshark, do you know who is responsible for these pieces ? in germany grooveshark is completely blocked i wonder if we shouldnt hide scopes you cant use in a country
[17:25] <ogra_> based on language or location selection
[17:26] <popey> good question!
[17:26] <popey> Saviq: ?
[17:28] <davmor2> popey: so grooveshark is working for me.
[17:29] <popey> works here too
[17:29] <popey> but audio level is low
[17:29] <popey> lower than music app or preview in scope
[17:29] <popey> bug 1297420
[17:29] <cyphermox> kgunn: what part have you changed for silo 4?
[17:31] <davmor2> popey: stay on the call
[17:31] <popey> kk
[17:36] <thostr_> sil2100: what do I need to do on silo1?
[17:39] <plars> sergiusens: if you replied I missed it, sorry. I had a power outage here
[17:39] <didrocks> balloons: you're not coming to the landing team meeting anymore?
[17:39] <sergiusens> plars, no, I saw you disconnect and waited,
[17:40] <sergiusens> plars, that info was in the email discussion with doanac; fwiw device autodetection works from fastboot
[17:41] <sergiusens> plars, I can add a --from-adb flag to the cli if you want; but wanted to avoid it being automatic for people not to mix a bootstrap from a wipe
[17:41] <sergiusens> a wipe just adds format data to the ubuntu commands
[17:41] <sergiusens> a bootstrap flashes more that what the system image updater would do and also cleans up much more
[17:42] <balloons> didrocks, sorry, not used to the new time and been a bit distracted. But yes I still plan on being there
[17:42] <plars> sergiusens: oh ok, so I could just adb reboot bootloader, wait a few seconds, and run u-d-f with --bootstrap and it should work?
[17:42] <t1mp> plars:
[17:42] <t1mp> tim@ideapad:~/dev/touch/scripts$ TESTS=all APPS=all ./run-autopilot-tests.sh
[17:42] <t1mp> plars: ^ I think that didn't run any autopilot tests
[17:42] <sergiusens> plars, no need to wait
[17:43] <plars> sergiusens: does --wipe add anything useful when you are already using --bootloader?
[17:43] <sergiusens> plars, ubuntu-device-flash will wait appropriately
[17:43] <t1mp> plars: http://pastebin.ubuntu.com/7152306/ at least I don't see autopilot logs
[17:43] <plars> t1mp: eh, yeah I think those only work if you are using run-smoke
[17:44] <sergiusens> plars, --bootstrap? no; it's the other way around :-)
[17:44] <t1mp> plars: test_results.xml http://paste.ubuntu.com/7152310/ or should I make the image writable manually?
[17:44] <dbarth> sil2100 or robru: we finished testing silo 002 (both phone ande desktop mostly): can you publish?
[17:44] <plars> sergiusens: not sure I understand, let me rephrase... so I want --bootstrap so that I get the new  bootloader (if any), but there would be nothing more done if I added --wipe on top of that right?
[17:45] <robru> dbarth, it's just oxide-prep right? doesn't actually bring in oxide right now?
[17:45] <plars> t1mp: if you use provision, it will do all the extra setup steps required like making it writable (if you use -w) and running phablet-click-test-setup
[17:45] <davmor2> popey: http://www.w3schools.com/html/tryit.asp?filename=tryhtml_audio_5
[17:45] <t1mp> plars: ok. I used provision.sh but without the -w
[17:46] <t1mp> plars: and if I want the autopilot tests for the non-click apps? I need run-smoke, but if I understand it correctly that automatically runs provision so I cannot use my own packages?
[17:46] <t1mp> is that correct?
[17:47] <plars> t1mp: are your packages in a ppa?
[17:47] <t1mp> plars: no they are the output of jenkins CI on an MR
[17:47] <sergiusens> plars, wipe is incompatible with bootstrap; you only want bootstrap
[17:47] <plars> t1mp: we could look at adding an option to skip provisioning from run-smoke
[17:47] <plars> sergiusens: got it, thanks
[17:48] <t1mp> plars: or add an option -z that downloads a zip-file with .deb's to test :)
[17:48] <t1mp> plars: but yes, without provisioning would be useful
[17:48] <t1mp> plars: then I can run all the tests on the device set-up the way I want it tested
[17:48] <plars> t1mp: there's also a way to add hooks that happen after the device is provisioned, but I haven't played with that at all
[17:50] <davmor2> popey: when you hang up on a hang out do you get what look like a mock chrome os desktop?
[17:50] <popey> i dont hang up
[17:50] <popey> i close the tab
[17:52] <davmor2> popey: hang up next time :) I got this http://ubuntuone.com/5iFagCM9dUDOvOklPKjUKb
[17:54] <popey> heh
[17:56] <plars> t1mp: untested branch at lp:~pwlars/ubuntu-test-cases/touch-no-provision
[17:57] <plars> t1mp: just specify run-smoke --no-provision
[18:03] <sergiusens> robru, silo-012 is fine according to Chipaca; can we publish?
[18:03] <sergiusens> robru, can we also make Chipaca the lander for ubuntu-push?
[18:05] <robru> sergiusens, did he get lander training?
[18:05] <kgunn> cyphermox: sorry, was eating lunch...
[18:05] <kgunn> cyphermox: so i addded an mp for indicator-sound
[18:06] <dbarth> robru: nope; oxide is in universe;and then we'll give a try to the silo on line 18 now
[18:07] <dbarth> robru: so silo 002 is safeto land i tested it without oxide and it works fine; and if oxide is added to my system it works fine as well here
[18:07] <robru> dbarth, yeah, but that's the problem. unity-webapps-qml is in main but is depending on oxide which is in universe. you need to MIR oxide for this
[18:07] <robru> dbarth, http://162.213.34.102/job/landing-002-2-publish/lastSuccessfulBuild/artifact/packaging_changes_unity-webapps-qml_0.1+14.04.20140324-0ubuntu1.diff
[18:07] <dbarth> robru: there should not be a build dep
[18:07] <dbarth> robru: is there one?
[18:07] <robru> dbarth, there's a dep dep
[18:08] <cyphermox> kgunn: you know that it conflicts with landings 5, 8 and 15 right? (just confirming again)
[18:08] <dbarth> robru: ugh; we'll turn that into a suggest, cause that's not a hard depend
[18:09] <sergiusens> robru, no, just ramped up by me; can we get him a training?
[18:09] <dbarth> ie, it works without oxide; but we'lll need to rebuild the ppa just to verify
[18:09] <robru> dbarth, ok, if you can add the change to one of the existing MPs then rebuild, then I can publish
[18:09] <alex_abreu> robru, yup doing it right now
[18:09] <robru> alex_abreu, great thanks
[18:09] <dbarth> robru: i will re-test very quick, and then ask to publish ;)
[18:09] <ogra_> mentioned seed change uploaded ...
[18:09] <robru> sergiusens, I don't see why not, but it's not really my call. poke didrocks about it i guess
[18:09] <sergiusens> Chipaca, ^^
[18:09] <robru> dbarth, great, thanks
[18:10] <Chipaca> i'll poke him in the morning
[18:10] <kgunn> cyphermox: yep...testing only
[18:11] <Chipaca> robru: was there an answer to sergiusens's first question (wrt publishing it) ?
[18:11] <kgunn> it actually landed/but got revereted from what i understand
[18:11] <robru> Chipaca, oh sorry, i still have to look at that
[18:11] <Chipaca> k
[18:11] <Chipaca> robru: should I wait, or should I walk the dog and get dinner? :)
[18:12] <robru> Chipaca, nope, just looking at it now
[18:12] <robru> Chipaca, remind me though, you had packaging fixes from didrocks right?
[18:12] <ogra_> kgunn, it broke the desktop
[18:13] <Chipaca> robru: from several people, yes; didrocks gave me a +1 to pre-new, if that means something to you?
[18:13] <kgunn> ogra_: yep...not complaining...just letting cyphermox know
[18:13] <ogra_> yeah
[18:13] <robru> Chipaca, yeah, that's what I was looking for ;-)
[18:13] <Chipaca> :)
[18:13] <robru> Chipaca, ok, if didrocks says +1, then I can publish it
[18:14] <Chipaca> cool
[18:15] <alex_abreu> robru, dbarth ok done
[18:15] <t1mp> plars: thanks, I'll try it tomorrow
[18:16] <robru> Chipaca, ok, so i just published silo 12. what happens now is that the package gets uploaded to trusty-proposed and it'll probably get stuck there for a while (partly because it's NEW, partly because of beta freeze etc etc) but eventually it's gonna make it into the archive, and at that point you'll want to "merge & clean" the silo. i'll handle it for today because you probably don't have permission, but just be aware of the workflow
[18:16] <robru> "ask for a silo -> build your stuff yourself -> test it yourself -> ask to publish it -> merge & clean yourself"
[18:17] <Chipaca> robru: so merge & clean only after it's gotten into the archive?
[18:17] <robru> Chipaca, yep. if you try it too soon it'll just error saying "not in archive yet". the goal is to have project trunks match what's in distro.
[18:18] <Chipaca> robru: ignoring the beta freeze, if something were in -proposed I'd expect to ask people to review the package to help it get out of proposed; is that still the case?
[18:18] <robru> Chipaca, only if it's stuck somehow. 90% of the time packages flow through -proposed in about an hour.
[18:18] <Chipaca> ah! ok
[18:19] <Chipaca> robru: so... when does the beta freeze thaw?
[18:19] <robru> Chipaca, it might get stuck by depwait on certain arch or perhaps an autopkgtest failure, or maybe a manual block by a release team member.
[18:19] <robru> Chipaca, hmmm, lemme check
[18:20] <robru> Chipaca, looks like friday.
[18:20] <robru> hmm
[18:22] <robru> alex_abreu, great. did dbarth start the rebuild or should I?
[18:22] <robru> alex_abreu, nm, i see the build job running ;-)
[18:27] <sergiusens> Chipaca, ubuntu touch specific packages (and more so unseeded ones) don't need to necessarily wait until friday
[18:27] <Chipaca> sergiusens: ah! ok. Now, how do I get seeded?
[18:30] <sergiusens> Chipaca, that's easy; but needs to be in archive first
[18:30] <sergiusens> Chipaca, it's an MR against ~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.trusty and get a core dev to germinate and push
[18:31] <Chipaca> sergiusens: ok... tell me more about not needing to wait for the thaw
[18:43] <Chipaca> sergiusens: ... :)
[18:43] <robru> Chipaca, right, sorry. in theory ubuntu-push is part of the touch FFe. however I think the release team will be too swamped with freeze work to care much for a new package...
[18:44] <Chipaca> robru: ah, ok. Ta.
[18:44]  * Chipaca adjusts plans accordingly.
[18:44] <robru> Chipaca, you might try pinging in #ubuntu-release. not sure how receptive they'll be
[18:44] <Chipaca> I'll do so. Tomorrow. :)
[18:45] <robru> hehe
[18:45] <Chipaca> robru: thank you!
[18:45] <robru> Chipaca, you're welcome!
[18:49] <sergiusens> Chipaca, it's just a matter of approaching the release team nicely
[18:49] <ogra_> or being not nice but bringing the right bribe with you
[18:49] <ogra_> ;)
[18:57] <ogra_> robru, (i'll bring that up tomorrow in the meeting too but thought it might help you today already) please done flush silos while the release team has not accepted a landing, else the packages are lost, the freeze block kicks in before the package is in -proposed (for desktop packages, touch will get auto-accepted after a few mins by a bot)
[18:57] <ogra_> s/done/don't/
[18:57] <ogra_> cyphermox, for you too ^^^
[18:59] <robru> ogra_, what? how do the packages get lost?
[19:00] <ogra_> robru, the freeze blocks them from going to proposed ...
[19:00] <ogra_> so the only actual copy is in the silo ... until they get accepted
[19:00] <robru> ogra_, but the merge & clean job errors if the package isn't already in archive? who is running merge & clean before packages get to -proposed? did i do this?
[19:01] <ogra_> robru, thats just a warning, nobody did anything wrong
[19:01] <robru> ogra_, oh ok, i thought you were saying some packages were lost, i was worried i did something wrong ;-)
[19:01] <ogra_> desktop packages are generally blocked and need manual approval from the release team before even going to proposed
[19:01] <ogra_> just a heads up ... only flush the silo if you are sure the packages made it into proposed
[19:02] <cyphermox> ogra_: robru: it breaks the silo scripts for publishing, that's what it is
[19:02] <cyphermox> the packages aren't list, they're just in the queue
[19:02] <cyphermox> *lost
[19:02] <ogra_> cyphermox, i was just told they are lost ... in #ubuntu-release
[19:02] <cyphermox> ah?
[19:02] <cyphermox> well, if they said it
[19:03] <ogra_> they arent oin proposed ... and the only actual copy is in the siol
[19:03] <ogra_> silo
[19:03] <cyphermox> i'd expect things to just fall into the queue for trusty-proposed
[19:03] <ogra_> if you flush at that point you might lose them
[19:03] <cyphermox> yes
[19:03] <cyphermox> oh, I think I see
[19:03] <robru> ogra_, it seems an empty warning since the official policy is to not run merge & clean until after the packages are in *distro*, not before they even get to *proposed*
[19:03] <cyphermox> syncs
[19:03] <ogra_> no, they dont fail, they wait for approval, i dont think there is any extra queue
[19:04] <cyphermox> ogra_: but since it's syncs it might be special.
[19:04] <ogra_> yeah
[19:04] <ogra_> ask the release team for details ... i just promised to forward the info :)
[19:05] <ogra_> i think seb128 nearly ran into such an issue, thats what brought up the topic
[19:05]  * ogra_ picks up GF ... brb
[19:06] <seb128> cyphermox, things hit "unapproved", not trusty-proposed, because of freeze
[19:06] <seb128> or unapproved doesn't have a copy of the source
[19:06] <seb128> just a pointer to the ppa content
[19:06] <seb128> so if the ppa is cleaned before the upload is approved, the file is not available anymore and the upload can't be accepted
[19:07] <seb128> I know that "merge only when package are in distro"
[19:07] <seb128> but you are going to run out of silos today if you stick to that
[19:07] <seb128> since nothing goes in distro in the next days
[19:07] <seb128> so you are not going to be able to claim back any silo
[19:07] <seb128> (well, nothing that is on images, which is still a good part of what we land)
[19:08] <seb128> talk to infinity or stgraber if that starts being an issue I guess
[19:08] <seb128> imho they should just let stuff in proposed at least and britney block them
[19:11] <kgunn> cyphermox: were you gonna reconfig silo4 ?
[19:11] <kgunn> maybe i got lost in the churn
[19:21] <cyphermox> kgunn: sorry, you did for a while but I did reconfigure silo 4 already
[19:21] <cyphermox> seb128: that's what I was saying
[19:22] <cyphermox> because it's a binary package copy the actual packages are only in the PPA until they really do hit proposed
[19:23] <cyphermox> if all else fails we can also reclaim silos and have some people start over testing if they really aren't ready; provided they agree of course
[19:24] <robru> cyphermox, seb128: clearly the correct solution is to have 5x more silos ;-)
[19:24] <cyphermox> clearly
[19:25] <cyphermox> let me put that to the agenda for tomorrow, since we're still on procrastination day
[19:25] <robru> lol
[19:25] <robru> cyphermox, ehh, i'll add it to the agenda later
[19:25] <cyphermox> alright
[19:36] <robru> alex_abreu, dbarth: how's silo 2? tested?
[19:38] <kgunn> cyphermox: huh....ever seen where a reconfig didn't take ?
[19:38] <kgunn> http://162.213.34.102/job/landing-004-1-build/78/console
[19:38] <kgunn> at least, its still whining that indicator-sound isn't there
[19:40] <dbarth> robru: just now; i noticed it's built now
[19:40] <dbarth> robru: won't be long
[19:40] <robru> dbarth, no worries, just checking. thanks
[19:41] <cyphermox> kgunn: looking
[19:46] <cyphermox> build should pass now I think
[19:47] <cyphermox> I didn't find anything funky, I jsut reran your project reconfigure and it passed
[19:47] <dbarth> robru, alex_abreu: all good here wit hthe 'suggests'; it installs fine with or without oxide; and runs fine after that
[19:47] <dbarth> robru: can you publish please?
[19:47] <dbarth> i'll merge and clean tomorrow morning and will get the next oxide silo ready
[19:47] <robru> dbarth, sure
[19:48] <dbarth> thanks
[20:20] <asac> robru: where do we have the silo temp bzr branches?
[20:21]  * asac remembers we had something like that
[20:21] <robru> asac, which? like the branches post-publish but pre-merge?
[20:22] <robru> asac, it's all here: https://code.launchpad.net/~ps-jenkins
[20:23] <asac> robru: ah we only have that post-publish?
[20:23] <asac> thought we had that everytime you reconfigure
[20:24] <asac> e.g. you can check whts in current silo basically
[20:24] <robru> asac, no, reconfigure just changes which branches it knows to pull from.
[20:24] <asac> robru: hmm. so building merges the MPs? I guess thats where pushing the branches woudl be great
[20:24] <robru> asac, best way to check what's in the silo is to just look at the most recent build job. grep for "trying to merge", it includes the MP link and the revision number that it pulled from there. you can open all the links and compare revision numbers to see if the silo has the latest of each MP
[20:25] <asac> basically, at the very first moment we havce them, publish them
[20:25] <asac> robru: well, thinking having the real tmp/merge branch would be nice
[20:25] <asac> but who knows :)
[20:26] <robru> asac, could be, i never had a need for it personally. didrocks is the one who controls the code, nothing I can do ;-)
[20:27] <kgunn> robru: cyphermox ...or anyone, how would one determine what the version of bzr is that's being used by the silo builders ?
[20:27] <cyphermox> kgunn: you'd need to ask ;)
[20:28] <cyphermox> or maybe look at the ppa logs
[20:29] <kgunn> mmm
[20:30] <cjwatson> I don't think bzr is run in a PPA-like environment
[20:30] <cjwatson> For the purpose of silo merging
[20:51] <asac> kgunn: thats what i just asked above. they take your bzr trunk, and merge stuff together, but dont publish that at the build stage; think thats what I hope the train could do; similar to the branches published when you hit merge and publish: https://code.launchpad.net/~ps-jenkins/+branches?memo=100&start=100
[20:52] <asac> kgunn: since we don thave that it seems, you probably have to look at the logs :/
[20:52] <robru> asac, you should fild a bug against cupstream2distro and assign it to didrocks ;-)
[20:53] <asac> robru: whats the jenkins URL again?
[20:54] <asac> the CItrain jenkins that is
[20:54] <robru> asac, you mean http://162.213.34.102/ ?
[20:54] <asac> right
[20:55] <asac> thats where the logs would be, right?
[20:55] <robru> asac, yeah, you can get links to the right jobs from the spreadsheet, and then the logs are listed on the left side of the page
[20:56] <asac> robru: hmm. where would the bzr branching/merging be done for the build prep?
[20:57] <robru> asac, in the build logs. depends per silo
[20:57] <robru> asac, each silo has a unique build job, so you have to know what silo you care about first, then look at it's build logs
[20:57] <asac> robru: right :) ... but at what stage?
[20:57] <asac> cant find anything in reconfigure and build
[20:57] <asac> ... bzr bd...
[20:57] <asac> does that pull the source auto?
[20:57] <robru> asac, what is it that you are looking for?
[20:58] <asac> robru: the bzr versions branched
[20:58] <asac> and merged
[20:58] <asac> those revs
[20:59] <robru> asac, ok, so lets say you are wondering about silo 3. go to the silo 3 tab in the spreadsheet. click on the 'build' button. see the logs on the left, most recent build is #85. click the blue dot, and you get this: http://162.213.34.102/job/landing-003-1-build/85/console
[20:59] <robru> asac, grep for 'Trying to merge' and it tells you the merge URLs with revisions pulled
[21:00] <asac> ic
[21:00] <asac> so the command is hidden :)
[21:00] <asac> good
[21:02] <robru> asac, yeah, it's not like a makefile where it echoes the commands it runs. it's a python script and it just logs certain details at certain points
[21:05] <robru> although I'm starting to think that a makefile would work better considering how many commands we run... ;-)
[21:10] <cyphermox> robru: is this still about the bzr version used for the silos?
[21:10] <robru> cyphermox, yeah
[21:10] <cyphermox> not revisions, correct?
[21:10] <robru> cyphermox, I'm pretty sure they are asking which MP revisions got branches for the most recent silo build job.
[21:11] <cyphermox> isn't that in the build logs?
[21:11] <robru> cyphermox, yes it is, I explain it in the scrollback
[21:11] <cyphermox> alright :)
[21:11] <cyphermox> if you need the actual bzr utility version; then you can ssh on the jenkins box to see it
[21:12] <cyphermox> it's 2.5.1-0ubuntu2
[21:12] <robru> cyphermox, yeah, that's what I thought they were asking at first but I don't think that's it
[21:44] <balloons> sergiusens, ping
[21:45] <t1mp_> plars: I just noticed that ubuntu-test-cases is using phablet-flash. Do you know that that is deprecated and ubuntu-device-flash should be used?
[21:46] <plars> t1mp: yes, we're aware. phablet-flash does still work, and the host system needed to be updated before patching this. I have an MP out to update to ubuntu-device-flash that is being reviewed and tested by another person right now. It worked in local testing, so I expect we'll probably merge it this evening
[21:47] <balloons> ogra_, cyphermox sergiusens, cjwatson perhaps? Can any of you help me in fixing a click package being built by cmake? I need it to include a plugin
[21:58] <sergiusens> balloons, which project?
[21:58] <balloons> sergiusens, filemanager. https://code.launchpad.net/~nskaggs/ubuntu-filemanager-app/fix-armhf-build/+merge/211621
[21:58] <sergiusens> you should just follow what reminders app does
[21:58] <balloons> sergiusens, yes but the plugin is external to the project
[21:59] <sergiusens> balloons, you need to include it; that's the click model
[21:59] <balloons> sergiusens, qtdeclarative5-nemo-qml-plugin-folderlistmodel
[22:00] <sergiusens> balloons, the plan with dpm was to include it in source
[22:00] <balloons> sergiusens, can I use find_package or find_library or ?
[22:01] <balloons> in the source would certainly help
[22:01] <sergiusens> balloons, you can use what I use in click-ready, you should notice a plugins.json in source
[22:01] <sergiusens> it's still bad though
[22:01] <sergiusens> and pull-lp-bin for it
[22:01] <sergiusens> or apt-get download it
[22:02] <sergiusens> but you need to add the ppa to your sources.list for the latter to work
[22:02] <balloons> ahh pull-lp-bin.py
[22:03] <balloons> sergiusens, so there's no really good way to do this apart from having the plugin inside the project?
[22:04] <sergiusens> balloons, nope; I talked to zoltan about having some 'plugin' repository and reusing that; but I guess they had no time yet to design and implement
[22:06] <balloons> ugh that's ugly. ok, so I will try and hardcode some hackery to make it build against a static copy of the plugin and push to get it into the project
[22:07] <balloons> ty sergiusens .. I reserve the right to ping you again, but this makes sense enough now I think
[22:25] <t1mp> plars: okay, nice
[22:40] <plars> bfiller: boiko: around?
[22:40] <bfiller> plars: yes
[22:40] <plars> bfiller: boiko: wondering if you saw the new crashes with dialer and messaging: http://ci.ubuntu.com/smokeng/trusty/touch/mako/261:20140325.3:20140304/7375/dialer_app/
[22:41] <bfiller> plars: argh, no
[22:41] <plars> I wondered if those might be related to the fix that recently went in
[22:41] <plars> bfiller: on the plus side, all the tests pass! :)
[22:46] <bfiller> robru: can I get a silo please for line 46
[22:47] <robru> bfiller, on it ;-)
[22:47] <robru> bfiller, ok, you got silo 9, please build
[22:48] <bfiller> robru: thanks!
[22:48] <robru> bfiller, you're welcome!
[23:43] <Saviq> popey, ogra_, re: blocking content based on geo, that's the smart scopes server's task