/srv/irclogs.ubuntu.com/2013/03/27/#ubuntu-devel.txt

TheMuso@pilot in00:04
=== udevbot changed the topic of #ubuntu-devel to: Ubuntu 12.10 released | Archive: Feature Freeze | Devel of Ubuntu (not support or app devel) | build failures -> http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and discussion of hardy -> quantal | #ubuntu-app-devel for app development on Ubuntu http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots: TheMuso
xnoxScottK: doko: I don't see a reason for that, either. Mind sharing your train of thought here?01:33
* ScottK didn't either.01:34
=== jono is now known as Guest2834
=== ironhalik_ is now known as ironhalik
TheMuso@pilot out03:59
=== udevbot changed the topic of #ubuntu-devel to: Ubuntu 12.10 released | Archive: Feature Freeze | Devel of Ubuntu (not support or app devel) | build failures -> http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and discussion of hardy -> quantal | #ubuntu-app-devel for app development on Ubuntu http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://bit.ly/lv8soi | Patch Pilots:
DigvijayUbuntuHello people05:17
DigvijayUbuntuI needed help understanding how the ubuntu-software-center code is layout05:17
DigvijayUbuntuany pointers on where i can start?05:17
sarnoldDigvijayUbuntu: I really love ctags, id-utils, and cscope05:19
DigvijayUbuntusarnold: i am a bit confused05:19
DigvijayUbuntuso you can help me?05:19
sarnoldDigvijayUbuntu: I can try :) hehe05:19
DigvijayUbuntuok greate =)05:20
DigvijayUbuntuwait lemme pm u05:20
pittiGood morning06:00
tvosspitti, good morning :)06:08
vibhavguten morgen06:08
tvossguten morgen :)06:15
=== MaJian is now known as BruceMa
KeithWHello folks. We just released mosh 1.2.4 and uploaded 1.2.4-1 to Debian. What is the appropriate way to respectfully request a sync to Ubuntu raring (universe) and a FFe to make the release?06:41
infinityKeithW: Point me at a 1.2.3 -> 1.2.4 changelog?06:58
KeithWinfinity: http://mailman.mit.edu/pipermail/mosh-devel/2013-March/000447.html06:58
infinityKeithW: Looks good to me, syncing.06:59
infinityOr... I would if LP knew about the version in unstable yet. :P06:59
infinityI'll sync when it does.07:00
infinityKeithW: FWIW, I would have called the mm:ss disconnect display a bugfix, not a feature.  Being offline for eighty bazillion seconds has always annoyed me. :P07:00
KeithW:-)07:01
KeithWThanks much!07:01
dholbachgood morning07:33
dholbachcan somebody from the desktop people reply to https://twitter.com/larsschuetze/statuses/316652779605225473 maybe? :)07:34
mardythomi: ping (need some help with autopilot)07:42
mardydholbach: hi! Do you know how to write autopilot tests?07:58
dholbachmardy, I'm afraid not - I wrote an autopkgtest once though :)07:59
mardydholbach: do you know anyone who could help me, who is currently online? It's a rather newbie-like question, I presume08:00
dholbachmardy, maybe just best ask the question?08:03
dholbachpitti, Mirv: ^ do you know much about autopilot?08:05
mardydholbach: let's try :-)  How do I wait till an application becomes visible, after calling start_app()?08:05
pittimardy: perhaps you can pick a particular widget and use Eventually() to wait until its visible property becomes tru?08:06
pittitrue08:06
* pitti only did some baby steps with autopilot as well, I'm afraid08:06
Mirvdholbach: I can run the autopilot tests of unity and tweak if something fails, but not that much. I'd take a look at unity autopilot tests for a basis.08:07
mardypitti: I'll try -- but I guess that I won't be able to get a handle to any widget, until the window is created08:07
Mirvmardy: Eventually seems to be used in Unity..08:08
mardyMirv: I tried that, but on what property? "user_visible" doesn't work ("Eventually is only usable with attributes that have a wait_for function or callable objects.")08:09
mardyfunny thing is that none of the tests in http://bazaar.launchpad.net/~ubuntu-testcase/ubuntu-autopilot-tests/trunk/files/head:/ubuntu_autopilot_tests/ check for the window visibility08:11
mardymaybe start_app() actually waits for the window visibility itself?08:11
=== smb` is now known as smb
=== zyga is now known as zyga-afk
dholbachseb128, salut - do you know who could reply to https://twitter.com/larsschuetze/statuses/316652779605225473?08:38
seb128dholbach, salut08:38
seb128looking08:38
seb128dholbach, you could reply? ;-)08:40
dholbachno08:40
dholbachI don't know the answer08:40
seb128no reason we can't if there are applications proving good enough08:40
seb128I don't see qml equivalent/replacements of our desktop core apps atm though08:41
seb128but I'm pretty sure that will happen in the futur, hard to predict when and what apps though08:41
thomimardy: Hi08:41
dholbachseb128, ok08:42
thomimardy: did you get your problem resolved?08:42
mardythomi: hi! Just trying to write some autopilot tests for Online Accounts, and I've a few questions :-)08:43
mardythomi: does the start_app() actually wait for the window to be visible, before returning?08:43
thomimardy: hmmm, let me check08:44
mardythomi: none of the tests in http://bazaar.launchpad.net/~ubuntu-testcase/ubuntu-autopilot-tests/trunk/files/head:/ubuntu_autopilot_tests/ is waiting for that, they all seem to assume the window is already available08:44
thomimardy: it doesn't check for window visibility inside that method, no. I belive some of the tests have their own way of checking that the application has been initialised08:45
thomiwhich would be the correct way to do it08:45
thomisince an application window can be visible, but not initialised yet08:45
thomimardy: so Ideally you'd start the app (via launch_test_application), and then read application state to ensure that it's been set up correctly08:46
mardythomi: OK, the second question is: are start_app() and launch_test_application() completely separate? or can I get an ApplicationProxyObject froma BamfApplication object?08:47
thomimardy: they are completely separate, and no, you can't. launch_app just starts the application normally, whereas launch_test_application does the "magic glue" that allows autopilot to introspect the application08:48
mardythomi: well, I guess that if I use launch_test_application() directly, the question above doesn't even apply08:48
mardythomi: OK, now it's clear. Thanks! :-)08:48
thomimardy: lots of people make that mistake, and we're making it clearer in newer versions of autopilot :)08:48
thomithe launch_app methods are there if you want to test how your application interacts with opther apps.... which is really useful if you happen to be testing a desktop shell =)08:49
thomiautopilot is showing it's lineage :)08:49
thomimardy: It's getting late here, so I'm going AFK. If you run in to any other problems, feel free to dop me an email: thomi.richards@canonical.com08:52
* thomi waves08:52
mardythomi: mmm... just got one more question :-) Are my "print" statements eaten up somehow?08:54
mardythomi: actually, it seems that launch_test_application() never returns08:58
mardythomi: nevermind, I was missing libautopilot-gtk :-)09:08
vrodicwhy is ubuntu insiting on trying to ship outdated versions of mesa. 13.04 only has 9.0.3, but drivers in 9.2 are much better09:12
=== ckpringle_ is now known as ckpringle
=== schmidtm_ is now known as schmidtm
=== ckpringle_ is now known as ckpringle
cjwatsonseb128: we seem to have a gnome-settings-daemon that breaks the current version of indicator-datetime, and no new indicator-datetime - do you know what's up with the latter?09:36
cjwatsonlooks like policykit-desktop-privileges and gnome-settings-daemon aren't currently coinstallable so image builds are failing as a result09:36
seb128cjwatson, yeah, cyphermox was supposed to autoland indicator-datetime to go with the other bits and didn't for some reasons :-(09:40
seb128cjwatson, I'm looking at it09:40
cjwatsonseb128: thanks - can you let me know when it's landed so I can retry image builds, please?09:42
seb128cjwatson, will do, sorry about the issue09:42
seb128cjwatson, I overlooked that britney was not checking for installability to migrate stuff, when I uploaded I though it would stay in proposed until the breaks were resolved09:43
cjwatsonseb128: it checks for installability, but not coinstallability of arbitrary sets09:51
cjwatsonseb128: (which gets into the seriously NP-complete side of things, never mind being exception city :) )09:51
seb128cjwatson, was there any discussion about checking for installability of ubuntu-desktop?09:51
pittiisn't that pretty much what daily live fs builds do?09:52
seb128or (*)ubuntu-desktop09:52
pitti(installing ubuntu-desktop and telling us about failures)09:52
=== ckpringle_ is now known as ckpringle
seb128pitti, I was talking about plugging the result of that check in britney's migration09:52
pittitricky; britney would take ages with that, I guess09:53
cjwatsonseb128: we do check for installability of ubuntu-desktop09:53
seb128pitti, we broke daily iso with an uncomplete transition09:53
cjwatsonseb128: we don't check for installability of the ubuntu-desktop *task* (i.e. including all Recommends)09:53
cjwatsonyou've fallen into the crack there09:53
cjwatsonYou can tell this because http://people.canonical.com/~ubuntu-archive/testing/raring_probs.html doesn't show ubuntu-desktop as uninstallable09:54
cjwatsonIt might be worth generating some kind of per-task virtual packages that depend on everything in each task09:55
cjwatsonThough we'd have to be careful about when the task changes09:55
cjwatsonOr we could get proposed-migration to promote Recommends to Depends selectively or something ...09:56
cjwatsonAll seems fairly hackish but I suppose possible09:56
seb128yeah, I'm not sure what is best either...09:57
seb128cjwatson, I'm going to do a manual upload for indicator-datetime, utah seems to be out of order on the daily builds09:57
seb128e.g http://10.97.0.1:8080/job/ps-indicators-autopilot-release-testing/183/label=autopilot-nvidia/console09:57
cjwatsonseb128: got it, thanks10:00
=== tkamppeter_ is now known as tkamppeter
mardypitti: any idea how I can introspect a GtkTreeView contents in autopilot? I don't see the "model" property there...10:12
pittimardy: no, I don't; unfortunately autopilot-gtk is rather limited right now10:12
mardypitti: OK, tough luck :-) I'll file a bug in the meantime, let's see how it goes...10:13
seb128cjwatson, indicator-datetime uploaded/built everywhere, I guess it just need to be published and to be moved to raring proper with gnome-settings-daemon and we should be ok for the iso10:18
seb128cjwatson, we will need systemd-services/systemd-shim promoted as well (MIR are approved, we just need to promote the binaries)10:19
=== iainfarrell1 is now known as iainfarrell
cjwatsonseb128: ok, cool, I'll keep an eye on the hints10:28
cjwatsonand I'll go and poke those MIRs10:28
dholbachpete-woods, AFAICT sphinx-voxforge-en looks good - it might be worth to put your comment (#10) from the bug report into either debian/copyright or debian/watch or maybe both to make it clear where the files are coming from and how they are constructed... also in case somebody else might want to update/fix the package and should not know how to go about it10:33
dholbachpete-woods, if you reupload the source package (or somewhere else) with that fix, you should just have to update the .dsc and .debian.tar.gz file - I can then upload it for you10:34
pete-woodsdholbach: thanks for the review. I will put some more information in the debian/watch file - at the moment it doesn't mention the training tool I write10:35
cjwatsonseb128: I can't find the MIR for systemd-services - can you point me at it?10:35
pete-woods*wrote10:35
seb128cjwatson, https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/115218710:35
ubottuLaunchpad bug 1152187 in systemd (Ubuntu) "[MIR] systemd" [Undecided,Fix released]10:35
seb128cjwatson, comment #12 has the ack10:35
seb128cjwatson, and #14 from security10:37
cjwatsonseb128: oh, it's closed, ok10:48
cjwatsonseb128: moved all those to main now10:51
seb128cjwatson, thanks a lot10:51
cjwatsonseb128: looks like it's been promoted now10:56
seb128cjwatson, good, I just got the launchpad emails that indicator-datetime and gnome-settings-daemon moved to raring10:59
pete-woodsdholbach: did you see my comment on the sphinxtrain package? (https://bugs.launchpad.net/ubuntu/+bug/1144015)11:55
=== MacSlow is now known as MacSlow|lunch
ubottuLaunchpad bug 1144015 in Debian "[needs-packaging] New package sphinxtrain" [Unknown,New]11:55
pete-woodsit's not installable in the state in the bug tracker11:55
pete-woodsI can't update it myself, as you already have extra changes in there11:55
pete-woods(also have learned by 0ubuntu1 lesson :) )11:56
vibhavDo we need a FFE for syncing ncpfs?12:17
vibhavhttp://packages.debian.org/changelogs/pool/main/n/ncpfs/ncpfs_2.2.6-9/changelog , if anyone wants to read the changelog12:19
=== _salem is now known as salem_
vibhav(the latest release fixes from critical bugs though)12:19
tumbleweedvibhav: the latest release has also been synced12:20
tumbleweedit's from november12:21
vibhavwierd, ubuntuwire still shows it.12:22
vibhav(must be update lag)12:22
tumbleweedprobably because it's stuck in -proposed12:22
vibhavah12:23
* vibhav takes a look12:23
vibhavyes, it is12:23
vibhavtumbleweed: thanks12:23
=== mmrazik is now known as mmrazik|afk
tumbleweedvibhav: in fact, IIRC, I even commented about why it's stuck on the rcbugs page on ubuntuwire12:24
tumbleweedit looked non-trivial, but we should unstick it12:25
vibhavtumbleweed: yeah, I saw it12:25
mlankhorstoops12:44
mlankhorstcan someone delete the xxv-modesetting upload I did? forgot to fixup distro12:44
pittimlankhorst: where did you upload it to?12:45
mlankhorstprecise-proposed12:45
pittimlankhorst: seems someone already removed it, or the upload was rejected12:46
pittihttps://launchpad.net/ubuntu/precise/+queue?queue_state=112:46
mlankhorstprobably in new12:47
mlankhorstyeah12:47
mlankhorstI don't think precise had the modesetting driver :)12:54
=== timchen`` is now known as timchen119
=== MacSlow|lunch is now known as MacSlow
cjwatsonmlankhorst: rejected from NEW13:15
mlankhorstty13:16
ogra_xnox, for your switch back to metacity, dont we also need to seed it again ?13:19
xnoxogra_: don't think so. compiz uses libmetacity to draw decorators. so 2/3 binary packages from src:metacity was still on the cd anyway.13:20
ogra_just wanted to make sure ...13:21
xnoxogra_: and today's image wasn't build yet (due to respin with g-s & stuff) so will be testing once it lands.13:22
ogra_++13:22
leagrisHey, hello13:33
leagrisThemuso, I wanted to reach you if you have a little bit time to talk about Ubuntu low sight accessibility13:34
cyphermoxseb128: tests failed, the autopilot vms exploded overnight, that's why it didn't land13:36
seb128cyphermox, ok, would have been nice to do a manual upload to unbreak installability of the desktop/cds13:37
seb128cyphermox, but no worry, we sorted it out this morning13:37
cyphermoxwould have been, but you asked and I was no longer around13:37
cyphermoxsorted it how? is there something to merge back into indicator-datetime now?13:38
leagrisFor all, I have strong concerns if Compiz E-Zoom could dissapear in the next ubuntu releases (while dropping Xorg). EZoom is the best suited screen magnifier for my 10% signtness. Comparing to Gnome Orca magnifier or its KDE counterpart or even the closed source world. Ezoom is responsive, smooth, easy and work anywhere even with full screen applications.13:38
seb128cyphermox, I did a manual upload of lp:indicator-datetime/13.04 to raring13:38
cyphermoxseb128: ok!13:39
seb128cyphermox, so I guess yes, we will need that merged back (or overwritten, I don't care much about the changelog entry)13:39
dholbachpete-woods, I'm not sure what this means......... is the bug fixed already?13:39
cyphermoxseb128: yeah, I'll just merge it and it's going to be fine13:39
seb128cyphermox, thanks13:39
seb128cyphermox, the utah guys looked at their issue this morning and retried the tests, which failed because systemd-services was not promoted yet, I retried again13:40
pete-woodsdholbach: sorry, I got my wires crossed and contacted the wrong person about it :$13:41
dholbachpete-woods, ok... no worries :)13:41
cyphermoxit failed again in this case :)13:41
seb128cyphermox, indeed :-(13:44
seb128one day utah will work for more than 2 runs in a row13:45
ogra_wishful tinking :)13:45
=== wedgwood_away is now known as wedgwood
=== kentb-out is now known as kentb
=== wedgwood is now known as wedgwood_away
=== wedgwood_away is now known as wedgwood
=== wedgwood is now known as wedgwood_away
=== wedgwood_away is now known as wedgwood
=== zyga-afk is now known as zyga
=== davidcalle_ is now known as davidcalle
=== rickspencer3_ is now known as rickspencer3
mlankhorstslangasek: can I reassign bug 982889 to you?18:05
ubottubug 982889 in lightdm (Ubuntu) "X trying to start before plymouth has finished using the drm driver" [Critical,Confirmed] https://launchpad.net/bugs/98288918:05
slangasekmlankhorst: if you wish... but I'd also be happy for you to follow through on it :)18:05
mlankhorstI don't trust myself touching upstart on other computers18:07
=== psivaa is now known as psivaa_afk
=== rickspencer3_ is now known as rickspencer3
=== xxiao is now known as ausxxh
slangasekwhy is update-manager upgrading packages that it's not showing me in the details list?19:16
ScottKIt's an over achiever?19:16
slangasek... specifically, packages that I'm hacking on locally and want to hold back but have not done a dpkg hold on, expecting update-manager to let me unselect them19:16
xnoxslangasek: talk to mpt. It's the new design saying that "technical packages" are lumped together into "Core OS" item.19:17
* xnox looks for the wiki / definition page.19:17
slangasekxnox: there's a "Base OS" pulldown that lets me see lots of packages... but not the ones I care about19:17
slangaseklibdbus - listed.  libdecoration0 - not listed19:17
slangasek"Ubuntu base", it's called here19:18
xnoxslangasek: https://wiki.ubuntu.com/SoftwareUpdates#Expanded_presentation_of_updates19:18
* xnox upgrades using cmd-line only =/19:19
* xnox ponders if I should be using the "normal" updates method19:19
slangasekxnox: yes; I assert that, even within the context of that design spec which I disagree with, the current behavior is buggy because I have no way at all to see and control the complete list of updates being installed19:23
=== racarr is now known as racarr|lunch
blitzkrieg3does anyone have a good way to test the timezone picker in ubiquity?19:43
=== Lutin_ is now known as Lutin
=== G4MBY is now known as PaulW2U
=== salem_ is now known as _salem
=== Tonio_ is now known as Tonio_aw
xnoxblitzkrieg3: please explain more? to test what?22:22
xnox(it has geoip detection, visuals/timezone highlights, setting correct timezone for the target user, offline geoip-less mode)22:23
xnoxit's actually a few things the meat of the timezone picket is libtimezonemap which is also used by gnome-control-center as the time picker.22:23
KeithWI am impressed that you guys synced mosh 1.2.4 and correctly reverted the appropriate commit to fix the FTBFS on armhf with no drama!23:00
=== wedgwood is now known as wedgwood_away
coderangerAnyone know if apt (10.04+) actually still uses the MD5 and SHA1 fields in repos?23:01
coderangerGuessing it is just crufty old stuff and everything hopefully uses the sha256 sums now, but I've not looked into the apt source23:01
sarnoldcoderanger: https://bugs.launchpad.net/ubuntu/+source/apt/+bug/109873823:01
ubottuLaunchpad bug 1098738 in apt (Ubuntu) "apt-get source only checks md5 hashes in Sources files" [High,In progress]23:01
coderangersarnold: I haz a sad now23:01
coderangerWill include all of them then, thanky23:02
infinityKeithW: Oh, I was going to make drama for you, but you weren't online. :P23:02
KeithWinfinity: Haha, well, impressive! We just pushed 1.2.4a to Debian to do the same thing.23:03
infinityKeithW: Didn't feel like fixing the broken test code instead?  The revert was just the path of least resistance for me. :P23:04
KeithWWe already had our own end-to-end test of OCB with the same test vectors in src/tests, so it didn't seem worth it. The real problem was that the built-in ocb.cc test doesn't even give a bad return code on failure, so it wasn't even working for us as an automated test!23:04
infinityKeithW: Hah.  Smooth.23:05
infinityKeithW: Alright, I'll resync 1.2.4a-1 over my upload later, then.  After I've seen it build on all the Debian buildds.23:05
=== Tonio_aw is now known as Tonio_
infinityKeithW: (Which I should have checked before doing the sync for you before, mea culpa)23:05
KeithWOk, I think your existing 1.2.4 is fine (the only relevant changes in 1.2.4a warnings on RHEL 5 and ARM/MIPS/s390 that failed the build), but also fine if you want to sync it.23:06
KeithWEr, the only changes in 1.2.4a are to fix build warnings.23:07
infinityOh, I'm sure mine's fine, I'd just sync it to get it off my plate, so it autosyncs correctly again later.23:07
infinity(We don't autosync packages with Ubuntu changes, for obvious reasons)23:08
KeithWOh, ok. Sounds good.23:08
=== Tonio_ is now known as Tonio_aw
=== Tonio_aw is now known as Tonio_
=== Tonio_ is now known as Tonio_aw
xnoxMirv: when are you planning to upload qt5 again? I really want the Gtk theming fix in, but there are a few more things stagged in the bzr branch.23:30
ScottKxnox: What and is it upstream?23:31
xnoxScottK: don't panic. We missded -gtkstyle & libgtk2.0-dev build-dep in qt5 package vs qt4 package, hence gtk theming support didn't build and right now qt5 hello world app looks like it's from 198723:32
ScottKOK.23:32
ScottKWhat are the few more things?23:32
sarnold(can we keep that as an option? :)23:32
=== Tonio_aw is now known as Tonio_
xnoxScottK: see unreleased stanza: https://bazaar.launchpad.net/~kubuntu-packagers/kubuntu-packaging/qtbase-opensource-src/view/head:/debian/changelog23:34
xnoxScottK: also do you have highlight on my nick or just on "qt" ? =)23:34
ScottKJust happend to notice.23:41
ScottKOK.  Not too scary.23:42

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!