=== bradm_ is now known as bradm === LoganCloud is now known as Guest49309 === Guest49309 is now known as LoganCloud [05:42] Good morning === freeflying_away is now known as freeflying === iahmad is now known as iahmad|afk [08:01] good morning === oSoMoN_ is now known as oSoMoN === doko_ is now known as doko === iahmad|afk is now known as iahmad === Sweetsha1k is now known as Sweetshark === _salem is now known as salem_ === seiflotfy_ is now known as seiflotfy === apachelogger is now known as DarthVader === DarthVader is now known as hsitter === jibel_ is now known as jibel [14:06] pitti, hmm, so i'm maintaining my cdimage branches on a precise desktop ... it uses python-mock ... and apparently uses tests that use the trusty python-mock, do you plan to SRU this ? [14:07] (so i can go on testing the branch on an LTS) [14:07] ogra_: err, sorry, context/ [14:07] ? [14:07] http://paste.ubuntu.com/6508783/ [14:07] ogra_: i. e. SRU what? [14:07] thats what i get when running the self tests under precise [14:07] pitti, python-mock changes [14:08] ogra_: so apparently python-mock 0.7 doesn't yet know this syntax [14:08] right, thats what i mean [14:08] ogra_: if you want them to work under precise, I suggest you use the 0.7 API (which hopefully is forward compatible)? [14:08] do you plan to backport such stuff for people developing on the LTS [14:08] ogra_: why me, specifically? [14:08] also, we don't generally SRU new versions [14:08] isnt python-mock your child ? [14:08] we can certainly ask for a backport [14:09] ogra_: no, it's not; I guess you mean dbusmock :) [14:09] i know how to work around that, but was wondering about others that use LTS versions for developing tip branches of newer stuff [14:09] ogra_: anyway, python-mock is fairly isolated, so we can build it as backport or put it into a PPA or something if needed [14:12] ogra_: I've hit the same issue, FWIW. But I didn't think that new features were SRU-able in this way. Annoyingly this means that I can't run tests as part of a package build for a precise target. [14:12] well, you can run them in a trusty chroot [14:13] just forces you to maintain two branches or to bind mount it into the chroot or some such [14:13] or just change the tests to split the command into two, and set mymock.return_value explicitlY? [14:13] I can, but that's no good when the final target is the cloud archive on precise. Though we may be able to backport python-mock there; I'll have to check. [14:13] (that syntax works with older p-mock) [14:14] that seems like the least painful approach [14:14] it's indeed nice to be able to run tests on precise to ensure that all deps are there, and the actual code doesn't use anything from newer pythons, etc. [14:16] it requires that upstreams have that self discipline though === dholbach_ is now known as dholbach === barry` is now known as barry_ === barry_ is now known as barry [14:51] rsalveti, ping [14:52] Saviq: pong [14:52] rsalveti, hey man, I've a .crash with... a LOT of question marks [14:53] rsalveti, I'm starting to think it's crashing on android side - do you have any pointers on getting some useful data out of there? [15:01] cjwatson: Can you please check why wxwidgets3.0 still isn't getting auto-synced? The conflicting binary package has been dropped from wxwidgets2.8 in -proposed. [15:02] Saviq: sure, https://wiki.ubuntu.com/Touch/Core/UbuntuDebugAndroid [15:03] Saviq: easier if you have the android build locally, let me know if you don't, then I can provide the lib with debug symbols [15:03] once you check it with maps [15:03] rsalveti, it's the first time I'm doing it, so no, I don't have anything [15:04] Saviq: ok [15:04] rsalveti, it's unity8... it's going to have everything in its maps... [15:04] rsalveti, any way to reduce the scope? [15:05] Saviq: but checking the stack trace, you should be able to find the first maps that brings the ??? [15:05] and the related library [15:06] rsalveti, here's the .crash, what should I be looking at? http://paste.ubuntu.com/6509647/ [15:07] let me check [15:07] mitya57: i filed a bug about it. basically wx-common binaries need removal, and a sync forced. [15:07] mitya57: cause at the moment wx-common binary has "ubuntuX" suffix version in -release pocket. And wxwidgets2.8 cannot migrate without new wx-common from 3.0, which can't be synced because of the first reason =) [15:08] I expected the autosyncer to look at -proposed [15:08] mitya57: so someone with archive admin powers needs to stab it =) i hope infinity or cjwatson will get around to doing it. [15:08] Saviq: the stack seems to be in the ubuntu side, maybe opening the core dump with gdb to check [15:09] xnox: OK [15:09] Saviq, we got a new glib exactly that day the install is from [15:09] mitya57: well it can't, because -proposed is an overlay (e.g. there is no way to tell between wx-common binary dropped in -proposed, vs wx-common is at *ubuntuX version at the moment) [15:09] well, not exactly ... but its definitely using the new one there [15:09] mitya57: -proposed & -release are merged and then compared. [15:10] mitya57: from archive point of view there "wx-common is fully migrated to release pocket" =) [15:12] Right. [15:27] ogra_, shall I try in the previous image then? [15:28] well, might make sense to do some comparison [15:28] Saviq, http://people.canonical.com/~ogra/touch-image-stats/20131129.2.changes [15:28] ogra_, will do, thanks [15:28] thats the respective change set [15:30] rsalveti, yeah, there's just no more symbols I can think to install to get anything - it's all ??s [15:30] rsalveti, in gdb, too [15:33] Saviq: have the dump somewhere I can take a look? [15:40] If I have uploaded an earlier version of my application and thus I had a sponsor for it, should I contact him for a new version of the app and a new library not in the repos which is a dependency of this new version of the application? === freeflying is now known as freeflying_away [16:01] mitya57: interesting I got accept from wxwidgets3.0 source =/ it's now in bin new. [16:01] mitya57: \o/ [16:05] xnox: \o/ [16:31] mhall119: do you know who manages irc cloaks for ubuntu members? [16:31] zyga: #ubuntu-irc if I remember right [16:59] are there any ubuntu developers from ukraine present here? [17:08] cjwatson: can you have a look at this merge request when you have time? https://code.launchpad.net/~penghuanmail/ubiquity/lp.1197220/+merge/195712 [17:11] zyga: I'm not aware of any Ukrainian Ubuntu developers. === bfiller is now known as bfiller_afk === _salem is now known as salem_ [19:03] balloons: DMB time. [19:03] Sorry [19:03] barry: DMB time. [19:22] doko: I don't suppose you're aware of some sort of porting guide that allows me to port an application from graphviz 2.26 to 2.34 [19:24] doko: currently stuck on http://paste.ubuntu.com/6510818/ [20:15] Does Ubuntu place mount and umount into /bin? [20:24] Nevermind. happyaron helped me. Long story short, I am writing an upstream patch and I wanted to be certain that it worked on Ubuntu. === salem_ is now known as _salem [21:16] anyone have an example of a service that does not have an upstart job ? [21:16] ie, sysvinit only ? === Ursinha_ is now known as Ursinha [21:22] smoser: I think maybe apache & postgres? (I would have to check the latest packages though) [21:22] smoser: apparmor profiles, x11 common - in the default install. [21:23] JanC, apache2 seems good. [21:23] fwiw, this is https://bugs.launchpad.net/ubuntu/+source/sysvinit/+bug/1257036 [21:23] Ubuntu bug 1257036 in sysvinit (Ubuntu) "services started with eatmydata should remove eatmydata by default" [Medium,Confirmed] [21:45] hm.. [21:45] so that bug above, i assumed was present [21:45] but it seems that most services clean their own environment [21:45] (tested so far postgress, apache2, postfix) [21:46] if i'm going to make cloud-init install packages by default with libeatmydata [21:46] should i fix this even thought it seems that most packages don't need the help ? [22:02] cjwatson: I just uploaded a merge of LTSP and a sync for LDM, I think you were technically TIL for those two because of small arm64/d-i fixes, I hope you don't mind (I wrongfully assumed that I was TIL on both ;)). [22:05] smoser: hm, would cloud-init leak LD_PRELOAD to e.g. juju / user logged-in session? ( i guess not, since cloud-init is managed by upstart). If it doesn't leak it, no need to fix "service" command since it's environment is clean. [22:06] invoke-rc.d should be enough. [22:06] xnox, you're right in that service is enough in that case. [22:07] but fixing service *also* is a more genericly safe path. [23:14] slangasek, tyhicks: apparmor 2.8.0-2 uploaded to Debian with multiarch and dh(1). [23:14] \o/ [23:15] kees: nice! :)