2012/07/20/#ubuntu-desktop.txt

[00:41] <bcurtiswx> dupondje, i can get chat windows to come up, but it's a tad delayed
[00:50] <bcurtiswx> robert_ancell, bug #1026629 , I believe empathy upstream made it so only their default adium theme could be used or it was the default, and we need to patch it to allow the ubuntu adium theme as default or change to theirs (which i don't think we have) not sure though
[00:50] <ubot2> Launchpad bug 1026629 in empathy "empathy-chat crashed with SIGSEGV in empathy_adium_path_is_valid()" [Medium,Confirmed] https://launchpad.net/bugs/1026629
[00:59] <bcurtiswx> kenvandine, busy man today
[00:59] <kenvandine> almost ready to call it a day :)
[01:26] <bcurtiswx> if a package was sync'd from debian, where would I get the branch from ?
[01:28] <jbicha> bcurtiswx: apt-get source is one way
[01:29] <bcurtiswx> ah and create a branch, jbicha thx
=== jalcine is now known as Jacky
[03:22] <pitti> Good morning
[03:22] <robru> hello
[03:23] <kenvandine> good morning pitti
[04:31] <thumper> robert_ancell: ping?
[04:31] <robert_ancell> thumper, hello
[04:31] <thumper> robert_ancell: I'm hoping you can help, just upgraded a laptop to quantal
[04:31] <thumper> and now X won't start
[04:31] <thumper> ati, and fglrx
[04:32] <thumper> :(
[04:32] <thumper> not sure how to move forwards with it
[04:32] <robert_ancell> thumper, what does /var/log/lightdm/lightdm.log say
[04:34] <thumper> waiting for X... then Process 1157 exited with return value 1
[04:34] <thumper> x crashing I think
[04:34] <thumper> I'm looking in /var/log/Xorg.0.log
[04:35] <robert_ancell> yeah it sounds like it
[04:35] <thumper> X seg fault
[04:35] <thumper> inside fglrx_drv.so
[04:35] <thumper> :(
[04:36] <thumper> robert_ancell: how would I get it to use the other driver?
[04:36] <thumper> any ideas?
[04:36] <robert_ancell> thumper, uninstall fglrx
[04:36] <thumper> just with apt-get remove?
[04:37] <thumper> or do I need to purge or something
[04:38] <robert_ancell> thumper, just a normal remove should do it
[04:38] <RAOF> thumper: amd64?
[04:38] <thumper> hmm it says fglrx not installed
[04:38] <thumper> RAOF: yep
[04:38]  * thumper tries to install fglrx
[04:39] <thumper> fglrx-updates being removed
[04:39]  * thumper crosses fingers
[04:41] <thumper> hmm... download stalled at 15%
[04:41] <thumper> just sitting there
[04:41]  * thumper notices
[04:41] <thumper> RAOF: actually download says i386
[04:41] <thumper> I thought it was a 64 bit install
[04:41] <thumper> but perhaps not
[04:42] <RAOF> There *was* a known bug with fglrx on 64bit 1.12 xservers (like those found in, say, Quantal)
[04:42] <RAOF> Where it would segv rather than run.
[04:42] <thumper> RAOF: but it was fixed?
[04:42] <RAOF> I'm not sure what the status of that is; it was an fglrx bug, so I've got limited visibility.
[04:42]  * thumper wonders if the kids are hogging the bandwidth, download keeps stalling
[05:00] <robert_ancell> RAOF, btw, do you have any missing icons in unity? I was wondering if it's the PPA that's done it for me
[05:01] <robert_ancell> thumper, I take it my missing icon problem is not widespread..
[05:01] <thumper> robert_ancell: I can't get X running yet, let alone unity
[05:01] <thumper> robert_ancell: at 77% download for fglrx
[05:02] <thumper> seems to be crawling for me, no idea why
[05:02] <robert_ancell> thumper, I figured you'd have heard if it was so
[05:02] <thumper> other connections seem fine
[05:02] <thumper> which icons are you missing?
[05:02] <robert_ancell> thumper, panel and alt-tab, everything else works fine
[05:03] <robert_ancell> can't see anything in particular in .xsession-errors
[05:03] <thumper> what do you man missing icons in alt-tab?
[05:03] <robert_ancell> since that update the other day. Was is 6.0?
[05:03] <thumper> yeah
[05:03] <thumper> have you restarted?
[05:03] <thumper> the icons were moved from .../5/..
[05:03] <robert_ancell> yes
[05:03] <thumper> to ../6/...
[05:04] <thumper> hmm
[05:04] <thumper> thomi didn't report anything like that for his update
[05:04] <thumper> other issues, yes, but not that
[05:05] <thomi> ...having said that, I can't actually see my alt-tab window right now, so for all I know the icons might be missing
[05:06] <robert_ancell> thumper, /watch?v=QbjAjmssAdw&feature=p­layer_embedded
[05:06] <robert_ancell> http://ubuntuone.com/3pJh53lESd53tF7zYpsNuK
[05:06] <robert_ancell> I mean
[05:06] <robert_ancell> same thing for alt tab, i.e can still see the text but no icons/previews
[05:06] <thumper> robert_ancell: haha, that is freaky weird
[05:07] <thumper> robert_ancell: but I think I know why...
[05:07] <robert_ancell> so it's sufficiently usable for me not to spend too much time tracking it down :)
[05:07] <thumper> robert_ancell: it is bamf not matching the desktop filea
[05:07] <thumper> and returning nothing
[05:07] <thumper> we just landed a branch that will make them '?' icons
[05:07] <robert_ancell> ah
[05:07] <thumper> but we should fix the matching
[05:07] <thumper> not sure why your matching is failing though
[05:08] <robert_ancell> so they work fine in the dash, but do they come from apt/zeitgeist?
[05:09] <RAOF> robert_ancell: Yeah, it's the mesa in the PPA.
[05:10] <didrocks> good morning
[05:11] <RAOF> robert_ancell, thumper: Apparently if you apitrace it you find that unity is receiving (and ignoring) a bunch of glErrors from the uniform code.
[05:12] <didrocks> RAOF: you are speaking about the invisible icons in the launcher using the compositor ppa?
[05:13] <RAOF> didrocks: Yup.
[05:13] <didrocks> nice! When I first saw that on the french forum, I was scared as we just released a new version and they weren't clear they were using this ppa :)
[05:14] <thumper> which compositor ppa?
[05:14] <thumper> which uniform code?
[05:14] <thumper> X still isn't starting... :(
[05:15]  * thumper removes fglrx
[05:15] <RAOF> https://bugs.freedesktop.org/show_bug.cgi?id=52140 is the mesa bug.
[05:15] <ubot2> Freedesktop bug 52140 in Mesa core "Ubuntu Unity - Launcher and switcher icons disappeared" [Normal,New: ]
[05:15] <RAOF> thumper: ppa:ubuntu-desktop/system-compositor
[05:16] <thumper> RAOF: is this on top of wayland?
[05:17] <thumper> didrocks: the fglrx driver is crashing X on my quantal install
[05:17] <RAOF> No (although it also appears there, but there's actually no difference in the unity rendering path X vs X-on-wayland); this is either (a) unity doing something technically incorrect which mesa's now strict about or (b) core mesa breakage.
[05:17] <thumper> I'd go probably for (a)
[05:17] <thumper> although (b) is a possibility :)
[05:19] <thumper> so... I have X now...
[05:19] <thumper> kinda
[05:19] <RAOF> xorg.0.log?
[05:24] <thumper> RAOF: well, it is there, but dropped me into 2d
[05:25] <thumper> just took ages
[05:25] <thumper> I'm used to my SSD
[05:25] <RAOF> Ah.
[05:25] <RAOF> This suggests that you've broken 3d in some fashion. The packages *shouldn't* let that happen...
[05:26] <RAOF> ...but if everything that shouldn't happen didn't happen I'd be more interestingly busy. :)
[05:28] <RAOF> You haven't done anything that's likely to make me want to shout at you, like trying to manually install the fglrx drivers from amd's website, have you? ☺
[05:38] <thumper> no, just using quantals
[05:44]  * dupondje hopes empathy gets fixed asap :)
[08:00] <seb128> hello desktopers, happy friday!
[08:01] <didrocks> happy friday to you too seb128!
=== zyga-afk is now known as zyga
[08:24] <chrisccoulson> hey seb128, happy friday to you too!
[08:24] <seb128> chrisccoulson, hey, happy friday! how are you?
[08:24] <chrisccoulson> seb128, yeah, good thanks. still quite tired though. how are you?
[08:24] <didrocks> hey hey chrisccoulson, achieving getting some sleep? :)
[08:24] <chrisccoulson> didrocks, oh, i'm getting hardly any sleep at all now ;)
[08:25] <didrocks> hehe, that was what I reckoned! :)
[08:25] <seb128> chrisccoulson, I'm good thanks
[08:39] <seb128> Sweetshark, libreoffice didn't like arm*, failed to build there
[08:48] <dupondje> wasn't there a thing to report launchpad bugs upstream automaticly ?
[08:49] <dupondje> I remember like something :)
[08:56] <seb128> dupondje, no, launchpad can sync status and datas from upstream bugs though
[08:56] <seb128> dupondje, we never went the other way around to not "spam" upstream bug trackers
[08:56] <dupondje> seb128:  ok :)
[08:56] <dupondje> well not automaticly, but maby with some report-upstream.sh :)
[08:57] <dupondje> anyway did it manually now
[08:57] <dupondje> btw, the fact that Empathy has no menu bar is already on the radar?
[08:57] <dupondje> cause its quite annoying :)
[09:04] <seb128> dunno, I'm not using quantal yet, that would be a question for ken when he's online
[09:05] <mlankhorst> happy bday seb128 :)
[09:05] <mlankhorst> fday*
[09:05] <mlankhorst> evil keyboard
[09:05] <seb128> lol
[09:05] <seb128> I was going to say :p
[09:06] <seb128> mlankhorst, happy friday! ;-)
[09:06] <mlankhorst> friday's are better anyhow :D
[09:07] <pitti> ah, finally this channel becomes alive :)
[09:08] <mlankhorst> sadly so
[09:10] <dupondje> i'll stalk ken when he's online then :)
[09:11] <dupondje> Empathy seems to be a thing that breaks on every update :(
[09:15] <dupondje> btw, do we have some specific changes to gnome3 that could explain the menubar is gone in Empathy? Or 99% sure an upstream bug?
=== tkamppeter_ is now known as tkamppeter
[09:21] <tkamppeter> pitti, hi
[09:23] <pitti> hey tkamppeter
[09:26] <Sweetshark> moin!
[09:26] <tkamppeter> pitti, I have something ver strange: Gutenprint package updates PPD via CUPS trigger, I have 6 queues, all 6 updates confimed by "PPD of XXX updated", but first two queues have still old PPD. Looks like that the trigger is executed while the files of the printer-driver-gutenprint package are still copied and so the replacement of the /usr/lib/cups/driver/gutenprint.5.2 happens only after updating the PPD for the second and before upda
[09:26] <tkamppeter> ting the PPD for the third queue. Is this possible?
[09:28] <Sweetshark> seb128: yes, I see it. it segfaults in collecting the help pages. that help page seems to be huge. temporary disabling help on arm could help.
[09:28] <pitti> tkamppeter: hm, as far as I know, triggers happen after unpack; they may happen before configuring, though; would that make a difference?
[09:28] <Sweetshark> seb128: we are building the help packages on i386 anyway, but I dont know if LO picks them up if I disable help on arm.
[09:29] <pitti> tkamppeter: i. e. does gutenprint's postinst configure do something to the PPD files?
[09:30] <Sweetshark> seb128: Id suggest the following: I disable help on armel,armhf now and we upload that to -proposed as it might give us a package at least. in the meantime i am looking at this and possible further issues on a porterbox.
[09:31] <seb128> Sweetshark, that seems fine
[09:32] <seb128> Sweetshark, I'm out for lunch outside and some errands but I will catch up with the backlog and sponsor when I'm back
[09:39] <dupondje> bcurtiswx: there?
[09:41] <dupondje> or somebody else with crashing empathy-chat and quantal :)
[09:46] <tkamppeter> pitti, looked into the built maintainer scripts /var/lib/dpkg/info/*guten*.postinst and none shows anything influencing the PPDs.
[09:58] <tkamppeter> pitti, I also have seen in terminal output that the "unpacking" message is before the message for the CUPS triggers. Does the dpkg installation process parallelize like "make" or is always happening one step after the other?
[09:58] <pitti> tkamppeter: it shouldn't parallelize unpack with triggers; in fact, there are never multiple dpkg steps being executed at the same time
=== MacSlow is now known as MacSlow|lunch
[12:16] <bcurtiswx> dupondje, im here
[12:23] <dupondje> bcurtiswx: saw the comments on the empathy bug?
[12:23] <dupondje> seems to be packaging issue :)
[12:24] <bcurtiswx> dupondje, yes i saw
[12:25] <dupondje> you'll fix or ? :)
[12:25] <bcurtiswx> i want to
[12:25] <didrocks> bcurtiswx: some files are not shipped for the theme
[12:25] <didrocks> bcurtiswx: the empathy packaging should really use dh_install --fail-missing :)
[12:26] <bcurtiswx> so i'll need to add deps on the new adium theme
[12:26] <dupondje> would be cool if I could chat again this evening ^^:D
[12:36] <didrocks> bcurtiswx: so, got some confirmation: the built-in adium theme in empathy is used as a fallback when you have no theme shipping the necessary widgets
[12:36] <didrocks> bcurtiswx: so you need to put it in the -common package I guess (and please add --fail-missing to debian/rules) :)
[12:44]  * desrt got an email about fresh dconf crashes
[12:45] <didrocks> desrt: happy Friday \o/
[12:45] <bcurtiswx> didrocks: so i should add /usr/share/adium to empathy-common.install ?
[12:45] <desrt> didrocks: happy friday
[12:45] <didrocks> bcurtiswx: yeah, I just talked to upstream, how big it is?
[12:46] <desrt> is there a way i can query launchpad for fresh bugs with stack traces containing certain symbols showing up?
[12:47]  * desrt has a very low quality stacktrace and wants to look for better ones
[12:48] <didrocks> desrt: do you have any dup to that one?
[12:48] <desrt> someone emailed me a stack trace
[12:48] <didrocks> desrt: because if you keep it and apport detects a dup, it will remove the stacktrace while dupping
[12:48] <desrt> that's where i'm starting from :)
[12:48] <didrocks> ah :)
[12:50] <didrocks> desrt: recent crash on d-conf:
[12:50] <didrocks> https://bugs.launchpad.net/ubuntu/+source/d-conf/+bugs?field.searchtext=SIGSEGV&search=Search&field.status%3Alist=NEW&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.assignee=&field.bug_reporter=&field.omit_dupes=on&field.has_patch
[12:50] <didrocks> =&field.has_no_package=
[12:50] <desrt> nope
[12:51] <didrocks> well, seems you will have to invent the stacktrace content then :)
[12:51] <desrt> ya
[12:51] <desrt> i'm working on that
[12:51]  * desrt has found some hints
[12:51] <desrt> the dconf debug symbols are not installed, but the glib ones are
[12:51] <desrt> so there is some help
[12:53] <didrocks> well, not a nice game though :/
[12:53] <desrt> thomi: ping
[12:53] <desrt> thomi: i guess this is your stack trace, actually :)
[12:53] <didrocks> is he using ubuntu? like having the .crash file
[12:53] <desrt> thomi: thumper forwarded it to me
[12:53] <didrocks> you can retrace on your system then ;)
[12:54] <desrt> if i was running quantal :)
[12:54]  * desrt really doesn't like the look of this bug....
[12:58] <bcurtiswx> is there a good online resource for understanding debian/rules in depth
[12:59] <bcurtiswx> i can't find one
[13:00] <dupondje> google debhelper :)
[13:01] <didrocks> bcurtiswx: you want to know about --fail-missing right?
[13:01] <bcurtiswx> didrocks, yes I'm trying to learn w/o bothering
[13:02] <didrocks> bcurtiswx: no worry! so man dh_install first ;)
[13:02]  * didrocks looks if the package is using cdbs or debhelper
[13:02] <didrocks> ok, cdbs, I don't have an example of hand, let me look
[13:02] <didrocks> ah, in fact, you already have list-missing
[13:03] <didrocks> common-binary-post-install-arch:: list-missing
[13:03] <didrocks> replace that by fail-missing
[13:03] <didrocks> then, build the package (without any modif)
[13:03] <didrocks> and you should see the build failing
[13:03] <bcurtiswx> didrocks, yes I see that, but I want to know how I'm supposed to know that I was supposed to do that
[13:03] <didrocks> bcurtiswx: for cdbs, it's quire tricky, I usually read the cdbs source
[13:04] <didrocks> so, what's in /usr/share/cdbs/1/
[13:04] <didrocks> for packages using debhelper, man debhelper
[13:04] <didrocks> for having an overwrite of debhelper 7/8/9: http://joeyh.name/blog/entry/debhelper_dh_overrides/ (and man dh)
[13:05] <didrocks> overview*
[13:06] <dobey> anyone know anything about the new e-d-s?
[13:06] <dupondje> kenvandine: you know about the issue that empathy has no menu in gnome3?
[13:06]  * dobey loves it when things break api
[13:06] <kenvandine> dupondje, no...
[13:07] <dupondje> you can simulate it or?
[13:07] <kenvandine> i don't have gnome-shell installed atm
[13:07] <dupondje> oh ok
[13:07] <didrocks> dobey: cyphermox maybe can help you
[13:08] <kenvandine> dupondje, i know it has the single appmenu thing now
[13:08] <kenvandine> but haven't tried in shell
[13:08] <kenvandine> it works in untiy
[13:08] <kenvandine> 3.5.4 does seem buggy though
[13:09] <didrocks> kenvandine: ah, FYI, I'm asking bcurtiswx to ship the missing theme in the -common package as it's breaking empathy and also changing list-missing by fail-missing
[13:09] <kenvandine> ah, cool!
[13:09] <dupondje> hmz :) seems like you need to open it another way: http://www.linuxidc.com/upload/2012_07/12071410368377.png
[13:10] <dupondje> damn thats hidden :s
[13:10] <dupondje> :)
[13:10] <bcurtiswx> didrocks, make: *** No rule to make target `fail-missing', needed by `common-binary-post-install-arch'.  Stop.
[13:12] <kenvandine> dupondje, ah... yes that is the new way :)
[13:12] <dupondje> if you never used it, you really can search hours for that imo :p
[13:13] <didrocks> bcurtiswx: ah, I thought fail-missing had a target like list-missing
[13:13] <didrocks> bcurtiswx: so looking at the source, you should remove that line and do:
[13:13] <didrocks> DEB_DH_INSTALL_ARGS := list-missing
[13:13] <bcurtiswx> you mean fail-missing?
[13:13] <didrocks> see /usr/share/cdbs/1/rules/debhelper.mk
[13:13] <didrocks> oupss yeah :)
[13:14] <didrocks> DEB_DH_INSTALL_ARGS := --fail-missing
=== MacSlow|lunch is now known as MacSlow
[13:19] <bcurtiswx> didrocks, im guessing (like it was with bug control for me) that you can't really learn these things on your own.. It just takes asking millions of questions as you package to learn
[13:20] <didrocks> bcurtiswx: you can ask or grep the code, I did the second on that case :)
[13:20] <didrocks> (googling, saw that there was no really killer documentation on cdbs and so, back to the makefiles of cdbs)
[13:20] <didrocks> it's not that complicated
[13:20] <didrocks> just takes time…
[13:21] <didrocks> asking is good as well :)
[13:21] <dupondje> feel free to make documentation ;)
[13:22] <didrocks> well, cdbs is more or less dead by now :)
[13:22] <didrocks> so maybe doesn't worth it
[13:23] <bcurtiswx> i really struggle trying to make connections from A) Oh there's an issue to B) I don't know what to do to C) Where do I look to find out what to do.. there isn't many things that can get me to C.. which is probably why i'll have to ask
[13:23] <didrocks> bcurtiswx: this is called experience I guess :)
[13:23] <didrocks> and maybe why you are not fully ready yet to have upload privileges, but that will come!
[13:23] <bcurtiswx> didrocks, thats my original point. It's gonna take repetition and asking lots of questions
[13:24] <bcurtiswx> didrocks, that is the reason, there's no doubt. ;)
[13:24] <didrocks> bcurtiswx: btw, I made some years ago my own cheat page, if that can help you: https://wiki.ubuntu.com/DidierRoche/MOTU/memorandum
[13:28] <jbicha> bcurtiswx: asking questions is fine :)
[13:28] <bcurtiswx> didrocks, thanks
[13:29] <bcurtiswx> didrocks, http://paste.ubuntu.com/1101797/
[13:29] <bcurtiswx> didrocks, so i see these errors, what should be my first thought
[13:30] <didrocks> bcurtiswx: so, if the package was compiling successfully with dh_install --fail-missing before (not the case here), it means that upstream added in the new release (or that you added a configure option) new files that are installed for them
[13:31] <didrocks> bcurtiswx: the error here is "those are the files that upstream installs and you didn't ship them in any package"
[13:31] <didrocks> and indeed, you can see in that case that everything in usr/share/adium/message-styles is not shipped
[13:31] <didrocks> so, you need to assess if it worths shipping them
[13:31] <didrocks> asking upstream if necessary
[13:32] <didrocks> and ship them in an existing package or a new one
[13:32] <didrocks> in that case, those are theme files and we  know we should ship them by default
[13:32] <didrocks> (see previous conversation)
[13:32] <didrocks> as the theme files are not architecture specific, it's best to get them built in a arch:all package (I hope this notion is known?)
[13:33] <didrocks> the empathy-common package seems to be the right spot :)
[13:33] <didrocks> now comes the 4 first ones
[13:33] <didrocks> most of the time, we don't ship static library nor .la files in ubuntu
[13:33] <didrocks> so we need to make dh_install --fail-missing happy and not crying because of those
[13:33] <didrocks> right?
[13:33] <bcurtiswx> didrocks, yup
[13:34] <didrocks> what I generally do, just before the dh_install --fail-missing call, I rm -f them in debian/tmp
[13:34] <didrocks> then dh_install --fail-missing is run and find nothing in debian/tmp that is in any package
[13:35] <didrocks> with debhelper, this is easy to do
[13:35] <didrocks> i don't remember the right target for cdbs, let me check that :)
[13:35]  * didrocks look for install in /usr/share/cdbs/1/rules/debhelper.mk
[13:37] <bcurtiswx> deb_dh_always_exclude?
[13:38] <didrocks> bcurtiswx: I think that won't work, we really need to find a makefile target to remove the files
[13:40] <didrocks> let me build to confirm :)
[13:40] <bcurtiswx> didrocks, what are you testing?
[13:42] <dupondje> you can rm them in debian/rules also ?
[13:43] <didrocks> dupondje: isn't what I said? :)
[13:43] <didrocks> bcurtiswx: hum, I can't build the package, seems egg-list-box is missing
[13:43] <didrocks> even if I apt-get build-dep empathy
[13:43] <bcurtiswx> i build fine
[13:44] <kenvandine> didrocks, valac
[13:44] <didrocks> oh
[13:44] <kenvandine> your default valac needs to be 18
[13:44] <didrocks> yeah, my default isn't the right one :)
[13:44] <didrocks> ok, thanks kenvandine!
[13:44] <kenvandine> that is a missleading error :)
[13:45] <didrocks> we should really have an autotool check the current valac --version
[13:45] <didrocks> not that it's juts installed or not :)
[13:45] <kenvandine> indeed
[13:45]  * didrocks turns 0.18 then
[13:45] <bcurtiswx> shouldn't the control file take care of that?
[13:45] <didrocks> bcurtiswx: no, it's juts ensuring that you have valac 0.18 installed
[13:46] <didrocks> not that it's your current valac
[13:46] <didrocks> valac can be installed in parallel and either is 0.14, 0.16 or 0.18 in quantal for instance
[13:46] <didrocks> (you can see that with valac --version)
[13:46] <bcurtiswx> ok
[13:46] <didrocks> this is because valac is an alternative, look at man update-alternatives
[13:48] <Sweetshark> seb128: 3.6.0~rc2-0ubuntu3 for -proposed on chinstrap
[13:50] <bcurtiswx> so you have /etc/alternative/valac set to 0.16 or something other than 0.18 ?
[13:51] <bcurtiswx> didrocks, ^^
[13:51] <didrocks> bcurtiswx: yeah, I had 0.14 for the unity lenses
[13:51] <didrocks> and I set it up to 0.18
[13:58] <didrocks> bcurtiswx: ok, I think the only way to do that with cdbs is something I don't really like
[13:59] <didrocks> meaning installing the .*a to the package and remove them
[13:59] <didrocks> (that's the only part I see, even in the official documentation: http://build-common.alioth.debian.org/cdbs-doc.html)
[14:00] <bcurtiswx> didrocks, so we have to do it this way or are we going another route?
[14:01] <didrocks> bcurtiswx: no, doing this I guess
[14:02] <bcurtiswx> didrocks, OK so what needs to be done next?
[14:02] <didrocks> bcurtiswx: I can show you in the branch, do you have your pending work anywhere?
[14:05] <didrocks> (confirmed it works)
[14:05] <bcurtiswx> pushing to lp:~bcurtiswx/ubuntu/quantal/empathy/3.5.4-0ubuntu2
[14:06]  * didrocks grabs
[14:22]  * bcurtiswx thinks didrocks has a slow internet connection :P
[14:22] <didrocks> bcurtiswx: no, I just changed my mind about the approach to take, building it :)
[14:23] <bcurtiswx> don't screw up the approach, im gonna be doing it this way from now on :P
[14:23] <didrocks> and done! :)
[14:23] <didrocks> bcurtiswx: well, I hope you will have some debhelper packages later on :)
[14:23] <didrocks> it's way less cryptic than cdbs
[14:24] <didrocks> (IMHO)
[14:24] <bcurtiswx> didrocks, i'm assigned to two packages so I imagine i'll find one :)
[14:24] <didrocks> bcurtiswx: have a look: lp:~didrocks/empathy/3.5.4-0ubuntu2
[14:24] <didrocks> so I added to empathy-common the theme
[14:24] <didrocks> I told "install the .la, .a" in the .install files
[14:25] <bcurtiswx> didrocks, how did you know to add it to empathy-common?
[14:25] <didrocks> and then, I remove them post-binary installation in the package
[14:25] <didrocks> oupss
[14:25] <didrocks> one sec, need to change something
[14:25] <didrocks> bcurtiswx: see above
[14:25] <didrocks> 15:32:17      didrocks | in that case, those are theme files and we  know we should ship them by default
[14:25] <didrocks> 15:32:26      didrocks | (see previous conversation)
[14:25] <didrocks> 15:32:51      didrocks | as the theme files are not architecture specific, it's best to get them built in a arch:all package (I hope this notion is known?)
[14:25] <didrocks> 15:33:06      didrocks | the empathy-common package seems to be the right spot :)
[14:27] <jbicha> almost all the gnome stuff uses cdbs so you'll have to be at least familiar with it
[14:27] <didrocks> that's seb128's fault
[14:27] <jbicha> lol, is it?
[14:27] <didrocks> he should have used doctor's who box and go to the future :)
[14:28] <didrocks> jbicha: look at debian/copyright on most of gnome (starting from gnome 2) packages :)
[14:32] <didrocks> bcurtiswx: see my last commit, is that all right with you?
[14:32] <seb128> heh
[14:33] <bcurtiswx> the install.* in rules
[14:33] <bcurtiswx> is there doc that says do that somewhere?
[14:33] <bcurtiswx> install/*
[14:34] <didrocks> bcurtiswx: well, I used the documentation pointed above (http://build-common.alioth.debian.org/cdbs-doc.html), but I never find it quite clear enough to me
[14:34] <bcurtiswx> didrocks, ah i see it, thx
[14:35] <didrocks> yw :)
[14:35] <bcurtiswx> didrocks, im good with that please do build it :)
[14:35] <bcurtiswx> but
[14:35] <bcurtiswx> actually
[14:35] <didrocks> yeah?
[14:36] <bcurtiswx> didrocks, add bug #1026629 as it will be fixed by this update i believe
[14:36] <ubot2> Launchpad bug 1026629 in empathy "empathy-chat crashed with SIGSEGV in empathy_adium_path_is_valid()" [Medium,Confirmed] https://launchpad.net/bugs/1026629
[14:38] <didrocks> bcurtiswx: indeed, doing that
[14:38] <didrocks> do you think bug #1026081 is linked?
[14:38] <ubot2> Launchpad bug 1026081 in empathy "empathy finds no accounts and contatcs" [Undecided,New] https://launchpad.net/bugs/1026081
[14:38] <didrocks> or the error is separate from the theme not showing them?
[14:39] <bcurtiswx> i don't understand the SS, but we'll dup it and see
[14:40] <didrocks> ok
[14:40] <didrocks> uploading then :)
[14:41] <bcurtiswx> didrocks, thanks for your help
[14:41] <didrocks> bcurtiswx: you're really welcome :)
[14:51] <bcurtiswx> hmmm
[14:57] <bcurtiswx> didrocks, have you tested the built empathy?
[14:58] <didrocks> bcurtiswx: I installed it and run it, why?
[14:58] <bcurtiswx> i got depracation warnings on install.. and i still can't see my contacts from google-talk
[14:59] <didrocks> I just have some google account and couldn't reach it before, so I'm not really online
[14:59] <didrocks> bcurtiswx: was already the case with 3.5.4-0ubuntu1
[14:59] <Sweetshark> seb128: ping?
[14:59] <bcurtiswx> didrocks, i figured it was a theme issue, but it did do that in 3.5.4-0ubuntu1
[15:00] <didrocks> bcurtiswx: what do you mean by "it did do that", so the gmail failure is not only theme, right?
[15:00] <didrocks> the theme issue is about empathy crashing AFAIK
[15:00] <bcurtiswx> didrocks, haha, i think it is fixed and i somehow wasn't running the new empathy even after dpkg -i them
[15:00] <didrocks> or maybe our theme doesn't really the new empathy, dunno
[15:00] <seb128> Sweetshark, hey
[15:01] <seb128> Sweetshark, I'm about to sponsor your update
[15:01] <didrocks> bcurtiswx: ah, same for me
[15:01] <didrocks> bcurtiswx: yeah, let's try to kill it
[15:01] <didrocks> and run it again
[15:01] <didrocks> got trapped by that, I don't really know empathy
[15:01] <seb128> mterry, howdy, how are you?
[15:01] <mterry> seb128, hello, good
[15:01] <bcurtiswx> didrocks, i think you have to kill quite a few things
[15:01] <Sweetshark> seb128: kk, no hurry. I just wanted to know if you might have missed it ;)
[15:02] <bcurtiswx> telepathy-*, empathy, MC5
[15:02] <Sweetshark> seb128: guess not ;)
[15:02] <seb128> mterry, question for you, you did the accountsservice patch for the hasmessage stuff right?
[15:02] <bcurtiswx> dupondje, hopefully empathy will be ready in a little bit :)
[15:02] <mterry> seb128, yeah
[15:03] <didrocks> bcurtiswx: can you try that and let me know (or better, check with ken who is more knowledgeable on empathy than I)
[15:03] <seb128> mterry, was the "X" standing for something in XHasMessage?
[15:03] <seb128> mterry, is that like X-Vendor-Key to say it's non official or something?
[15:05] <bcurtiswx> kenvandine, what needs to be killed when trying to restart empathy? telepathy-*, MC5, empathy , anything else?
[15:05] <kenvandine> sometimes empathy-chat and empathy-call
[15:05] <mterry> seb128, supposed to be like X-Vendor-Key, but can't use dashes with some of the dbus-code-generation helper scripts
[15:06] <seb128> mterry, ok, thanks, gdbus-codegen makes XHasMessages a "xhas-messages" property
[15:06] <seb128> mterry, which is incompatible with the old x-has-message
[15:06] <seb128> mterry, (code being ported to gdbus)
[15:07] <mterry> seb128, huh.  hm
[15:07] <seb128> mterry, I was trying to determine what to do ... nothing out of accountsservice relies on the gobject property right, renaming it xhas-messages doesn't matter?
[15:07] <seb128> mterry, they all use the dbus one?
[15:07] <mterry> seb128, oh yeah, that's likely true.  Let me do a quick grep in lightdm
[15:07] <dupondje> bcurtiswx: sweet! thanks for the work
[15:07] <bcurtiswx> hmm, i'll wait until empathy is built and installed
[15:08] <seb128> mterry, other option is to add an another to the xml to say force the property name to be x-has-messages (you can do that)
[15:08] <bcurtiswx> dupondje, thank didrocks he did all of it :) i was just there for the learning
[15:08] <seb128> mterry, I would just like to avoid adding hacks when not needed
[15:08] <didrocks> bcurtiswx: I tried to kill some of the things, but it crashes on startup here, seems to have some Camel issue, so better for kenvandine and you to look at it :)
[15:08] <didrocks> if it confirms :)
[15:09] <bcurtiswx> it doesn't crash here, i just get tp_connection_upgrade_contacts fail
[15:09] <mterry> seb128, yah, lightdm only uses dbus, and I bet the same is true for control-center
[15:09] <cyphermox> jbicha: ping?
[15:09] <jbicha> cyphermox: good morning!
[15:09] <didrocks> bcurtiswx: try a guest session maybe?
[15:09] <cyphermox> hey, good morning :)
[15:10] <cyphermox> I wish it was a better one, I'm still sick and feeling like crap
[15:10] <mterry> err, not gnome-control-center
[15:10] <cyphermox> jbicha: do you use VPNs much, with gnome-shell?
[15:10] <jbicha> no, I don't have any VPNs
[15:10] <cyphermox> ah, alright
[15:11] <mterry> seb128, yup, indicator-messages also directly uses dbus.  no reason for hacks then, the property can change names
[15:11] <seb128> mterry, excellent, thanks for confirming
[15:11] <bcurtiswx> didrocks, same issue there.
[15:17] <didrocks> bcurtiswx: you should really check with ken then. I even don't know which components are involved
[15:17] <bcurtiswx> ken is busy with packaging things for quantal for the webapps
[15:17] <bcurtiswx> i'll grab him later i guess
[15:18] <desrt> can someone upload a dconf package with a few patches?
[15:18] <didrocks> yeah, he sponsors, he fixes :)
[15:18] <didrocks> desrt: friday evening? ;)
[15:18] <desrt> hmm
[15:18] <desrt> good point :)
[15:21] <seb128> bcurtiswx, didrocks: if you have a stacktrace maybe ask about it on #empathy or #telepathy
[15:22] <desrt> i think i nailed that issue with the bad stacktrace
[15:22] <desrt> apparently it was discovered by something called autopilot
[15:22] <desrt> that seems like something that should be automatically filing launchpad bugs?
[15:24] <seb128> desrt, you mean? what bad stacktrace,
[15:24] <desrt> seb128: i had a crappy stacktrace in my inbox this morning from tim
[15:24] <desrt> i was trying to find a way to ask launchpad for similar traces (possibly with retracing)
[15:27] <seb128> ah
[15:29] <bcurtiswx> how can you get a stacktrace without a crash?
[15:29] <desrt> pet peeve: when you call some business or government office and have to sit through 100 recorded messages telling you "did you know that you didn't really want to call us you didn't really have to call us?"
[15:29] <desrt> bcurtiswx: attach gdb to the running program
[15:30] <desrt> gdb /usr/bin/whatever <pid>
[15:30] <desrt> that'll probably require root on a modern ubuntu system....
[15:30] <bcurtiswx> i run gdb /usr/bin/empathy then 'run' and i have to kill it before i can strace?
[15:31] <desrt> no.
[15:31] <desrt> wait... strace?
[15:31] <desrt> you can also attach strace to a running program with strace -p pid
[15:31] <desrt> again, needing root
[15:31] <bcurtiswx> stacktrace
[15:31] <desrt> strace = system call trace :)
[15:32] <desrt> you can just say 'bt' in gdb
[15:32] <bcurtiswx> backtrace == stacktrace ?
[15:32] <desrt> yes
[15:33] <desrt> if you run a program under gdb you can suspend it at any point with ^C
[15:33] <desrt> and resume with gdb command 'c'
[15:33] <desrt> so just run it... then ^C at whatever point
[15:33] <desrt> then 'bt'
[15:33] <desrt> i assume you are trying to find a backtrace at the point of a UI freeze?
[15:53] <seb128> Sweetshark, still there?
[15:57] <Sweetshark> yep
[15:57] <Sweetshark> seb128: yep
[15:57] <seb128> Sweetshark, your chinstrap files a -sa, i.e include orig tarball for upload
[15:57] <seb128> ups, a->are
[15:59] <Sweetshark> seb128: ah, meh. missed the checkbox in my jenkins
[15:59] <Sweetshark> seb128: should I fix it? Ill have a new upload in 5 minutes.
[15:59] <seb128> Sweetshark, yes please, thanks
[16:15] <seb128> bcurtiswx, did you manage to get a stacktrace? did you ask on #empathy?
[16:16] <bcurtiswx> seb128, yes and yes
[16:17] <seb128> bcurtiswx, is it getting anywhere? can I help you?
[16:17] <bcurtiswx> seb128, it is not getting anywhere, i'm trying to code dive, i should get kenvandine involved he knows a lot more than myself
[16:18] <bcurtiswx> seb128, kenvandine here a pastebin of a stacktrace http://paste.ubuntu.com/1101990
[16:18] <seb128> bcurtiswx, you go no help,reply from the empathy guys?
[16:18] <bcurtiswx> seb128, no reply
[16:20] <seb128> bcurtiswx, that's not a segfault ... you said it's hanging?
[16:20] <bcurtiswx> seb128, i asked in both channels, i didn't ping anyone in #empathy and pinged smcv in #telepathy no reply
[16:20] <seb128> bcurtiswx, what's happening exactly when doing what?
[16:20] <bcurtiswx> seb128, the contacts are not showing for google-talk
[16:20] <bcurtiswx> but they work with facebook
[16:20] <bcurtiswx> which are both -gabble
[16:21] <bcurtiswx> that error you see is what comes up
[16:21] <seb128> xclaesse, ^ do you know if that's a known issue?
[16:23] <seb128> Sweetshark, getting it ready?
[16:25] <Sweetshark> seb128: its there now
[16:25] <seb128> Sweetshark, thanks
[16:32] <bcurtiswx> bzr bd-do says it only works for merge mode packages.  I branched a package, did a merge-upstream and i figured I could do a bzr bd-do
[16:32] <bcurtiswx> what am I doing wrong?
[16:33] <seb128> bcurtiswx, bzr bd-do only works for debian-dir only vcs-es
[16:33] <seb128> bcurtiswx, why would you need that in a full source tree?
[16:33] <seb128> you can directly use bzr and quilt in the directory
[16:33] <seb128> without doing bzr bd-do
[16:33] <bcurtiswx> ah ha, nvm
[16:33]  * bcurtiswx hides
[16:33] <seb128> ;-)
[16:44]  * didrocks waves good evening and good week-end!
[16:44] <bcurtiswx> im getting debian/rules some of the includes are not found ("no such file or directory"), is it as simple as they don't exist because they're not needed?
[16:46] <bcurtiswx> /usr/share/gnome-pkg-tools/1/rules/uploaders.mk for example
[16:46] <bcurtiswx> was gnome-pkg-tools renamed?
[16:47] <bcurtiswx> wait, i need it installed first
[16:48] <bcurtiswx> new machine, i forget to build-dep things :-)
[16:50] <seb128> bcurtiswx, yeah, that would be weird
[17:02] <bcurtiswx> seb128, is there another place to check for dep bumps other than a NEWS file ?
[17:05] <seb128> bcurtiswx, usually diffing the configure.ac or configure.in from both versions
[17:05] <bcurtiswx> seb128, thx
[17:05] <kenvandine> bcurtiswx, the only thing i trust is diffing the configure.ac
[17:11]  * Sweetshark finds the LibreOffice4 thread on the dev-list. fun indeed.
=== jdstrand_ is now known as jdstrand
[17:44] <bcurtiswx> seb128, dpkg-shlibdeps: warning: Can't extract name and version from library name `libgvfscommon.so' soname bump time?
[17:49] <bcurtiswx> anyone?
[17:49] <seb128> bcurtiswx, no, just ignore it, that one is a "private lib" to be used by gvfs components only, the warning was probably already there on the current version
[17:50] <bcurtiswx> dpkg-shlibdeps: warning: debian/gvfs-daemons/usr/lib/gvfs/gvfsd contains an unresolvable reference to symbol g_mount_source_show_unmount_progress: it's probably a plugin
[17:51] <bcurtiswx> ignore as well?
[17:53] <seb128> bcurtiswx, yes, you can probably ignore all the unresolved symbol warnings on that one
[18:01] <bcurtiswx> seb128, feel like sponsoring? https://code.launchpad.net/~bcurtiswx/ubuntu/quantal/gvfs/1.13.3-0ubuntu1/+merge/116046
[18:09] <dobey> mterry: http://askubuntu.com/questions/166028/moving-deja-dup-backups-from-dropbox-to-ubuntu-one
[18:10] <mterry> dobey, I'm not actually sure.  I think another user did something similar and wrote about it.  I think you do have to adjust something.  /me looks
[18:11] <dobey> i wonder why apt-get update seems to be so horribly slow on my machine now :(
[18:19] <bcurtiswx> dobey, not just you
[18:23] <dobey> well, it's just the "Reading packages" part
[18:23] <dobey> and this is still 12.04
[18:24] <dobey> on my laptop it's fast though, and that's on 12.10
[18:43] <seb128> bcurtiswx, hey
[18:44] <bcurtiswx> seb128, hi
[18:44] <seb128> bcurtiswx, I will have a look but not tonight, it's 9pm on a friday night here
[18:44] <bcurtiswx> seb128, sure. I understand
[18:45] <dobey> seb128: why are you on irc? :)
[18:45] <bcurtiswx> seb128, grab a beer (or like 6) and watch some television
[18:46] <seb128> dobey, because I pinged the SRU guys about the unity SRU before going for dinner and didn't want to miss their reply,questions if there were any ;-)
[18:46] <seb128> bcurtiswx, sounds like a plan!
[18:46] <dobey> heh
[18:46] <seb128> bcurtiswx, looking through your changelog entry, did you run dh_install --list-missing in the build dir after build?
[18:46] <bcurtiswx> seb128, no
[18:47] <seb128> bcurtiswx, I didn't check so I could be wrong but they added a backend for recentlyused files and I think a .install will need to be updates to include it
[18:47] <bcurtiswx> seb128, looking now. thx
[18:47] <seb128> bcurtiswx, try the dh_install trick, it will tell you what stuff are installed by make install and not shipped in any binary package
[18:47] <bcurtiswx> how do i preserve build-dir in pbuilder?
[18:47] <seb128> bcurtiswx, you don't, I would recommend not using pbuilder for work
[18:48] <seb128> bcurtiswx, usually you want to do your update and once you are done to a pass through pbuilder for checks
[18:48] <bcurtiswx> seb128, darn, my desktop(precise) does like 10 times faster in building then my laptop(quantal)
[18:48] <seb128> bcurtiswx, oh, so how I do that is: pbuilding login
[18:48] <seb128> you get in the pbuilder
[18:49] <seb128> then you can apt-get build-dep and build like you would normally do
[18:49] <seb128> pbuilder login --save-after-login
[18:49] <seb128> edit your /etc/apt/sources.list to enable the deb-src
[18:49] <seb128> then install maybe devscripts and your prefered editor and exit 0
[18:50] <seb128> so your pbuilder "base" has those and you don't need to enable them everytime
[18:50] <seb128> it makes things easier
[18:50] <bcurtiswx> i can stick with the laptop, it's only semi-painful :P
[18:50] <bcurtiswx> so after a bd how do i stay in build-dir ?
[18:50] <bcurtiswx> bzr bd*
[18:52] <bcurtiswx> --dont-purge looks like what i want
[18:55] <seb128> bcurtiswx, bzr bd doesn't purge
[18:55] <seb128> bcurtiswx, if it's debian dir only, cd ../build-area/source-version
[19:02] <bcurtiswx> i wonder why it's not doing --list-missing or --fail-missing on build
[19:04] <bcurtiswx> kenvandine, do we ship .h files? (all i know is not .la or .a files)
[19:05] <seb128> bcurtiswx, we do yes
[19:05] <kenvandine> in devel packages
[19:22] <cyphermox> wow, there's goign to be a documentary on TV here about people trafficking pangolins
[19:22] <cyphermox> I had no idea Ubuntu was this popular ;)
[19:26] <bcurtiswx> kenvandine, theres a while entire blob of usr/share/man/man1/gvfs* since thats not arch specific im guessing arch:all would be best (gvfs-common)
[19:35] <dobey> it's also weird that html5 video and flash seem to now require the entire video to download before being able to play smoothly