2012/12/08/#ubuntu-devel.txt

[00:00] <jtaylor> ._.
[00:00] <achiang> nod. looking at changelog in ubuntu, last time we sync'ed was precise
[00:00] <infinity> jtaylor: Did you just want it synced wholesale, or do we have a delta you want to merge/preserve?
[00:00] <jtaylor> I didn't look at it yet
[00:00] <jtaylor> I only fixed two bugs for quantal, I have no history in the package
[00:01] <achiang> perhaps i could persuade janimo to do so, he's made a few uploads...
[00:01] <achiang> s/a few/one/
[00:02] <infinity> achiang: You could possibly talk me into it.
[00:02] <infinity> achiang: But it won't be this evening.
[00:02] <jtaylor> if you do, I'd also like a backport to precise ;)
[00:02] <infinity> achiang: Want to give me a URL to the svn commit of the bit you're interested in, rather than the bug with your patch?
[00:02] <achiang> infinity: yes, you've been a good robot and you need to power down for a few hours while we change your oil
[00:04] <achiang> infinity: let me check what is in unstable. my patch touched an internal API that changed between our version in Q and what is in svn head (or trunk or whatever they call it)
[00:04] <achiang> infinity: so i have 2 patches, one that works w/ubuntu and the one that landed upstream
[00:04] <infinity> Unstable is 3.8.1
[00:04] <achiang> (just trying to save you some time)
[00:04] <infinity> achiang: dget http://http.debian.net/debian/pool/main/v/valgrind/valgrind_3.8.1-1.dsc
[00:05] <achiang> thanks, that's handy
[00:05] <infinity> ('pull-debian-source valgrind' would do the trick too, I'm still trying to train my fingers to do that)
[00:06] <ScottK> So.  I have this bzr branch that I've pushed to LP.  How to I make a merge request out of it?
[00:07] <infinity> ScottK: Wherezit?
[00:07] <ScottK> https://code.launchpad.net/~kitterman/+junk/ubiquity-encryptcheckbox
[00:07] <infinity> ScottK: You need to use predictable namespacing to make it work.  No junk.
[00:07] <ScottK> Sigh.
[00:07] <ScottK> OK.
[00:07] <infinity> ScottK: bzr push lp:~kitterman/ubiquity/my-fix
[00:07] <ScottK> OK.
[00:07] <infinity> I believe.
[00:07] <achiang> infinity: ok, unstable has the new API, so the version from svn should Just Apply(tm).
[00:08] <infinity> I never use this feature. :P
[00:08]  * ScottK tries.
[00:08] <infinity> ScottK: If the namespace matches, then you magically get LP sorting out ancestry and offering you the MP option.
[00:08] <ScottK> Right.  It made a stacking branch this time, so I think it got it.
[00:08] <ScottK> Thanks.
[00:09] <infinity> achiang: Shiny.  URL me up, I'll drag it out of backcroll.  Food time.
[00:09]  * achiang attempts to find a url to valgrind's svn
[00:09] <xnox> ScottK: do i have something to review and merge =))))?!
[00:10] <infinity> achiang: A pointer to the repo and revision number works too, doesn't need to be a silly web view. :P
[00:10] <achiang> infinity: oh, heh. ok. sorry, "URL" confused me. ;)
[00:10] <ScottK> Depends on if you want to make sure I didn't screw up a Qt .ui file change.
[00:10] <achiang> infinity: email ok or irc preferred?
[00:11] <ScottK> xnox: You should have mail.
[00:11]  * xnox waits for launchpad to generate review diff and finally send mail
[00:12] <ScottK> The .ui file presented correctly in Qt Designer, so I imagine it's correct ....
[00:23] <xnox> ScottK: i'm not entirely familiar with qt format. But I do wonder if this option is displayed correctly in either gtk & kde with a Right-To-Left locale.
[00:23] <ScottK> Good question.
[00:24] <ScottK> Would you leave a comment in the merge proposal.
[00:24] <ScottK> Perhaps Riddell will know (I marked him down as reviewer since it's Qt .ui stuff)
[00:24] <xnox> ScottK: if it doesn't work I open a new bug.
[00:26] <xnox> ScottK: I have been told many times RTL "just works" but I'm not conviced.
[00:29] <slangasek> marrusl: ping
[00:36] <marrusl> slangasek, hey.
[00:36] <slangasek> marrusl: hi - did you see my follow-up questions on gdbM?
[00:36] <slangasek> gdb
[00:36] <slangasek> IWFM
[00:36] <marrusl> slangasek, yes.  thank you kindly.  just haven't had a chance to catch up with it.
[00:36] <marrusl> :)
[00:37] <slangasek> ok cool
[00:37] <slangasek> just making sure you'd seen
[00:37] <slangasek> from my POV, we should be able to consider the SRU verified
[00:38] <marrusl> slangasek, I won't argue with that.  I have a feeling I know what I'm missing.  but I need to go back.
[00:38] <marrusl> technically, well, I take your word for it!
=== slank is now known as slank_away
[01:40] <stokachu> slangasek: i think we should mark that gdb one verified as it fulfills what the actual issue is being reported
[01:42] <slangasek> stokachu: yep, done
[01:42] <stokachu> cool man, thanks!
[02:01] <hallyn> slangasek: no, i did rename the upstart job.  in .dfsg.5.
[02:05] <hallyn> ah, and seems to be right now.  at least it's shipped with qemu-system.*.deb.
=== cpg is now known as cpg|away
[03:47] <puff> Evening, I'm looking for the proper channel to discuss something odd that happened when I upgraded from 11.10 to 12.4, to see if I should file a bug report.
=== cpg|away is now known as cpg
[05:11] <slangasek> jelmer: fyi your samba4 sru .changes looks like it was generated on Debian and is missing the bug closure field; also, the setoption conversion from perl to python doesn't look like an SRU-appropriate minimal change.  I'm rejecting the package from the queue; if there's a reason setoption needs to be converted to python instead of just adding a dep on perl-modules, please document this in the appropriate bug report, and in either case pl
[05:11] <slangasek> ... with a correct .changes
[05:38] <mitya57> Is there a way to generate correct .changes on Debian?
[05:40] <slangasek> mitya57: set DEB_VENDOR
[05:53] <mitya57> thanks slangasek
[06:38] <Monotoko> is there any particular reason that the amazon search isn't run through HTTPS?
=== trijntje_ is now known as trijntje
[09:53] <mlankhorst> slangasek: yay, just need to figure out how two make mesa -dev packages coexist now and xorg package approval and then it should just be maintenance:)
=== doko_ is now known as doko
=== bdrung changed the topic of #ubuntu-devel to: Ubuntu 12.10 released | Archive: Open | Dev' of Ubuntu (not support or app devel) | build failures -> http://qa.ubuntuwire.com/ftbfs/ | #ubuntu for support and dicussion 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:
=== cpg is now known as cpg|away
[11:29] <vibhav> If a new version of a package in Debian ONLY introduces a Ubuntu change is merging it useless
[11:29] <vibhav> ?
=== lifeless_ is now known as lifeless
[11:48] <israeldahl> xnox: are you around?  I wondered if you had any links to fixing those lintian warnings
[11:53] <vibhav> xnox: Do you mean https://wiki.ubuntu.com/UbuntuDevelopment/BugFixingInitiative
[11:53] <vibhav> We have some lintian warnings there
[11:54] <israeldahl> vibhav: xnox was helping me yesterday and noticed the program had some lintian warnings that needed fixed, so now I want to fix them since I have to re-merge the code anyway
[11:55] <israeldahl> vibhav: do you know much about lintian warnings?
[12:59] <ubuntu-tester> Hi people, is there special channel for Unity dev?
[13:00] <ubuntu-tester> or where can I reach Unity team?
[13:02] <ubuntu-tester> found it, sorry for bothering
[13:04] <israeldahl> ubuntu-tester: no bother
=== yofel_ is now known as yofel
=== nerd is now known as database
=== iulian is now known as Guest54287
[16:32] <ScottK> jamespage: Thanks again for fixing sikuli.  There is an armhf specific FTBFS for sivp that also looks Java'ish.  Could you have a look at that as well?  That's the last blocker for the opencv transition.
=== database is now known as developer
=== Nisstyre_ is now known as Nisstyre
=== developer is now known as unix
[17:29] <OdyX> tkamppeter_: would it be possible to have the printers supported by splix be shown in the Openprinting database ?)
[17:52] <slangasek> mlankhorst: nice!  (well, at least until 6 months from now when we have to do the same thing for the raring stack ;)
[18:08] <mlankhorst> why do you think I used a script :p
[18:08] <mlankhorst> no way I'm going to keep track of all the changes by hand!
[18:26]  * ScottK wonders how many gcc uploads there will be this weekend to ensure powerpc never catches up ...
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
[19:30] <kees> i want syncpackage to work without gnome-keyring
=== cpg|away is now known as cpg
[19:55] <SpamapS> kees: s/syncpackage/launchpadlib/
[19:55] <SpamapS> kees: actually if there's no gnome keyring it will use the more generic keyring thing
[19:57] <SpamapS> kees: but yeah, its annoying when it finds a gnome keyring but cannot use it because you're actually just ssh'd in
=== jasoncwarner_ is now known as jasoncwarner
[20:07] <ricotz> hallyn, hi, maybe you are familiar with that problem http://paste.debian.net/plain/215121 -- this is raring armhf ppa builder failing, always on setting up libgtk-3-0
=== korn_ is now known as c_korn
[20:16] <israeldahl> dpkg-source: error: add data/themes/default/icon.png in debian/source/include-binaries if you want to store the modified binary in the debian tarball      anyone know what include-binaries is?
[20:18] <israeldahl> do I have to tar up the source and use that or use a commit?  anyone know where i need to go?
[20:18] <ScottK> israeldahl: Start with man dpkg-source.  Look in the section on format version 3.
[20:21] <israeldahl> ScottK: thanks but what or where is debian/source/include-binaries
[20:22] <ScottK> It would be in the debian directory of the source package.  You may have to create it.
[20:23] <israeldahl> ScottK: Thanks is it a text file that I put file paths into or a directory?  The only thing there is format
[20:24] <israeldahl> I don't mean txt file, sorry... I just mean a file with txt in it
[20:27] <ScottK> include-binaries is a vile
[20:27] <ScottK> file
[20:27] <israeldahl> ScottK: OK, thank you and I just put a path like data/themes/default/icon.png i suppose
[20:28] <ScottK> IIRC, yes.
[20:30] <israeldahl> Thank you immensely!!
[20:35] <israeldahl> ScottK: it worked to get me past the error!  I appreciate your help very much!
[20:35] <ScottK> You're welcome.
[20:52] <hallyn> ricotz: sorry, no.
[21:07] <ricotz> hallyn, i see, since the version the builders are running is older than the current raring one, maybe this is something which got fixed, and i am hoping you can pass that into the right direction
=== tkamppeter_ is now known as tkamppeter
[21:10] <hallyn> ricotz: any chance you could file a bug on that?
[21:11] <hallyn> ricotz: was that an update on an arm box?  an error during a pkg build on builder?
[21:12] <hallyn> multilib pkg upgrade on x86?
[21:12]  * hallyn bbl
[21:12] <ricotz> hallyn, this is building on an official armhf ppa builder
[21:36] <hallyn> ricotz: yikes.  can you paste a link to the full build log on launchpad here?  i'm not sure who supports the arm builders  (stgraber or infinity?)
[21:40] <cjwatson> hallyn: infinity deals with the chroots
[21:40] <stgraber> hallyn: so, my understanding of the armhf PPA builders is that those are essentially armhf chroots running on a x86 precise host or maybe even a precise VM on top of a hardy xen server (like the other PPA builders)
[21:40] <hallyn> ah
[21:41] <stgraber> hallyn: those are pretty different from the distro builders and non-virt PPA that build directly on pandaboards and so don't use any of the crazy qemu-user-static magic
[21:41] <cjwatson> and yes, the armhf PPA builders are emulated
[21:41] <cjwatson> the virtualised ones I mean
[21:42] <stgraber> so it may be that qemu-user-static can be updated in the armhf chroot which may or may not fix the issue, infinity should be able to help with that. If it's a problem with the qemu outside of the chroot, then I guess you'd need to talk to IS
[21:43] <stgraber> in any case, it should be easy to check whether the failure happens with a recent qemu, just create an armhf chroot or container on raring and do the build in there, if it passes, then there's hope that updating the qemu on the builders will fix it, if not, well, that'd be a qemu bug I guess
[21:43] <hallyn> would still be worth first seeing the source pkg which caused this so we can try and reproduce it - though i guess if IS runs the hosts it'll be hit-or-miss on reproducing it anyway
[21:43] <hallyn> ricotz: ^ worth trying as stgraber said
[21:43] <hallyn> probably using a lxc-armhf container on raring :)
[21:44] <stgraber> yeah, lxc-armhf is probably the easier, install lxc and qemu-user-static, then just do: lxc-create -t ubuntu -n builder -- --arch armhf --release raring
[21:52] <ricotz> hallyn, stgraber, i have checked with a raring pbuilder chroot which worked fine here
[21:53] <ricotz> hallyn, and sorry, i canceled the failing builds to free the builders since they didnt stop itself (also not after 2 hours)
[21:53] <ricotz> so no log
=== Guest54287 is now known as iulian
=== Quintasan_ is now known as Quintasan
=== mike is now known as Guest27454