2012/06/26/#ubuntu-desktop.txt

[00:02] <robert_ancell> RAOF, this is a MIR bug 1016158
[00:02] <ubot2> Launchpad bug 1016158 in libgusb "[MIR] libgusb" [Undecided,New] https://launchpad.net/bugs/1016158
[00:02] <RAOF> Yeah, saw that.
[00:27] <bryceh> RAOF, hey if you have a few minutes could you pump through the xdiagnose SRU?  Both bugs it fixes got validated-none.
[00:27] <bryceh> RAOF, one of the fixes is to turn off the gpu hang apport hook.  Apparently it's still been firing off, even though apport is disabled.
[00:27] <RAOF> Certainly,
[00:39] <RAOF> bryceh: Hm. xdiagnose hasn't been in there for the standard 7 days; are the changes are low-risk, so pushing it out a little prematurely reasonable.
[00:57] <bryceh> RAOF, hasn't it?  the date on the changelog is the 15th.  when did I upload it...?
[00:58] <RAOF> It got accepted 6 days ago, according to http://people.canonical.com/~ubuntu-archive/pending-sru.html
[00:58] <bryceh> ah, so one day shy
[00:58] <bryceh> ok, yeah seemed like it's been about a week.
[00:59] <bryceh> but yes, the changes are low risk.  it only affects the bug filing logic, so even if everything exploded it's just going to affect bug reporting
[01:01] <bryceh> RAOF, of course, no harm in letting it sit one more day if that's the required process.
[01:02] <RAOF> Doesn't xdiagnose also include failsafe-x?
[01:02] <bryceh> RAOF, it does
[01:02] <RAOF> In unlikely scenarios that could blow up pretty badly. ☺
[01:03] <bryceh> none of the changes actually alter any of that code path, but yeah that's there too
[01:34] <jbicha> hmm, a gnome-menus transition, that'll be fun :/
[03:40] <pitti> Good morning
[03:56] <TheMuso> Morning pitti.
[05:05] <robert_ancell> RAOF, so, what's the chance we can get these pandaboards to composite?
[05:09] <RAOF> robert_ancell: High; they've got GLES drivers, right?
[05:09] <RAOF> Unless the question is ‘what's the chance we can run the system compositor on them’ in which case... I *think* there's an open driver in the works.
[05:09] <robert_ancell> that was the question
[05:11] <RAOF> Failing anything else, I believe a software-rendered compositor would be possible now. That's not exactly awesome, though.
[05:11] <robert_ancell> RAOF, that would actually be fine
[05:14] <RAOF> I'll give it a whirl once I've set it up properly.
[05:28] <TheMuso> Its probably worth grabbing a USB hard drive, netboot installing using the alternate installer onto the USB drive, and using the board that way. Less thrashing of an SD card, and a little snappier.
[05:29] <TheMuso> You still need an SD card to boot, but once booted, the USB drive does the rest.
[05:29] <TheMuso> And jockey offers to install the pvr-omap package.
[05:30] <TheMuso> jbicha: Gnome shell will be depwait atm, at-spi2-atk is still in binary new.
[05:36] <TheMuso> Mine is currently running precise, only because for some reason the quantal kernel didn't boot, but I might have done things incorrectly with the SD card. WIll try again later.
[05:36] <RAOF> Oh, it doesn't boot from USB?
[05:37] <pitti> RAOF: no, I learned that the hard way last week, too
[05:38] <TheMuso> I think you can get it to boot via the USB port on the side, i.e so the board is a USB client, but not with a USB host port no.
[05:38] <pitti> I bought an 8 GB SD card on the weekend
[05:38] <RAOF> ...
[05:38] <RAOF> Ok, SD card time.
[05:38] <pitti> you can put the main file system on an USB drive, but you still need a small SD card for early uboot stuff
[05:38] <TheMuso> pitti: Yeah, thats what I did, as above.
[05:39] <TheMuso> THings actually perform acceptabbly from a USB drive, and plain suck from SD card only.
[05:39] <TheMuso> Not to mention killing the SD card.
[06:09] <didrocks> good morning
=== jalcine is now known as Jacky
[06:54] <pitti> bonjour didrocks
[06:55] <didrocks> guten morgen pitti
[07:35] <didrocks> greyback: hey, are you around?
[07:36] <greyback> didrocks: yep
[07:36] <didrocks> greyback: did you get a change to look at the stacktrace I gave to youN?
[07:36] <greyback> didrocks: not really tbh, was at conference last week.
[07:37] <greyback> didrocks: I should be able to give it some time later this evening
[07:38] <didrocks> greyback: excellent, thanks!
=== RAOF_ is now known as RAOF
[07:47] <Laney> slomo: look at libgstreamer0.10-dev libgstcoreelements.la libgstcoreindexers.la http://paste.debian.net/176353/
=== vuntz_ is now known as vuntz
[07:47] <slomo> Laney: ah those, a patch to just remove them would be good ;)
[07:47] <slomo> Laney: they're completely unneeded
[07:48] <Laney> yep, that's what I linked you to yesterday
[07:48] <slomo> Laney: i mean, remove the files, not just clear dependency_libs
[07:49] <Laney> oh, well if you're sure nothing depends on them
[07:49] <Laney> just a find -name *.la -delete in rules then :P
[07:50] <RAOF> There should be a dh helper for that :)
[07:50] <RAOF> dh_lakiller
[08:02] <seb128> hey
[08:02] <didrocks> salut seb128
[08:02] <didrocks> seb128: meeting reminder day!
[08:02] <seb128> lut didrocks, en forme ?
[08:02] <seb128> didrocks, oh, right, thanks!
[08:02] <didrocks> yw ;)
[08:02] <didrocks> ça va, et toi?
[08:04] <seb128> ca va nickel ;-)
[08:04] <Laney> can someone give-back https://launchpad.net/ubuntu/+source/gnome-font-viewer/3.5.3-0ubuntu1/+build/3605498 please? :-)
[08:07] <didrocks> Laney: done
[08:08] <Laney> thanking you
[08:35] <chrisccoulson> good morning everyone
[08:35] <RAOF> Hey chrisccoulson!
[08:35] <chrisccoulson> hi RAOF, how are you?
[08:36] <RAOF> I'm pretty good.
[08:37] <seb128> hey chrisccoulson, RAOF, how are you?
[08:37] <chrisccoulson> hi seb128
[08:37] <RAOF> Hey seb128!
[08:37] <chrisccoulson> i'm good thanks, how are you?
=== tkamppeter_ is now known as tkamppeter
[08:37] <seb128> chrisccoulson, I'm good thanks
[08:37] <seb128> RAOF, I've a graph for you, tell me what you think about it :p
[08:37] <seb128> RAOF, http://status.ubuntu.com/ubuntu-quantal/u/raof.html ;-)
[08:38] <seb128> RAOF, (don't worry you are not the only one in that case, but the result is that the team trend is similar to yours, which we need to fix ;-)
[08:39] <seb128> RAOF, though to be fair you seem to have collected all those lcms items that shouldn't be there
[08:39] <RAOF> Again?!  Bah :)
[08:39] <seb128> I will try to make them go away
[08:40] <seb128> RAOF, I unset the assigned of the color management spec, let's see how that goes
[08:40] <seb128> RAOF, is the work on libxrandr-utils still going on? bryceh and you have a fair share of items for that on the list
[08:41] <RAOF> It's kindof burning slowly.
[08:41] <RAOF> I need to rush off; I'll be back on later if you want to keep talking :)
[08:42] <seb128> RAOF, no, that's fine, I guess I just wanted to say "please make sure you update you work items when some are done" ;-)
[08:42] <seb128> RAOF, have a nice evening
[08:42] <seb128> chrisccoulson, ** No longer affects: firefox (Ubuntu)
[08:42] <seb128> chrisccoulson, you got annoyed by the spamming as well ;-)
[08:42] <chrisccoulson> seb128, yeah :(
[08:43] <seb128> I will do the same
[08:43] <chrisccoulson> i hate bugs with a bazillion tasks, when you get bug mail for months after fixing your own bits
[08:43] <seb128> I tried to argue with pitti and others before without success
[08:43] <seb128> chrisccoulson, yeah, same here
[08:43] <pitti> hey seb128
[08:43] <seb128> pitti, hey ;-) wie gehts?
[08:43] <pitti> seb128: gut, danke!
[08:44] <pitti> seb128: I'm currently working on dropping the "|| true" on the glib and GTK "make check" calls; they make me weep through the night :)
[08:44] <pitti> any objection? Seems we mostly inherited them from Debian
[08:44] <pitti> glib works perfectly for us, on Debian there are just two failures
[08:45] <seb128> pitti, no objection, I think they were mostly added for non common archs in debian
[08:45] <pitti> chrisccoulson: you could just have muted yourself from the bug instead of removing the affected packages..
[08:46] <chrisccoulson> pitti - ooh, i'd totally forgotten about that feature
[08:46] <seb128> chrisccoulson, pitti, though that's only moving the issue to the next subscriber to the source
[08:46] <seb128> I've dropped the desktopish lines from done items as well
[08:47] <seb128> pitti, did you check the logs to see if any of the gtk,glib testsuit got any failure in the recent cycles?
[08:47] <pitti> seb128: I checked the logs of the current upload
[08:47] <seb128> pitti, I didn't realize we were still doing || true, I got so used to the glib one failing the builds because it was locking until the build was getting killed :p
[08:48] <pitti> seb128: right, but || true doesn't help with the locking anyway :)
[08:48] <seb128> indeed ;-)
[08:48] <pitti> and now that we have newer buildds it should not happen in the first place
[08:48] <seb128> that and chrisccoulson and desrt fixed the buggy glib test in precise
[08:48] <chrisccoulson> :)
[08:49] <seb128> we need new buggy tests
[08:49] <seb128> chrisccoulson, you like debugging those right? ;-)
[08:49] <chrisccoulson> lol
[08:57] <Laney> slomo: do you have packaging repos for gstreamer? or do you want a debdiff?
[09:05] <slomo> Laney: git.debian.org but a debdiff is fine too
[09:06] <Laney> maybe I'll add you some Vcs headers while I'm there :P
[09:52] <seb128> Laney, speaking about gstreamer, do you have a CD drive? ;-)
[10:00] <didrocks> alf_: ogra_: so, building unity on armel with the compiz opengl is still an issue
[10:00] <didrocks> https://launchpadlibrarian.net/108687017/buildlog_ubuntu-quantal-armhf.unity_5.12-0ubuntu4~ppa2_FAILEDTOBUILD.txt.gz
[10:00]  * didrocks looks
[10:01] <didrocks> nuxgraphics still try to use the opengles path
[10:02] <didrocks> ah, got the guilty instruction
[10:05] <ogra_> well, feel free to disable *all* plugins if you like :)
[10:05] <ogra_> all we need is a dummy package that fulfills the deps to build images
[10:05] <alf_> didrocks: ogra_: I am not sure how well this will work. I think they are going to be API mismatches between unity-gl (i.e. unpatched) and compiz-gles2 (patched)
[10:05] <ogra_> alf_, we dont have any GLES drivers atm ...
[10:05] <ogra_> it doesnt need to work
[10:06] <didrocks> ogra_: right, but still, upstream not meeting their commitment is making distro loosing days of work :/
[10:06] <ogra_> but we cant build alpha2 images if the dependencies cant be fulfilled
[10:06] <ogra_> didrocks, yes
[10:06]  * ogra_ cant do much here ... neither can you i suppose
[10:07] <alf_> didrocks: ogra_: I am referring to build-time problems, unfortunately
[10:07] <ogra_> oh, you mean unity would ftbfs on top of an empty compiz build ?
[10:08] <didrocks> ogra_: yep ;)
[10:08] <ogra_> (or s/empty/old/
[10:08] <ogra_> )
[10:08] <didrocks> let's see
[10:08] <ogra_> sigh
[10:08] <didrocks> it's currently building
[10:08]  * ogra_ crosses fingers
[10:11] <Laney> seb128: yeah, but not any cds atm
[10:12] <seb128> Laney, not even music ones?
[10:12] <ogra_> didrocks, did you plan to have that version in the A2 images btw ?
[10:13] <didrocks> ogra_: well, at least in -proposed
[10:13] <didrocks> ogra_: if it's in, great, otherwise, will be post A2
[10:13] <ogra_> well, if the image builds break for a few days right after A2 i'm not muchly worried i must say
[10:14] <ogra_> but we only had one arm image for A1 so i need to make sure to get the ducks in a row here ... read: A2 is essential for arm
[10:15] <ogra_> if it can stay in proposed and we get a new reliable commitment from upstream thats not to far in the future, i wouldnt mind the breakage
[10:15] <ogra_> did anyone tell you when thy plan to be done ?
[10:15] <ogra_> *they
[10:17] <didrocks> ogra_: I'm trying to get the plan for weeks already
[10:17] <didrocks> was supposed alpha1
[10:17] <didrocks> then alpha2 at the latest
[10:17] <ogra_> yeah, i remember that part :)
[10:17] <ogra_> well, the A1 part
[10:17] <didrocks> so, can't really trust the schedule anymore :)
[10:17] <ogra_> right
[10:21] <seb128> Laney, context is that I'm trying to find somebody to look at finishing bug #945987 (ronoc did most of the work, there is just to put the pieces together, test and get them uploaded)
[10:21] <ubot2> Launchpad bug 945987 in rhythmbox "No Settings are available in "Preferred format", only preset defaults are used" [High,In progress] https://launchpad.net/bugs/945987
[10:21] <seb128>  
[10:22] <seb128> ogra_, you should ask asac to get Linaro to help on getting the compiz gles work finished ;-) the #ps guys are short on resources
[10:22] <ronoc> seb128, yup that is still an issue
[10:22] <ronoc> it's hardcoded to what i think is an acceptable profile but the user cannot change this
[10:22] <seb128> ronoc, yeah, I'm trying to make sure it ges addressed for precise and quantal
[10:23] <ronoc> seb128, good stuff
[10:23] <ogra_> seb128, well, we had a long discussion at UDS (didrocks was there too) and upstream said they would get along fine and have everything ready by the agreed data ... and its not like we havent aked twice if thats doable
[10:23] <ogra_> *date
[10:24] <seb128> ogra_, well, we are all on the same side, it's one of those things which is hard to estimate before you really started to work on it
[10:25] <seb128> ogra_, it seemed it turned out that the linaro patch is creating a number of regression and bugs and it's taking them time to sort those out
[10:25] <ogra_> right, and linaro is doing their duty already, not sure there is much they can help with
[10:25] <seb128> ogra_, well, they could help fixing the bugs in their patch ;-)
[10:25] <seb128> ogra_, that's where most of the delay is coming from
[10:25] <ogra_> oh, there is definitely commitment from jesse (linaro graphics manager) to help with upcoming bugs
[10:26] <ogra_> alf_, ^^^ are you aware of the bugs ?
[10:26] <ogra_> (did anyone from upstream talk to you ? )
[10:26] <seb128> ogra_, well, my understanding from what duflu said is that they still didn't roll it out because of the number of regressions the linaro changes created and they have nobody helping them to get those sorted
[10:26] <seb128> ogra_, alf_: https://bugs.launchpad.net/compiz/+bugs?field.tag=gles
[10:27] <ogra_> i'll talk to jesse
[10:27] <seb128> ogra_, thanks
[10:27] <ogra_> "and they have nobody helping them to get those sorted" ... i still wonder if they have asked at all :P
[10:28] <didrocks> well, most of the delay is on plugins we don't support unfortunatly
[10:28] <didrocks> I would rather they drop them
[10:28] <seb128> ogra_, I don't know, there might be some management to engineers mixed with cross team interactions issues
[10:28] <didrocks> but it seems as an upstream, they don't want to
[10:29] <ogra_> well, the have the choice between no support for arm at all or *some* support on arm ... i know what *i* would chose there
[10:30] <ogra_> (though i also know that GLES isnt arm only anymore ... poulsbo uses it too ...)
[10:30] <seb128> didrocks, it seems they are sort of agreeing on dropping those now though?
[10:31] <seb128> ogra_, didrocks: well bottom line is that it would help if linaro (or anyone else) could give an hand sorting the remaining bugs and do testing of the current vcs
[10:31] <ogra_> right
[10:31] <didrocks> seb128: no, see the bug list
[10:32] <didrocks> seb128: and they are readding even more exotic plugins to trunk
[10:39] <alf_> seb128: The remaining problems require more compiz internals expertise than GLES2 expertise, so linaro (which is essentially just me in this case) can't really offer any substantial assistance. In any case, I am keeping track of the issues and I have asked upstream to assign bugs to me if they think I am a better match for it.
[10:45] <Laney> seb128: I mean not *here*, in the office at uni. I have some at home.
[10:52] <Laney> ronoc: So all is left to do is to add that preset = rhythmbox-custom-settings to /usr/share/rhythmbox/rhythmbox.gep?
[10:52] <Laney> if I do that here I get a button that asks me to install additional software, but is greyed out
[10:56] <ronoc> seb128, did that get distro patched for p
[11:02] <ronoc> Laney, do you have the lame preset that the gep is referring to
[11:02] <ronoc> sorry i haven't looked at this since april, at which point i thought it was sorted
[11:02] <Laney> you mean rhythmbox-custom-settings?
[11:02] <ronoc> (until we port gnome-media-profiles to gtk3)
[11:02]  * Laney is a bit clueless atm
[11:02] <ronoc> Laney, yep
[11:03] <ronoc> okay
[11:03] <Laney> it's created by rb
[11:03] <ronoc> i solved this by created my gstreamer preset
[11:03] <Laney> in ~/.gstreamer-0.10/presets/GstLameMP3Enc.prs
[11:03] <ronoc> s/created/creating
[11:03] <ronoc> yes
[11:03] <ronoc> and then in the gep it should refer to that prs for it's encoding
[11:03] <Laney> I think he's saying that the missing part is to tell rhythmbox to use this created preset by adding those lines to the .gep file, correct?
[11:04] <ronoc> yes
[11:04] <Laney> does it fall back to the default if they don't exist?
[11:04] <ronoc> it should
[11:04] <Laney> I don't know what this install additional software thing is about, though. (but I am on unstable, which may make a difference)
[11:05] <Laney> there should be an upstream bug for this too then
[11:05] <Sweetshark> software archaelogy continues -- not for the faint at heart: http://nabble.documentfoundation.org/where-are-extras-source-templates-wizard-bitmap-used-td3992151.html
[11:05] <Sweetshark> cue Indiana Jones soundtrack
[11:06] <ronoc> Laney, the install additional software usually points to some issue with the gep, it doesn't recognize the preset or doesn't have the encoder for that preset
[11:07] <Laney> (12:07:14) [0x875c640] [impl_get_missing_plugins] rb-encoder-gst.c:778: didn't get request pad, profile oggvorbis doesn't work
[11:08] <Laney> ronoc: maybe you could make the patch here, as you know more about this than me :P
[11:08] <Laney> I'll take care of getting it into Q and filed upstream
[11:09] <ronoc> I'm unsure of what we are doing here. This was just a quick fix for P, the ideal fix is for gnome-media-profiles to be resurrected by porting it to gtk3
[11:09] <ronoc> the patch for rb is a one liner in the gep
[11:09] <ronoc> i don't think it belongs upstream
[11:09] <Laney> but they implemented the custom settings stuff upstream
=== MacSlow is now known as MacSlow|lunch
[11:10] <ronoc> Laney, in gstreamer ?
[11:10] <Laney> in rb
[11:10] <ronoc> so where is this new preset located on the file system ?
[11:10] <Laney> http://git.gnome.org/browse/rhythmbox/commit/?id=b2439d89df2a4aa5883d2d947fc07c150d0b3193
[11:10] <Laney> is is created by rhythmbox when you fiddle with the settings in the ui
[11:10] <Laney> so in your home dir
[11:12] <ronoc> i don't know  why  with this patch you are seeing the greyed out button
[11:13] <Laney> I probably Did It Wrong™
[11:13] <ronoc> it seems like it should work
[11:13] <ronoc> also what patch do you need me todo
[11:13] <ronoc> if rb is already patched
[11:13] <Laney> it seems like the .gep needs patching
[11:13] <Laney> at least that's what the guy on the bug says
[11:16] <ronoc> Laney, just to clarify, do you have the patched rb installed ?
[11:16] <Laney> yes, I see the new custom settings UI
[11:17] <ronoc> if so simply add that one liner to the gep
[11:17] <ronoc> if it doesn't work then there is aproblem with either your install or that patch
[11:17] <ronoc> i can patch rb
[11:17] <ronoc> 's gep
[11:17] <ronoc> or a problem with the encoder and our preset name
[11:18] <seb128> re
[11:18] <Laney> the install additional software button works on my home (Ubuntu) machine
[11:18] <seb128> Laney, ronoc: I was out for lunch
[11:18] <Laney> maybe it's a problem here
[11:18] <Laney> ronoc: I think all that's missing is adding the preset lines to the gep file indeed
[11:19] <seb128> Laney, you shouldn't have to install things if the gep is correct, the first version in the bug report had extra "" breaking the parsing but the bottom version should be fine
[11:19] <seb128> Laney, basically I think we need gstreamer to install the profile and rb to use it IIRC
[11:20] <Laney> I think RB now creates profiles that gstreamer can use, but doesn't tell gstreamer to use them because of a missing entry in the gep file
[11:21]  * Laney → lunch
[11:21] <seb128> Laney, enjoy
[11:22] <Laney> slomo: here are your patches for gstreamer0.10/debian http://paste.debian.net/176377/
[11:23] <slomo> Laney: thanks
[11:23] <ronoc> seb128, do  you know if xingmux is used in the encoding pipelines
[11:23] <seb128> ronoc, no idea
[11:24] <didrocks> ogra_: alf_: my amrhf build passed FYI :)
[11:24] <ogra_> \o/
[11:24] <ogra_> thanks so much for that work !!!!!
[11:25] <ronoc> brb
[11:25] <didrocks> ogra_: yw ;)
[11:26] <didrocks> pushing to -proposed onw
[11:26] <didrocks> now*
[11:26] <ogra_> yay
[11:26] <didrocks> and unblocking the compiz SRU
[11:26] <didrocks> phew ;)
[11:26] <slomo> Laney: looks good
[11:27] <Sweetshark> seb128: nothing blocking the 3.5.4 SRU which I need to action currently, right?
[11:28] <seb128> Sweetshark, not that I know about no
[11:29] <seb128> Sweetshark, it still has 2 days to wait in proposed and it still needs somebody to confirm that the SRU works fine and flag the bug verification-done
[11:31] <Sweetshark> k
[11:42] <alf_> didrocks: great!
[11:47] <seb128> mvo_, hey, is there any way you could get somebody from the s-c team to check the aptdaemon precise SRU and see if the bugs can be marked verified?
[11:58] <xclaesse> seb128, please remember to SRU telepathy-gabble. should I open a lp bug for that?
[11:58] <seb128> xclaesse, I was sort of waiting for Debian to update it so I could sync to quantal which is supposed to happen before a SRU (i.e update unstable first, then backport)
[11:59] <seb128> xclaesse, do you have anyone in debian you can ping to get it updated there?
[11:59] <xclaesse> bigon, ^
[11:59] <seb128> xclaesse, if not I guess I will just update quantal directly...
[11:59] <seb128> xclaesse, thanks for the ping though ;-)
[12:00] <xclaesse> seb128, pinged our usual debian guys on #telepathy, surely someone will pick it :)
[12:00] <seb128> thanks
=== MacSlow|lunch is now known as MacSlow
=== zyga is now known as zyga-afk
[12:20] <larsu> did empathy in quantal stop working for anyone else today?
[12:21] <larsu> it's connecting to gtalk, showing me my contacts and everything, but refuses to start a chat
[12:21] <seb128> larsu, try chekcing with kenvandine when he gets online or with xclaesse
[12:22] <larsu> seb128, will do, thanks
[12:22] <xclaesse> larsu, AFAIK empathy-chat segfault, should be fixed in latest empathy release
[12:24] <seb128> larsu, do you get any error if you run empathy from a command line? what happens if you try to chat? is empathy-chat segfaulting?
[12:24] <larsu> xclaesse, I'm on 3.5.1-0ubuntu3, is there a newer one upstream?
[12:24] <larsu> seb128, no messages except something about it not being able to write avatar pictures to the cache because of long filenames
[12:24] <larsu> no crashes either
[12:25] <seb128> larsu, well in any case I didn't see the issue mentioned before and I'm still on precise so I can't tell from here
[12:26] <xclaesse> hm, actually could be wrong... but I have that issue here on ubuntu precise, empathy 3.4.2.1
[12:26] <xclaesse> clicking on a contact in emapthy's contact list and nothing appear
[12:26] <larsu> xclaesse, same for me
[12:26] <xclaesse> actually the chat is visible in gnome-shell's notification
[12:26] <xclaesse> if I start empathy-chat process manually, then it works
[12:27] <xclaesse> larsu, start /usr/lib/empathy/empathy-chat manually, then click on a contact
[12:27] <larsu> xclaesse, ha! your're right
[12:27] <larsu> weird.
[12:27] <seb128> did that start recently?!
[12:28] <larsu> seb128, it worked this morning, I updated before going to lunch, after lunch it didn't work anymore
[12:28] <larsu> so... yeah :)
[12:28] <seb128> larsu, can you check in /var/log/dpkg.log what you updated?
[12:28] <larsu> seb128, sure
[12:30] <xclaesse> seb128, any idea how I can get a backtrace of empathy-chat when it crash at startup when it's dbus-activated
[12:30] <xclaesse> starting it manually does not crash
[12:30] <xclaesse> seb128, can apport do that?
[12:31] <seb128> xclaesse, yeah, it should, otherwise just mv empathy-chat empathy-chat.real and create a wrapper
[12:31] <seb128> xclaesse, is the issue logged in /var/log/apport.log?
[12:31] <larsu> seb128, weird, empathy was not updated today..
[12:32] <larsu> but it updated *a lot*, don't know where to start looking
[12:32] <seb128> larsu, check if there is any gnomish lib that empathy could be using?
[12:32] <seb128> larsu, do you run your own glib for the indicators hacking you do?
[12:32] <larsu> seb128, no, I run jhbuild
[12:32] <larsu> (to develop)
[12:33] <xclaesse> seb128, ah! ERROR: apport (pid 20394) Tue Jun 26 14:32:53 2012: wrote report /var/crash/_usr_lib_empathy_empathy-chat.1000.crash
[12:33] <larsu> the broken empathy uses all quantal libs
[12:33] <seb128> xclaesse, good, you can apport-unpack that file, there is a coredump in it you can use gdb on
[12:33] <seb128> larsu, is there anything in empathy>help>debug?
[12:35] <larsu> seb128, a lot. I have the same crash report as xclaesse, though, apport-unpacking it now
[12:36] <xclaesse> good, got the backtrace into gdb
[12:36] <xclaesse> and it makes sense, tp-glib bug
[12:36] <xclaesse> that one: https://bugzilla.gnome.org/show_bug.cgi?id=678807
[12:36] <seb128> why is it happening today?
[12:36] <ubot2> Gnome bug 678807 in Chat "empathy-chat crashed with SIGSEGV in _tp_base_client_handle_channels()" [Critical,Needinfo]
[12:37] <xclaesse> seb128, it has been like that for me since ages
[12:37] <seb128> like, what changing if it started today for you guys?
[12:37] <seb128> xclaesse, oh ok
[12:37] <seb128> let's see if larsu has the same issue
[12:38] <xclaesse> seb128, given that we don't have millions of crash reports, my guess is it's tp-glib 0.19.1 regression
[12:38] <xclaesse> quantal just upgraded to that version, which explains why we are starting to get users angry :)
[12:38] <larsu> seb128, exactly the same.
[12:39] <xclaesse> I'm running precise's empathy but tp-glib master here
[12:39] <larsu> xclaesse, seb128, thanks for your help
[12:39] <seb128> ok
[12:39] <seb128> larsu, telepathy-glib got updated on friday so it's likely you just got it
[12:39] <seb128> xclaesse, will you follow up upstream about it?
[12:40] <xclaesse> seb128, I'm digging tp-glib code as we speak ;)
[12:40] <seb128> xclaesse, great, thanks
[12:40] <seb128> larsu, xclaesse said it doesn't happen if you run empathy-chat manually so I guess you can do that as a workaround
[12:41] <seb128> larsu, i.e run it by hand and then try to talk to somebody on empathy
[12:41] <larsu> seb128, it must have been held back then for me on the weekend - I only got it this morning
[12:41] <larsu> that's what I'm doing and it works great
[12:41] <larsu> thanks again!
[12:42] <seb128> larsu, xclaesse: thanks
=== zyga-afk is now known as zyga
[12:54] <mvo_> seb128: yes, I will check that later today
=== zyga is now known as zyga-food
[12:59] <seb128> mvo_, danke
=== slomo_ is now known as slomo
=== zyga-food is now known as zyga
=== tsimpson_ is now known as tsimpson
[14:12] <jbicha> cyphermox_: do you know how this should be handled? http://paste.ubuntu.com/1060801/ from https://launchpadlibrarian.net/108687682/buildlog_ubuntu-quantal-amd64.gnome-shell_3.5.3-0ubuntu1_BUILDING.txt.gz
[14:13] <cyphermox_> sure
[14:13] <cyphermox_> you're missing libnm-glib-vpn-dev
[14:13] <jbicha> ok, I did that and it didn't create a depend on libnm-glib-vpn1, does it need that?
[14:14] <cyphermox_> I don't think so, let me check
[14:15] <cyphermox_> err, maybe it should, actually
[14:15] <cyphermox_> seems to me like the pacakge is properly configured for it
[14:16] <cyphermox_> jbicha: otherwise it could be that gnome-shell is missing some configure switch to build and link against libnm-glib-vpn1/-dev; but it still checks for the package
[14:18] <jbicha> cyphermox_: it looks like that library is a dependency of nm-gnome so I don't think there's a need to explicitly depend on it
[14:18] <jbicha> thanks
[14:24] <cyphermox_> jbicha: well, probably but I do think it should show up anyway, if it doesn't there may be something funky going on
=== cyphermox_ is now known as cyphermox
[15:10] <didrocks> sil2100: is the dee override working?
[15:10] <didrocks> and how is the unity release going? :)
[15:18] <sil2100> didrocks: I'm building a testing new-trunk-unity package for myself right now, so far so good!
[15:19] <sil2100> didrocks: the dee package was building fine, but then I noticed it's not installing the python3 directory gir things, so I'm looking into that in the meantime
[15:20] <didrocks> sil2100: ok :)
[15:26] <seb128> didrocks, kenvandine, chrisccoulson, Ursinha, Laney, mlankhorst, cyphermox, mterry, Sweetshark, tkamppeter: it's meeting time in 5 minutes if anyone has a topic (none on the wiki so far), also please update https://wiki.ubuntu.com/DesktopTeam/Meeting/2012-06-26 with things you worked on this week for those who didn't write anything yet
[15:27] <mlankhorst> hey
[15:28]  * kenvandine waves
[15:28] <mlankhorst> nothing here, been working on prime stuff but it's a challenge not to make it deadlock :-)
[15:29] <cyphermox> seb128: added my work on porting software-properties to python3 last week
[15:30] <seb128> cyphermox, thanks
[15:30] <cyphermox> nothing really to discuss in the meeting, though I'm wondering if anyone is working on gweather 3.5
[15:31] <Laney> Some people might be interested a little script I wrote to help me be alerted of self-introduced breakage. It's in bzr branch lp:~laney/+junk/lp-subscribe uploads — it subscribes you to bug reports of packages you recently uploaded for a defined period (default 7 days)
[15:31] <cyphermox> gweather 3.5 has some API changes that evolution doesn't support, just fyi :)
[15:31] <Laney> :-)
[15:31] <seb128> cyphermox, check http://people.canonical.com/~platform/desktop/versions.html
[15:31] <seb128> cyphermox, seems not
[15:31] <cyphermox> right :)
[15:32] <seb128> cyphermox, if there is any reason to not work on it please open a bug, tag it desktop-upgrade
[15:32] <seb128> cyphermox, then note your comments in it
[15:32] <cyphermox> sure
[15:32] <seb128> cyphermox, so it's listed on the page
[15:32] <seb128> cyphermox, thanks
[15:32] <seb128> cyphermox, robert_ancell started doing that, opening bugs and putting status in them when they are blocked or anything like that, it helps visibility
[15:32] <cyphermox> I think I'll probably look at gweather myself and fix evo at the same time, if it makes sense to do so without spending too many hours
[15:33] <seb128> cyphermox, sounds good
[15:33] <seb128> Laney, could, that's something that was mentioned several times at UDSes as "it would be good if launchpad could do that" ;-)
[15:33] <Laney> it would indeed, but in the absence of that here's a script you can run from cron :P
[15:34] <seb128> hehe
[15:34] <pitti> seb128, Laney: FYI, I'm currently touching glib for the test suite stuff; want me to update to 2.33.3 as well while I'm at it?
[15:35] <seb128> pitti, that would be great ;-)
[15:35] <Laney> pitti: sure, I'm not attached and I guess you want to do it anyway to see if you can make the testsuite failure fatal
[15:35] <pitti> Laney: I already did that now
[15:35] <Laney> oh, neat
[15:35] <pitti> now let's see how much 2.33.3 fails :)
[15:36] <Laney> still, I'll let your CPU have fun building it
[15:36] <pitti> yeah, it takes quite a while
[15:36] <kenvandine> :)
[15:36] <pitti> https://launchpad.net/~pitti/+archive/ppa/+packages now has a glib-2.0 2.33.2 with ftbfs-on-test-fails
[15:37] <pitti> I grepped all Debian exp and Ubuntu build logs for current failures
[15:43] <sil2100> didrocks: so, about the dee thing
[15:43] <didrocks> yeah?
[15:44] <sil2100> didrocks: maybe one moment...
[15:45] <sil2100> didrocks: since hm, even though David added the python3 bindings to the makefile, the respective python3 directory in lib/ during package build is not created
[15:45] <sil2100> I only see things for python2.7
[15:47] <didrocks> sil2100: it's not installed? maybe you can patch the makefile.am to get it installed then,
[15:50] <sil2100> didrocks: I think they're not installed since when it finds python both 2.7 and 3, it only creates paths for 2.7
[15:52] <sil2100> didrocks: not sure how to force configure and makefiles to consider both at once
[15:57] <didrocks> sil2100: I'm sure you can figure that out ;)
[15:57] <didrocks> otherwise hardcode
[15:57] <didrocks> sil2100: I made some tests
[15:57] <didrocks> sil2100: it needs to be installed in python3
[15:58] <didrocks> not python3.2 (like it was python2.7)
[15:58] <sil2100> ACK
[15:58] <sil2100> (hardcoding sounds sooo nice... ;) )
[16:17] <mterry> didrocks, thanks for the reviews!  sorry about the left-in print statements
[16:17] <didrocks> mterry: you're really welcome. The print leftover is nothing seeing how much good work those branches are :)
[16:18] <mlankhorst> how long does it take for mirrors to pick up from -proposed?
[16:18] <mterry> didrocks, there's also the split-release-upgrader branch if you're feeling ambitious
[16:20] <didrocks> mterry: well, I saw Gary looked at the pre-requisite branch, so I thought he will get to it, he won't? :)
[16:39] <mterry> didrocks, for https://code.launchpad.net/~mterry/update-manager/split-release-upgrader/+merge/109211 the pre-req got merged
[16:39] <didrocks> mterry: ok, good :)
[16:39] <didrocks> I'll have a look
[16:40] <mterry> didrocks, there's a sister branch linked in that merge for the other side of things (the reverse split)
[16:41] <mterry> didrocks, these merges were trickier and I likely missed something in the packaging
[16:44] <didrocks> mterry: hum, interestood, I'll try to decipher what I can :)
[16:57] <xclaesse> from apport's .crash file, is it possible to get stderr output of the process?
[16:58] <xclaesse> didrocks, ^
[16:59] <didrocks> xclaesse: stderr should be printed on ~/.xsession-errors if it's not captured elsewhere
[16:59] <didrocks> xclaesse: so ship it with the apport hook for the package
[17:04] <xclaesse> didrocks, ok, it's not by default :)
[17:04] <didrocks> xclaesse: nope ;)
[17:06]  * didrocks waves good evening!
[17:12] <bryceh> seb128, it's still on my agenda, just that various bugwork and other projects have been occupying me
[17:29] <dupondje> something broken with GtkScales (Think that are the ones)
[17:30] <dupondje> can't click on them anymore to change position
[17:30] <dupondje> really need to drag them
[17:33] <seb128> dupondje, distro version? in what ui?
[17:34] <seb128> bryceh, hey, ok, I'm mostly checking what we can start crossing or postponing, you trend line is mostly flat so far, we need to readjust
[17:34] <seb128> bryceh, btw so after today I think we should drop or adjust the fr space bar fix
[17:35] <dupondje> Quantal :)
[17:35] <dupondje> seb128: take for example the sound options
[17:35] <dupondje> there you have "output volume" gtkscale
[17:35] <dupondje> try clicking :)
[17:35] <seb128> dupondje, that's a bad example
[17:35] <dupondje> why? :)
[17:36] <seb128> dupondje, that slider had hacks to jump at the click position which became default in the new gtk but broke the hacks
[17:36] <seb128> dupondje, try one which was not hacked to behave differently
[17:36] <dupondje> Rythmbox volume slider => broken
[17:36] <seb128> same
[17:36] <dupondje> Audacious song position slider => broken :)
[17:36] <seb128> typically the sound control were hacked this way
[17:36] <seb128> right
[17:36] <seb128> try a non sound related one
[17:37] <seb128> those will need to drop their hacks
[17:37] <seb128> like in gtk3-demo
[17:37] <seb128> dupondje, or try the mouse panel in system settings
[17:37] <seb128> dupondje, do they work in that one?
[17:38] <dupondje> those do indeed :)
[17:38] <seb128> dupondje, ok, so that's what I said, http://git.gnome.org/browse/gnome-control-center/commit/?id=9859ffbc27aca0a4f1a58412233498cdf846da70
[17:39] <seb128> dupondje, basically gtk before was doing "click on a slider moves on step and not where you clicked", which works not so well for play bar or sound controls
[17:39] <seb128> dupondje, so apps hacked around the behaviour
[17:39] <seb128> dupondje, gtk does the proper thing in 3.5 but that seems to break the hacks
[17:40] <seb128> dupondje, thanks for pointing it, we will look at those :-)
[17:40] <dupondje> so those hacks will need to be removed from the 'broken' software ?
[17:41] <seb128> dupondje, yes, similar to the url I just pointed
[17:41] <seb128> dupondje, those hack do "if left clicked then change the left event for a middle click one"
[17:41] <seb128> dupondje, or middle click stopped to do what they want with gtk 3.5 :-)
[17:46] <dupondje> hmz, can't find it directly in audacious code
[18:05] <dupondje> seb128: seems like audacious uses mpris
[18:05] <dupondje> could it be that is bugged ?
[18:07] <seb128> dupondje, hum, mpris is a norm to control players
=== zyga is now known as zyga-afk
[18:28] <dupondje> can't find it directly :(
[19:09] <seb128> bryceh, still around?
[19:09] <bryceh> seb128, yeah
[19:10] <seb128> bryceh, did you catch up on the xkeyboard-config activity from the day?
[19:10] <bryceh> seb128, nope, been tending to some other bugs, what's up?
[19:11] <seb128> bryceh, basically no SRU moved to updates yet, so the french space change is still only in proposed
[19:11] <bryceh> oh right, yeah I saw that
[19:11] <seb128> bryceh, I tried lucid in a vm today and rctrl and space in rhythmbox were both working
[19:11] <seb128> bryceh, seems like the patch we had by then was half of the precise one
[19:11] <seb128> bryceh, i.e we didn't add the level5 line
[19:12] <bryceh> seb128, seems the SRU acceptance is all out of order and messed up
[19:12] <seb128> bryceh, I think at this point we should drop the rb space patch (or use the lucid version) and do another upload which -v the uploads we got since precise ... what do you think?
[19:12] <bryceh> seb128, I thought slangasek said that the ubuntu2 got accepted into updates, ahead of ubuntu1, or is that wrong?
[19:13] <seb128> bryceh, that's wrong indeed
[19:13] <seb128> bryceh, he said that ubuntu2 got accepted in proposed before ubuntu1 moved to updates
[19:13] <bryceh> seb128, so you're saying none of the SRUs have gotten through to -updates?
[19:13] <seb128> bryceh, which means ubuntu1 couldn't move to update since it was deprecated
[19:14] <seb128> bryceh, correct
[19:14] <bryceh> feh
[19:14] <seb128> bryceh, cf https://launchpad.net/distros/ubuntu/+source/xkeyboard-config
[19:14] <seb128> bryceh, no precise-updates line
[19:15] <bryceh> heh I thought sru's were supposed to be getting *easier* this time
[19:15] <seb128> bryceh, they are, you got unlucky, usually if there is an update in -proposed the next one should be put on hold until the first one move to -updates
[19:16] <seb128> bryceh, somebody in the SRU team screwed and accepted the second one before that happened
[19:16] <bryceh> seb128, yeah I thought that's how it's supposed to work
[19:16] <seb128> yeah, somebody did a mistake
[19:16] <seb128> well, anyway at this point I think we shouldn't break rctrl for french users
[19:16] <bryceh> seb128, well compounding things is that hardly anyone bothers to confirm the fixes, so they sit there a long time
[19:17] <seb128> bryceh, that's not really true this cycle, the .1 team does efforts to verify things
[19:17] <seb128> well anyway that was screwed up, nothing we can change at this point
[19:17] <seb128> bryceh, so, we are back to "no SRU made it to -updates yet" and "we don't want to break rctrl for french users"
[19:17] <bryceh> seb128, that's good, will they be able to do that for the xkeyboard-config changes?  I put time into making the test cases especially well spelled out hoping they could.
[19:18] <seb128> bryceh, so I think we have 2 options
[19:18] <seb128> - use the patch we had in lucid which doesn't include the level5 change line
[19:18] <seb128> - drop the patch for that SRU round until it's sorted
[19:18] <seb128> bryceh, which one do you prefer?
[19:19] <bryceh> well, I'd still most prefer to get upstream's feedback first
[19:19] <seb128> bryceh, it's basically "reupload what you uploaded yesterday with a -v in the changelog to list previous upload and with one of those changes"
[19:19] <bryceh> er, dput -v just prints the dput version
[19:19] <seb128> bryceh, ok, can you reupload the version you uploaded yesterday with the french change from ubuntu1 dropped and using -v?
[19:20] <seb128> bryceh, that's a debuild argument, basically debuild -S -v2.5-1ubuntu1
[19:20] <bryceh> sure
[19:20] <seb128> it means "include in the changelog all the entries since the -v version"
[19:20] <seb128> that will make the bugs from both ubuntu1.1 and ubuntu1.2 listed on the -changes and SRU page
[19:21] <seb128> bryceh, thanks, I'm rejecting your upload from yesterday
[19:22] <seb128> bryceh, confirming the fixes from the testcases will not be an issue, all the bugs got confirmed, the current version would have moved to -updates without the rctrl issue
[19:27] <bryceh> seb128, done
[19:27] <seb128> bryceh, thanks
[19:33] <dupondje> seb128: I need to report a bug on the audacious/and others gtkscales issue ?
[19:33] <dupondje> or you'll check it out ?
[19:38] <Sweetshark> weather is cold and cloudy for summer in hamburg. "21:37:52 up  9:58,  0 users,  load average: 15.69, 17.29, 18.26" helps to keep a warm lap.
[19:40] <seb128> dupondje, better to open bugs
[19:40] <dupondje> on what package ?
[19:42] <seb128> dupondje, those which have the issue
[19:42] <seb128> e.g audacity
[19:42] <seb128> no need to report the gnome-control-center one, that's already fixed in git
[19:49] <dupondje> https://bugs.launchpad.net/ubuntu/+source/rhythmbox/+bug/1018087
[19:49] <ubot2> Ubuntu bug 1018087 in rhythmbox "Seek bar not working with left click" [Undecided,New]
[19:58] <seb128> dupondje, thanks
[19:58] <dupondje> if some more affected, I will add them :) just noticed those 2 now
[19:59] <Laney> dupondje: rhythmbox seems to be widgets/rb-header.c:883 and 952 if you want to fix that
[20:00] <dupondje> I can ... :) but maby better to report upstream?
[20:00] <Laney> yes, with a patch :-)
[20:18] <Laney> looks like totem is affected too
[20:20] <seb128> Laney, like anything with a sound or play slider
[20:21] <Laney> I guess it's a common hack
[20:25] <seb128> Laney, like I said before it's mostly multimedia players, other apps didn't bother so much
[20:26] <Laney> ok
[20:26] <Laney> dupondje: I'll file these two upstream with patches unless you are doing it already
[20:28] <dupondje> didn't do it yet :) so feel free
[20:28] <dupondje> couldn't find it in audacious :s
[20:56] <Laney> if only I could jhbuild totem successfully
[21:09] <dupondje> Laney: you add totem to the bug ?
[21:10] <Laney> nah, you can if you want
[21:10] <dupondje> i'll do, so the issue gets grouped
[21:41] <Sweetshark> kubuntu kicking libreoffice out of the seed? ah, great.
[21:43] <micahg> Sweetshark: it's a trial
[22:18] <robert_ancell> seb128, hey, did you get the syntax right for the versions cron job?
[22:18] <robert_ancell> It looks like every 6 minutes to me
[22:18] <robert_ancell> or is that 6 minutes past the hour
[22:19] <seb128> robert_ancell, hey, are you insulting me there? ;-)
[22:20] <robert_ancell> seb128, no, just trying to work out why it's not working
[22:20] <TheMuso> lol
[22:20] <seb128> robert_ancell, it was "06,36", I deleted the ",36", so I think it's every :06
[22:20] <seb128> # m h  dom mon dow   command
[22:20] <seb128> says the comment
[22:20] <seb128> so it should
[22:20] <robert_ancell> seb128, is there a log anywhere as to why the script failed?
[22:20] <seb128> 06m of every hour
[22:21] <seb128> robert_ancell, http://paste.ubuntu.com/1061672/
[22:22] <robert_ancell> aha
[22:22] <seb128> robert_ancell, log -> type "mail" on the server
[22:22] <seb128> robert_ancell, it's sending email but locally, you can put your email in the cron if you want to get directly emailed
[22:23] <robert_ancell> seb128, crappy old technology
[22:24] <seb128> robert_ancell, I never enabled that because it emails any output and at some point the script was printing stuff at every run so you would get an email every half an hour
[22:24] <seb128> robert_ancell, theother option is to just always print infos in a log...
[22:25] <robert_ancell> seb128, yeah, we should just update the cron job to > versions.log
[22:25] <seb128> robert_ancell, that's the easy way right
[22:35] <robert_ancell> seb128, did you have an opinion on the libgee package?
=== bjsnider_ is now known as bjsnider
[22:50] <robert_ancell> meeting in 10 mins, anyone have anything to say?
[22:52] <seb128> robert_ancell, no strong opinion on gee, what you commited to debian works for me
[22:55] <TheMuso> Nope.
[23:45] <RAOF> Quite a lot of my apps seem to be semi-regularly segfaulting in cairo_region_is_empty(); disabling the overlay-scrollbars appears to avoid these crashes. Has anyone else noticed this?
[23:50] <seb128> RAOF, qt applications?
[23:51] <RAOF> seb128: GTK applications; both GTK2 and GTK3 - evolution just crashed with that for me, and Tomboy and Monodevelop have crashed in the past with that trace.
[23:52] <seb128> RAOF, not know afaik, maybe kenvandine knows about it though
[23:53] <RAOF> Sadly apport crashed while trying to upload the crash report, so I don't know if it's hit launchpad :)
[23:53] <seb128> RAOF, report it again, worth case you will increase the score of your first report :-)
[23:55] <RAOF> ‘slice indices must be integers or None or have an __index__ method’??
[23:56] <RAOF> Please be not passing raw python exception messages through to user visible dialogs without any context :)