=== skaet is now known as skaet_afk
chrisccoulsonlol @ bug 74535000:38
ubot2Launchpad bug 745350 in firefox "while working on Firefox/Gmail, my cat sit down on keyboard " [Undecided,New] https://launchpad.net/bugs/74535000:38
* RAOF is intrigued.00:38
chrisccoulsonhe just seems to be reporting a bug to tell me that his cat sat on his keyboard00:39
TheMusoThat is amusing.00:42
ryanpgHi there, not sure if this is the right place to ask but... I installed gnome-shell from the gnome3 ppa, all went well except themes are broken, can't install gtk3-engines because it depends on libgtk3.0-0 and that breaks everything :) solutions?01:05
mimicoryanpg, possibly gnome-icon-themes-symbolic01:10
ryanpgmimico, I'm not finding that package... googling it now01:12
mimicomy bad01:13
ryanpgahh... well both that and gnome-icon-theme are installed01:14
ryanpgthe issue is that gtk3-engines seems to want libgtk3.0-0 and not libgtk3-001:15
ryanpgand installing libgtk3.0-0 means dumping 90% of gnome01:15
ryanpgsomeone over in gnome-shell on irc.gnome.org just told me that PPA isn't fully installable atm01:17
ryanpgnope, 32... I guess I'll just purge and do jhbuild01:18
ryanpgmimico, but thank you for your help! :)01:18
=== asac_ is now known as asac
psusiI'm a little new to the whole gnome/gui programming scene... I've always preferred working on kernels and daemons, but the last few days I've been trying to change that.  I have ended up fixing a bug in gnome-power-manager that made it stop respecting the gconf settings to lock the screen or not on suspend or hibernate02:27
psusiand I realized that these settings seem like they belong in the preferences gui rather than requiring the user to go digging around in gconf.  Is that correct?  Secondly, do you think this new page I added to the gui looks good or should it be done differently?02:28
RAOFpsusi: Is there some reason why they're not identically equal to “lock screen on screensaver”?02:29
* psusi is struggling to figure out photobucket02:32
RAOFUse Do's imageshack upload plugin!02:32
psusiRAOF, yea... I want my system to lock on hibernate, but not on suspend or general screen saver kicking in02:32
RAOFI mean - that seems like a pretty *specific* desire.  Is it something that enough people will/should want to do that it is usefully exposed in preferences?02:34
psusiso you think that most people either want to always lock when the screen saver kicks in, or never?  so why bother adding the checkbox to the preferences dialog?02:36
RAOFWhich checkbox?02:37
psusithat's what I've managed so far02:38
RAOFSo, yeah.  I don't think having those options exposed makes sense.02:40
psusiso you think it should just start the screen saver and either universally lock or not based on the screen saver setting, unless someone figures out how to go into gconf and set the more specific settings?02:41
RAOFBecause those preferences are extremely niche.  If you want some security (lock my screen) then why is hibernate different from suspend?  And you should obviously also lock the keyring in that case, and have the screensaver unlock it.02:43
RAOFI'll admit that lock on screensaver might be sufficiently different to lock on suspend/hibernate, though.02:43
psusiRAOF, yea, I think plenty of people don't want it to lock every time they don't touch the keyboard for 5 minutes02:44
psusiI was also thinking of maybe using a drop down list with the options to lock on [suspend, hibernate, neither, both], but it was easiest to add a checkbox tied to the existing bool gconf key02:45
RAOFI'd say there should be at most one checkbox.02:46
RAOFI'd probably side towards locking iff the user needs to enter a password at login.02:47
psusiif the options are there though, why hide them from the gui?02:55
RAOFIt's a tradeoff for users like you who want to do something strange :)02:56
psusiI was thinking about that the other night... it seems to be part of the gnome philosophy right?  if it's kind of esoteric, don't bother exposing it in the gui, make people use gconf?02:56
psusiseems like what we need is a secret keystroke to enable the gui equivalent of --verbose... something you can hit to say show me all the goodies! ;)02:57
RAOFThat's the project you're thinking of :)02:57
psusihrm... not finding much on it so far...03:00
psusidoesn't look like it ever got off the ground ;(03:00
RAOFI think there's at least *some* code, but you're right, it's not exactly taking the world by storm.03:03
psusimany applications have an "advanced" settings button to put all the esoteric stuff behind03:04
psusiit seems like gnome doesn't like that idea03:04
RAOFRight.  It clutters things up, encourages making options where fixing a problem would be better, and *everyone* hits the advanced settings button at some point.03:05
psusihow does it clutter things up?  it reduces clutter by hiding esoteric options unless you really want them?03:06
RAOFWell, it's an extra tab / button that doesn't have an obvious reason.03:06
RAOFIt's a grab-bag of options that would naturally be elsewhere, but aren't because they're not important enough.03:07
psusiyea, but one button is a very small amount of clutter to continue to allow access to many options hidden behind it03:07
psusiI agree, rather than gather them all under one place, there should be a way to just make them show up in their natural position03:08
psusia sort of global --verbose or I'm an advanced user option that makes them show up all over rather than in a tightly constrained grab bag03:08
kklimondameh, you are talking like a real kernel hacker :}03:09
psusinot surprising ;)03:09
psusiI'm still getting used to this whole gui thing ;)03:09
kklimondathe magical button that toggles a number of options in the interface sucks - I've seen software that does that, and it's not pretty.03:10
psusithe underlines on accelerator keys are hidden until you hold down the alt key, maybe something like that could be done for more esoteric options03:10
psusiwhy isn't it pretty?  and more importantly, how is it worse than forcing people to dig around in gconf? ;)03:11
kklimondasame for the --verbose flag - the problem with adding more buttons, checkboxes, tabs, and text entries is that they really clutter interface, and only a small minority of users is going to ever need them03:11
psusiwell that's why they aren't shown normally03:11
kklimondayeah - but they have to be there, they have to be designed, and taken into account.03:12
psusiit's one thing if the devs don't feel like taking the time to add an option to the gui preferences screen because it is so esoteric... but if someone wants to take the time, the question is, how to do so that both avoids clutter for most people, but still provides the option for advanced users03:13
kklimondapsusi: but advanced users have gconf - there is nothing hard about it..03:14
kklimondapsusi: if something is popular enough there are tools like ubuntu-tweak or gnome-tweak (new tool in gnome3) that can provide ui for that.03:14
psusikklimonda, ADVANCED users do, yes... but... well... not quite so advanced users don't ;)03:14
psusithose struggling in the middle are screwed03:14
kklimondamiddle class is always getting shorter end of the stick :)03:15
kklimondaIt's simply impossible to create something that fits all needs so we have to compromise. GNOME has always been about creating simple, and usable desktop.03:16
kklimondaKDE3 went the other way03:16
kklimondaand I still remember applications that had 3 different configuration dialogs :)03:17
lifelesskklimonda: actually, gnome /became/ about that; it wasn't initially03:18
psusithere should be a way to get the best of both worlds03:19
psusithat's why I started thinking of a magic key you can smack to verbose++03:20
kklimondalifeless: right, I've oversimplified - I've meant GNOME 2, which has introduced HIG03:21
psusithat way it isn't clutter unless someone thinks to themselves, gee, I'm looking for something that seems like it should be here, I wonder if it's just hidden by default?03:21
kklimondapsusi: you will love the fact that in GNOME3 you can't even change Gtk+ theme without running gnome-tweak ;)03:22
psusiI've never even looked at gnome-tweak03:22
kklimondait's a new project, created for GNOME 303:23
psusiohh, I was thinking ubuntu-tweak03:23
lifelesskklimonda: metacity 4 eva03:23
kklimondait's called gnome-tweak-tool actually03:23
kklimondapsusi: you should really talk with mpt, or some other ux folks - they can explain the drawbacks of "advanced" button better.03:25
psusikklimonda, I agree that it isn't good to lump all esoteric options under an advanced button... but there should be some way to tell the system that you are looking for more options and don't care if it adds clutter... telling users to use gconf just seems like a cop-out03:27
psusithe magical "zoom in and enhance", hehe...03:29
* psusi asks for OPONIES while he's at it03:29
kklimonda-ENOPONIES :P03:30
lifelessso the whole thing about options is often misrepresented03:31
lifelessthe thing is that *if* you can make it better without an option, that is clearly a win: simpler code, less defects, less maintenance03:31
lifelesseasier for users to train each other because there is less variation03:31
psusiin other words, don't make it an option if it is a universally good change03:32
kklimondathere is no such thing like an "universally good change" but noone has said that you have to please everyone.03:32
RAOFAnd also *push back* on options before you've been convinced that it can't be done better.03:33
kklimondaright, once you have introduced some new setting, it's pretty much impossible to remove it :)03:33
lifelessjdubs recent series of posts about gnome culture have a pretty good analysis of this meme03:34
kklimondayou would have to rename your application, or rewrite a huge chunks of it.03:34
psusithat seems like an argument for changing/removing a feature rather than converting it to an option03:34
lifelesskklimonda: thats not true at all03:34
lifelessif you remove an option you need to be willing to wear the feedback03:34
kklimondalifeless: feedback, not the outcry for restoring it, and threats of forking your application.03:35
psusibut if you decide to make it an option, why should it NOT be accessible without going into gconf was more the issue I was contemplating03:35
lifelesskklimonda: depends on the option obviously03:35
lifelesskklimonda: but if you've got *that many objectors* its pretty good data about the importance of supporting their use case.03:36
psusilock on screen saver vs suspend vs hibernate is already an option, my question is, why is there no way to configure it in the g-p-m preferences dialog?03:36
RAOFIt's a cop-out, basically.03:37
kklimondapsusi: I see options that are not available in the gui as situational tweaks.03:37
kklimondapsusi: by your definition we should provide gui for g-p-m to switch between suspend on the remaining batter %, or the remaining time03:38
kklimondapsusi: and then additional controls to tweak the numbers03:38
lifelesskklimonda: thats not what psusi argued at all though03:39
kklimondalifeless: but you have to put the line somewhere.03:39
lifelesskklimonda: its not about the line03:39
lifelesskklimonda: its about asking how we can improve the syste,03:39
psusikklimonda, if it's an option you can set via gconf, yes, why not?03:40
lifelessfor instnace, on the time vs % issue, I don't see any use case for a % based limit *unless* time based limits are unreliable.03:40
lifelessAnd as it turns out, time based limits are terribly unreliable.03:40
lifelessg-p-m, for all its 500MB footprint, is terrible about remaining runtime estimation on my hardware.03:41
lifeless(and btw, 500MB for a battery manager? WTF!)03:41
kklimonda500M of what?03:41
psusihuh?  it's 10mb03:41
lifelessvirt, resident is 150MB03:42
lifelesssorry, 200MB03:42
lifeless 2354 robertc   20   0  448m 198m 2680 S    0  2.6   0:55.31 gnome-power-man03:42
kklimonda1000       850  0.0  0.7  78392 29164 ?        Sl   Mar29   0:00 gnome-power-manager03:42
kklimondaand I blame nvidia for that03:42
psusilifeless, rss here is 8.7m03:43
RAOFMine's 5.5m03:43
lifelessRAOF: want a bug ? :P03:43
lifelessRAOF: I just figured it was terrible03:43
kklimondawell, 500M is terrible :)03:44
lifelessyes, yes it is :)03:46
RAOFI'm not sure what to do with such a bug; a pmap might be interesting,though.03:46
kklimondathere are lots of settings in gconf that has probably been introduced years before, and now they can't remove them as they are used.03:46
lifelessRAOF: I can ubuntu-bug it up, if you like03:47
psusivsz is in general, ridiculously huge, and should be ignored.03:47
kklimondathere are keys like /apps/gnome-search-tool/disable_quick_search_second_scan with description "This key determines if the search tool disables the use of the find command after performing a quick search."03:47
kklimondaor even /apps/gnome-search-tool/show_additional_options heh03:47
RAOFlifeless: You might as well, attaching pmap -x $(pidof gnome-power-manager); it's certainly a bug!03:48
psusimy g-p-m has 8792 rss and 176984 vsz... I've long given up on even looking at the vsz of gnome apps03:49
lifelessRAOF: shouldn't that be a apport hook ?03:51
RAOFlifeless: It's not common information to ask for.03:52
psusihell, x-chat has a vsz of 426m but only 30.9m rss03:52
lifelessadmittedly, its not as bad as chromium03:52
lifeless740M for one web page03:53
lifeless 3085 robertc   20   0 1642m 739m 3100 S    0  9.6   1:10.51 chromium-browse03:53
psusilifeless, yep... similar results here.. hence, ignore vsz, it's meaningless03:54
lifelesspsusi: its only meaningless when its shared libraries03:54
lifelesspsusi: when its swap, its meaningful03:54
psusiwhen it's shared anything03:54
RAOFIt's annoying how difficult it is to distinguish between the two.03:54
psusiI've got no swap, but thunderbird is showing 858m virt, but only 139 rss and 30.9 shared03:55
lifeless0000000000942000       0  194932  194256 rw---    [ anon ]03:55
lifelessRAOF: ^03:56
RAOFSomething is horribly broken.  Valgrinding it might reveal leaks, I guess.  Does it start out so big, or grow so over time?03:57
ubot2Launchpad bug 745420 in gnome-power-manager "high res (and virtual) memory sizes" [Undecided,New]03:57
RAOFPlease be to not dirtying 200MiB of memory.03:57
RAOFHm.  In other “why are you so big” news… indicator-datetime-service probably shouldn't be using 270MiB RSS.03:59
lifelessRAOF: pmap is there04:00
lifelessRAOF: I'm not sure if it grows or not04:01
RAOFThat should be pretty easy to check, though; just restart it?04:01
lifelessRAOF: sure04:02
lifelessbut if there is stuff you want from the current process, should get that first04:02
RAOFNot that I can think of; my next step would generally be a valgrind check.04:03
lifelessRAOF: fresh started instance pmap attached to the bug04:06
RAOFRight.  So, it's leaking over time.04:07
lifelessRAOF: it seems to stabilise at 500m04:10
lifelessRAOF: so I suspect something like a cache04:10
lifelessRAOF: (500MB virt; 200M rss)04:10
RAOFThat could be its time-to-empty/charge history feature happening.04:11
lifelessI swap batteries from time to time, and go on off power a lot as I wander around04:11
RAOFI guess you could check natty?  I'll see if I can reproduce that by wandering around on/off power.04:12
lifelessRAOF: I should upgrade, just been afraid04:23
RAOFYou can be a beta 1 tester :)04:23
=== skaet_afk is now known as skaet
=== skaet is now known as skaet_
TheMusoRAOF: Is there a way you can query X as to how much VRAM is available? I'd like to get an exact figure of how much VRAM my Thinkpad's Intel GPU is taking from system RAM.06:18
TheMusoMore out of curiosity than anything else, the notebook has tons of RAM.06:19
brycehTheMuso, there is xrestop which shows pixmap memory usage, which isn't total vram but it might give some insights06:21
RAOFTheMuso: I'm not sure that the question actually makes sense any more; I'm not sure that the GPU actually steals memory.06:21
RAOFRather it'll dynamically acquire and release memory as necessary.06:22
TheMusoRAOF: Ah, that makes sense.06:22
RAOFCertainly before GEM/KMS it stole a fixed chunk of memory, but I don't *think* that's the case anymore.06:22
TheMusobryceh: Thanks.06:23
brycehTheMuso, cat /sys/kernel/debug/dri/0/*vram* shows some vram related data on my radeon.  ymmv06:24
bryceh# cat /sys/kernel/debug/dri/0/*vram* | tail -n 106:25
brycehtotal: 131072, used 37553 free 9351906:25
TheMusoOk thanks.06:25
* RAOF investigates what's new in intel /sys06:25
RAOFOh, that's right.  /sys/kernel/debug is now restricted.06:26
RAOF /sys/kernel/debug/dri/0/i915_gem_objects are what you're after.06:27
TheMusoah ok06:27
RAOFI have ~500MiB worth of objects there, most of which are inactive.06:28
TheMusoI like the fact that the amount of RAM used for intel GPUs is now dynamic.06:30
TheMusoAt least in Linux06:30
TheMusoOne can only hope Windows does the same...06:31
RAOFI'd presume so; one of the prerequisites for DX 10 hardware/drivers is a sane memory manager.06:31
TheMusoAh ok.06:31
TheMusoThis is weird, in /sys/kernel/debug/dri I have 0 and 64.06:32
RAOFI'm not entirely sure what the 64 is to be honest :)06:33
TheMusoWell I get the same on my desktop with my radeon card.06:34
TheMuso0 and 64.06:34
TheMusoSpeaking of Windows, one really nice thing that the Thinkpad power management software does is completely disable/hot unplug the optical drive when not in use for a period of time, to the point where Windows doesn't know its there, but power is still being fed to it.06:35
TheMusoSo you press the eject button, and windows sees it again and you can use it.06:35
TheMusoReally neat.06:35
RAOFThat's pretty cool.06:36
RAOFI'm in the market for a new laptop, but no-one seems to make the laptop I want :)06:37
TheMusoWhat do you want?06:38
RAOFBasically: I'd like a 13" macbook with a faster processor, discrete graphics card, and higher resolution screen :)06:40
RAOFThere are any number of things which satisfy some subset of these critera, but nothing that satisfies all ;)06:41
* TheMuso nods.06:42
TheMusoNothing in the lenovo range?06:42
RAOFOh, and a multitouch screen would be nice for actually, you know, *testing* the multitouch stuff Chase wants me to upload :)06:42
RAOFThe x220 is promising, but no discrete chip.  The new t420s might be what I'm looking for, although that's a bit bigger at 14" and has an nvidia chip rather than the ati chip I'd prefer.06:43
TheMusoNice to know the discrete chip you want is AMD.06:43
RAOFThat's not *entirely* because I expect it to work better; it's because it's the chip I'm most likely to be able to get wine working on with free drivers. :)06:44
TheMusolol ok.06:44
RAOFI'd also expect it to work better out of the box, though.06:44
TheMusoWhat are you wanting to use in wine?06:45
RAOFJust have worse proprietary drivers.06:45
RAOFCiv 5, Portal 2, that sort of thing…06:45
TheMusoOk so you want to play games, fair enough.06:45
TheMusoThings that are generally rather GPU intensive.06:45
RAOFThere 'aint any other windows software I'm pining for!06:45
TheMusoWell I didn't know that. :p06:46
TheMusorobert_ancell, RAOF, did either of you manage to see Four Corners on Monday night, about the A380 fault from last year?06:47
RAOFOh, no I didn't.  That sounds quite topical!06:51
RAOFIt should be on iview, though, so I'll watch it sometime!06:51
robert_ancellTheMuso, nope06:51
robert_ancellTheMuso, interesting?06:51
TheMusoYes, they went through and talked about exactly what happened.06:54
TheMusoIncluding a rather elaborate re-enactment.06:54
TheMusoYes, I suggest watching/downloading from iview.06:54
pittiGood morning07:03
pittibryceh: FYI, fglrx is on the DVDs, coordinating with #u-release to accept it07:17
tjaaltonpitti: huh?07:18
pittiin ship, no in the live session07:18
tjaaltonnvidia too?07:19
brycehpitti, ah right, forgot07:19
pittiI don't think it warrants rebuilding the DVDs, though; it's just in ship, not in the live session or anything07:20
pittii. e. if people install it, they should get it from the archive07:20
pittibryceh, tjaalton: ok, accepting07:22
pittibryceh: I guess this package still has the ABI dep problem on amd64?07:22
AmaranthRAOF: btw iirc DX 11 requires a sane memory manager but the requirements were relaxed for DX 10 because nvidia couldn't get going in time for vista07:22
* pitti -> vaccination, bbl07:22
RAOFAmaranth: nvidia?  I thought it was intel (cf: Vista Ready™ lawsuit)07:25
AmaranthRAOF: No, intel wanted a way for non-DX10 parts to be vista ready07:26
RAOFRegardless *some* version of DX requires a sane memory manager.  Oh, but if it's DX11 then there's no guarantee that intel has one; they have no DX11 parts.07:26
RAOFAmaranth: Ah, of course :/07:26
=== skaet_ is now known as skaet_afk
AmaranthRAOF: but back when we were having so much trouble with nvidia memory management and compiz I learned they didn't have it fully done on Windows either07:27
RAOFYeah, that's not really something they'd needed to care too much about before.07:27
AmaranthRAOF: Remember everyone complaining about nvidia drivers when Vista was released? They were slow, buggy, etc07:28
SweetsharkMorning all.07:28
Amaranthmorning Sweetshark07:28
RAOFGoooood morning.07:28
ricotzrobert_ancell, hello07:47
robert_ancellricotz, hey07:48
ricotzdid you had a chance to look at libwnck3, basically it is what you started07:48
robert_ancellricotz, I'm fine with the wnck package btw, please upload07:48
ricotzrobert_ancell, ok07:49
desrtrobert_ancell, ricotz; either of you guys know about this intense background flickering problem?08:00
robert_ancelldesrt, in?08:00
desrton login or setting the background from the control centre, the background flickers very rapidly between the desired image and complete blackness for a couple of seconds before randomly settling on one or the other (ie: sometimes the background, sometimes blackness)08:01
robert_ancelldesrt, in the GNOME3 PPA?08:01
ricotzdesrt, hmm, no, probably a driver issue?08:01
desrtit doesn't seem like a driver issue because the problem persists08:02
ricotzdesrt, you were able and updated everything?08:02
desrtricotz: i'm avoiding upgrades at the moment in order to reduce the number of things i accidentally pull in from your ppa :)08:02
desrtbut i'm up to date as of yesterday08:03
desrt(it was happening before then for a day or two)08:03
=== warp11 is now known as warp10
desrtit almost looks like two things are fighting with each other to have the background be black or not08:03
ricotzdesrt, is this on nvidia blob?08:06
desrt+1 for awesome use of the word 'blob', though :)08:07
didrocksgood morning08:08
desrtdidrocks: sup?08:08
didrockshey desrt, do you enjoy your trip? :)08:09
desrtya.  it's going pretty nicely so far.08:09
desrthaving fun with some of the PPAs at the moment08:09
ricotzdesrt, yeah, best way to name it ;) -- cant find any bug reports for intel, but nvidia had a similar problem08:09
desrtricotz: if we've seen the problem on nvidia and intel as well then maybe it's not a driver problem08:10
ricotzdesrt, i am on 270.30 without problems though08:10
desrthm.  interesting.08:10
desrtdo you still experience the problem with slowness when notification icons are present?08:10
didrocksdesrt: excellent. That's everytime the same kind of "fun" with drivers and such :)08:10
ricotzdesrt, yes08:11
desrtso the weird thing is that i never had it before and then after upgrading maybe 4 days ago i had the problem, and have it 100% of the time08:11
desrti should check my apt log...08:11
ricotzdesrt, maybe a clutter bug then08:12
desrtit could have been really a lot of things08:14
desrtthere was a massive number of upgrades08:14
desrtmaybe it was libpcre08:14
desrtwould be nice if we had apt-bisect :)08:15
desrthad an X server upgrade here too08:16
ricotzdesrt, there seems to be workaround for the nvidia issue, perhaps it solves intel too http://bugzilla-attachments.gnome.org/attachment.cgi?id=15732608:16
ricotzdesrt, you are running the x-stack from natty repo?08:17
desrtnothing fancy there08:17
desrtricotz: i'll try this patch.  give me a moment.08:17
ricotzjust apply it and restart gnome-shell should be enough08:18
desrtya.  that's what i'm doing08:18
desrtnot fixed08:18
desrtit was just changing the priority to 500, right?08:18
ricotzlooks so, yes08:19
desrtit's an a substantially different offset in my version of the file08:19
ricotzdoes the flickering effect the whole screen or only a part of it?08:19
desrtjust the background08:19
desrtand only when trying to set the background08:20
ricotzhave you turned on the nautilus-desktop?08:20
desrtmaybe i should try that, though :)08:21
desrtis that a dconf key or something these days?08:21
ricotzi was thinking if you are using it, it might be the cause08:21
ricotzuse can the gnome-tweak-tools08:21
desrtah.  i've heard rumours about this mythical tool08:22
ricotzyou can start it right out of the source folder08:22
* desrt hopes it is small08:22
desrthaving nautilus draw the desktop fixes it08:23
desrti love tweak tool!08:24
ricotzhmm, so i might need to turn it off then to see the problem :P08:24
ricotznope, works fine08:26
desrti'm tempted to leave nautilus rendering the desktop so i don't see the problem anymore :p08:28
pittihey seb12808:29
seb128hellio pitti08:30
seb128how are you?08:30
pittiseb128: pretty good, thanks!08:31
pittihow about yourself?08:31
* pitti prepares for a mixed release / patch pilot day08:31
seb128pitti, I'm fine thanks, ready for a mixed iso testing, bug triaging day08:36
seb128ok, pitti just beat me to the g-c-c button id thing while I was reviewing the commits08:51
seb128pitti, will you commit to the packaging vcs as well?08:51
pittioh, sorry for overlapping then08:51
pittiseb128: already at it08:51
pittiyou are too fast!08:52
seb128pitti, btw the gtk bug just assigned to our team is an indicator issue and fixed in natty08:52
pittioh, awesome, I didn't get to that one yet08:52
seb128pitti, btw do you watch all desktop bugs?08:52
pittiseb128: I don't, but I'm subscribed to oem-priority bugs08:53
seb128just wondering how you noticed the gcc one since it was not assigned to the team or anything08:53
seb128oh ok08:53
seb128it makes sense ;-)08:53
pittiI can't keep up with them all, I'm afraid -- I don't have seb128 powers08:53
didrockspitti: (it's blackmagic) :) hey pitti!08:54
* pitti hugs didrocks08:54
* didrocks hugs pitti08:54
* pitti hugs seb128 as well08:54
pittiseb128: and I need 4 more bug fixes to catch up with mvo :)08:56
* pitti pats https://bugs.launchpad.net/~pitti/+assignedbugs?field.status=Fix+Committed -- take that, mvo!08:57
* pitti hugs mvo08:57
* seb128 hugs pitti08:57
* pitti gets into the plane seat08:58
seb128pitti, hum08:59
seb128pitti, happy piloting! ;)08:59
pittithanks :)08:59
pittiseb128: what's the "hum"?09:00
seb128pitti, nothing, I was about to make a comment about bug fixed and didrocks but decided to not ;)09:01
pittiI've given up trying to chase him :)09:01
didrockshey rodrigo_09:03
seb128hey rodrigo_09:04
seb128we get quite some pygtk UnicodeDecodeError crashes nowadays09:04
pittididrocks: looking forward to another 43 from https://bugs.launchpad.net/unity/+bugs?field.status=Fix+Committed :)09:04
seb128it's weird09:04
seb128did something change in regard of encoding, utf handling?09:04
pittiseb128: yeah, I fixed another bunch of them yesterday09:04
pitti2.7 changed the unicode handling a bit to be a bit closer to 3.009:04
pittibut some things actually work better now09:05
mvopitti: oh? time to get into gears it seems ,)09:05
didrockspitti: we decided to not stop at 42, still 2 days to go! :-)09:05
mvo(for me)09:05
pittimvo: well, you are already -- you are first on the list (not counting ueber-didrocks)09:06
seb128mvo, hey, wie gehts?09:06
seb128pitti, bah, bug #745022, seems the bug pattern is not working?09:06
ubot2Launchpad bug 745022 in apport "apport-gtk assert failure: python: ../../src/xcb_io.c:462: _XAllocID: Assertion `ret != inval_id' failed." [Medium,New] https://launchpad.net/bugs/74502209:06
mvoseb128: danke, gut!09:06
pittididrocks: bug 737467  perhaps? :-)09:07
ubot2Launchpad bug 737467 in compiz "compiz crashed with SIGSEGV in g_closure_invoke()" [Medium,Confirmed] https://launchpad.net/bugs/73746709:07
seb128pitti, bug #738939 is weird but seems to affect some users09:07
ubot2Launchpad bug 738939 in apport "apport-gtk crashed with TypeError in __init__(): must be a subtype of GObject" [Undecided,New] https://launchpad.net/bugs/73893909:07
pittiseb128: looking09:07
pittiseb128: (the bug pattern one first)09:07
ricotzseb128, hi, are icon theme updates still possible?09:08
didrockspitti: should be fixed with the bunch of signal and handler disconnection that are already landed or about to lan :)09:08
pittiseb128: ah09:08
seb128the second one is not specific to apport, jockey has similar ones09:08
pitti<!-- bdmurray - 2011-03-23 disabling temporarily to see if it is still happening on natty.09:08
seb128pitti, oh ok, so yes, it is ;)09:08
seb128ricotz, define icon theme09:08
seb128ricotz, gnome-icon-theme-symbolic for example can be updated09:09
pittiseb128: bah, I can't even open the original bug any more09:09
seb128ricotz, we will not likely update g-i-t though09:09
ricotzseb128, elementary-icon-theme09:09
seb128ricotz, guess it can be updated, I've no clue about it but it's not on the default installation09:10
seb128so should be easier to get a freeze break approval09:10
ricotzseb128, yes the symbolic theme can be synced09:10
seb128ricotz, right, it's on my list of things to sync after the beta freeze09:10
ricotzelementary-icon-theme is the default for xubuntu and lubuntu09:10
ricotzseb128, ok09:11
seb128ok, so check with them09:11
seb128you will likely need a freeze break approval for the update with a rational09:11
ricotzseb128, alright09:11
seb128pitti, do you have any clue about that nautilus crash btw?09:17
pittiseb128: it's next on my list, but haven't had time for it yet09:17
pittiI don't seem to get it, so I'll need to play around with it (or stare hard at the trace and patches)09:18
pittiseb128: it's now the only "in progress" bug on my list, though09:18
seb128pitti, ok, rodrigo said he would have some extra time to help on some GNOME bugs now that he's done mostly with unity work09:18
pittijust patch pilot got in the way today09:18
seb128pitti, so maybe bounce him his way if you want09:18
pittioh, feel free to steal, of course :)09:18
seb128rodrigo_, bug #740765 if you want to have a go to this one09:19
ubot2Launchpad bug 740765 in nautilus "nautilus crashed with SIGABRT in __kernel_vsyscall()" [High,In progress] https://launchpad.net/bugs/74076509:19
didrocksdesrt: is there a convienent hidden function to transform a bitmask to a gvariant? :)09:19
rodrigo_seb128, ok09:19
pittirodrigo_: if you have some time, would you like to have a look at bug 707705?09:19
ubot2Launchpad bug 707705 in vino "Vino not accepting incoming connection to more than a single desktop session when the "network_interfaces=lo" option is set" [Medium,Triaged] https://launchpad.net/bugs/70770509:19
rodrigo_pitti, yes09:19
pittirodrigo_: that one is rather important for OEM09:20
desrtdidrocks: array of booleans, or something?09:20
pittirodrigo_: that would be appreciated09:20
seb128rodrigo_, ok, let's put those 2 on your buglist to start then09:20
rodrigo_pitti: are you working on https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/740765 ? it's in progress and assigned to you, or can I take it?09:20
ubot2Launchpad bug 740765 in nautilus "nautilus crashed with SIGABRT in __kernel_vsyscall()" [High,In progress]09:20
seb128rodrigo_, that's what I was just asking before pinging you09:20
pittirodrigo_: TBH I think the other is more important; I think I can deal with the nautilus one (just not today, but we're frozen anyway)09:21
seb128if you read the backlog ;-)09:21
didrocksdesrt: yeah, but that means I have to iterate to every bit in the mask, hence I wanted to know if there was already that in gvariant to avoid rewritting this. Seems not, so I'll, thanks! :)09:21
seb128rodrigo_, ok, start with the vino one then please09:21
seb128something since yesterday is breaking my keypad, annoying09:21
desrtdidrocks: i guess it will require about 4-5 lines :)09:21
seb128I swear it was working when I logged in09:22
desrtdidrocks: why not send the bitmask as a uint, though?09:22
pittiseb128: oh, wait -- I debugged that with claire the other day09:22
* pitti digs out IRC logs09:22
pittiseb128: does gconftool --recursive-unset /desktop/gnome/acces09:22
pittiseb128: does this help: gconftool --recursive-unset /desktop/gnome/accessibility09:23
didrocksdesrt: I want to output it through dbus as a debug statement without firing up gcaltool each time to convert it back ;)09:23
didrocksdesrt: ok, will iterate then!09:23
pittiseb128: it took me 2 hours to get to that when debugging claire's broken keypad after a maverick upgrade09:23
pitti(the upgrade apparently was unrelated, just coincidence)09:23
desrtdidrocks: makes sense09:24
desrtdidrocks: something that might be neat is an API in GObject to convert a registered flag type to a strv of the nicks of the individual bits that are set09:24
seb128pitti, ok, thanks, it was mousekeys_enable set to ture09:26
seb128I've turned a11y on and off the other day to try a bug but I didn't think it would have been related to the keyboard issues09:27
seb128pitti, you just saved me hours of head scratching ;-)09:28
seb128pitti, ;-)09:28
didrocksdesrt: yeah, we are using enums right now, but that would be neat :)09:28
kamstrupmvo: good morning, have you seen https://bugs.launchpad.net/oem-priority/+bug/745243 ?09:46
ubot2Launchpad bug 745243 in unity-2d "[dash] wrong search result in Chinese" [Critical,New]09:46
kamstrupmvo: i was wondering if you had some working knowledge in xapian vs ibus/chinese..?09:47
pittididrocks: btw, do you know about lp-set-dup <master> <bug1> <bug2> ..?09:52
pittididrocks: it's very handy for duplicating a lot of bugs (which have dupes themselves), and isn't affected by timeouts09:53
didrockspitti: oh no, I'll do that then, thanks a lot (no right now as I want to finish something first and the list of dup is long) :)09:53
chrisccoulsongood morning everyone09:53
pittihey chrisccoulson09:55
chrisccoulsonhi pitti, how are you?09:55
pittichrisccoulson: I'm great, thanks!09:55
pittichrisccoulson: hey, *shhh*, want something sponsored? only five dollahs09:56
seb128hey chrisccoulson09:56
chrisccoulsonheh :)09:56
chrisccoulsonhi seb12809:56
seb128pitti, you can also directly open the duplicate page for the bug which has not timeout issue09:56
pittididrocks: I can't do it myself, as you typoed the master bug (it's not 74167)09:57
didrockspitti: don't worry, I'll do it :)09:58
pittiseb128: so you are saying bug 551809 is fixed in natty in some indicator bits? do you happen to know where?09:59
ubot2Launchpad bug 551809 in gtk+2.0 "gnome-settings-daemon crashed with SIGSEGV" [High,Triaged] https://launchpad.net/bugs/55180909:59
pittiseb128: (I don't see indicator stuff in the trace)09:59
seb128pitti, the most recent libappindicator upload10:00
pittiseb128: ah, thanks10:00
pittiduping to bug 729150 then10:01
ubot2Launchpad bug 729150 in libappindicator "libappindication crashes in gtkstatusicon code on update" [High,Fix released] https://launchpad.net/bugs/72915010:01
seb128right, you can read the bug if you are interested10:01
seb128but it might be a bug in gtk as well, they workaround it now though by stopping using gicon10:01
seb128pitti, it's happening basically for things using the notification area fallback (i.e when no indicator applet is in use) when the icons are updated10:02
seb128 10:04
seb128kamstrup, hey10:04
seb128kamstrup, so the gwibber account desktop is named "compte de microblogage" in french10:05
seb128kamstrup, but when I type "blog" in the application place it has no match, is that a bug?10:05
seb128or you don't do substring matching this way?10:06
kamstrupseb128: no substring matching10:10
kamstrup(and I think that's a feature :-))10:10
seb128kamstrup, well, having nothing matching "blog" sucks ;-)10:10
seb128kamstrup, what you basically say is that it should be adressed by adding keywords10:11
kamstrupseb128: indeed. the .desktop metadata sucks in general, but that's part of a much bigger problem10:11
kamstrupthe .desktop files where just not designed to have search friendly metadata10:11
seb128why don't you do subtring searches?10:12
kamstrupI think substring matching works for apps like Do and Synapse because the show *1* result. When you show many results they will just flood you with useless junk because there are so many matches10:12
AmaranthYou could only do substring matching if you're not getting many matches normally10:13
kamstrupsubstring matching on 2,4k apps from the software center is not gonna be fun on a netbook10:15
kamstrupalso seems that substring matching is just a band aid because the .desktop metadata is so bad10:16
kamstrupband aids are ok, if it improves the UX of course10:16
seb128well GNOME3 has a Keywords key in their .desktop to solve that issue10:17
kamstrupbut my gut tells me that substring matching is a can of worms10:17
seb128so let's wait for next cycle10:17
AmaranthWhat are you searching now? The file name and the Name and Comment keys?10:17
seb128kamstrup, thanks10:17
kamstrupname, generic name, exec, and comment10:17
kamstrupwith ranking boosted descendingly in that order10:17
seb128pitti, wth?10:21
seb128pitti, how did you duplicate that g-s-d crash bug?10:22
pittilp-set-dup 729150  55180910:23
seb128pitti, https://bugzilla.gnome.org/show_bug.cgi?id=646209 -> wrong component, bugzilla has a l10n component10:23
ubot2Gnome bug 646209 in general "Wrong time format shown at login screen (gdm) for locale pt-PT" [Minor,Unconfirmed]10:23
seb128pitti, so .po patches should be opened on l10n, locale concerned10:23
pittiseb128: hm, gdm only has "general" and "docs"?10:24
seb128so it reaches the translators10:24
seb128pitti,"l10n", not "gdm" ;-)10:24
seb128sorry "product" l10n10:24
seb128component "pt"10:24
pittiah, ok10:24
seb128pitti, dup -> please use launchpad, your lp-set-dup sent a mail by bug it reassigned where launchpad do it silently10:25
pittiseb128: LP timed out :/10:26
seb128pitti, use <bugnumber>/+duplicate10:26
seb128that should not timeout10:26
seb128it's a page with only the dup entry10:26
pittiah, thanks; will remember next time10:26
seb128thanks ;-)10:26
pittiseb128: https://bugzilla.gnome.org/show_bug.cgi?id=646209 fixed10:27
ubot2Gnome bug 646209 in Portuguese from Portugal [pt] "Wrong time format shown at login screen (gdm) for locale pt-PT" [Minor,Unconfirmed]10:27
seb128pitti, danke10:27
seb128pitti, it's basically the equivalent of opening the bug on the langpack for us10:28
pittiah, after an unity --reset the "semi-maximize" actually works -- nice!10:28
seb128pitti, translators deal with translations on GNOME, not maintainers10:28
mvokamstrup: thanks for the pointer, I have not seen this one yet. I need to look into it10:35
mvokamstrup: though my chinese is that that great ;)10:35
kamstrupmvo: bugger, i was hoping you were fluent :-)10:39
mvokamstrup: I look into it now, probably the xapian index that is not properly build with the localized names10:44
seb128pitti, can you add bug #572260 to your review list?10:51
ubot2Launchpad bug 572260 in gnome-menus "package python-gmenu 2.30.0-0ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 2" [Undecided,New] https://launchpad.net/bugs/57226010:51
pittiseb128: sure10:51
seb128pitti, it seems lucid specific, just closed a bunch of duplicates though10:51
seb128well bug #60853510:52
ubot2Launchpad bug 608535 in gnome-menus "package python-gmenu 2.30.0-0ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/60853510:52
seb128could be the same issue10:52
seb128there is a suggested patch for it10:52
pittiwill look at them both10:52
Sweetsharkhmmm, when trying to do a build in my lucid pbuilder I get a lots and lots of errors after "dependency problems prevent configuration of pbuilder-satisfydepends-dummy", what might I be doing wrong?10:56
mvokamstrup: I added info to the bug, I suspect its a non-systemwide chinese setting, that is currently known to not fully work (i.e. the xapian index is build for the defualt language only currently)11:09
kamstrupmvo: yeah, i was thinking something like that as well11:15
kamstrupmvo: do you do anything special to handle chinese - because it appears the matching is wrong for installed apps as well (where I create an in-mem index of the gnome-menu contents)11:16
mvokamstrup: no special handling here, what is wrong exactly?11:18
mvokamstrup: installed stuff looks fine for me, but I can't do searches as I don't know what to type into ibus :/11:18
kamstrupmvo: from what I can tell the lower unity window here should match Empathy, not the stuff that it actually does https://launchpadlibrarian.net/67620390/Screenshot.png11:18
mvohm, ok11:22
mvoI think I will add code that builds multiple indexes11:22
mvofor each installed language11:22
Amaranthkamstrup: I thought you didn't do substring matching11:23
kamstrupAmaranth: prefix matching on each indexed term11:23
* mvo goes to get some lunch11:24
mvomaybe chinese today ;)11:24
kamstrupSo "aw" "bl" "po" all match "Awesome Blog Post!"11:24
Amaranthoh, I see11:24
AmaranthI thought you just did prefix matching of the entire Name11:25
AmaranthBut you split on spaces11:25
kamstrupoh, yeah, that would be a bit limited11:25
kamstrupwe split on non-alphanumeric characters right now I think11:25
kamstrupSo "disp" matches "gnome-display-properties"11:26
didrockspitti: oh nice, lp-set-dup dups as well duplicates of duplicates to the master bug as the UI does for a short time, excellent :)11:29
Amaranthkamstrup: That explains why I can search for "display" and get "Monitors" as a result11:29
pittididrocks: yes, that's the main point of it; but see seb128's concern about email spam above11:29
Amaranthkamstrup: since it matches display in gnome-display-properties11:29
pittididrocks: so it seems using /+duplicate is better11:29
AmaranthWhen I first saw that I actually thought you were doing tagging11:30
didrocksok, too late for this one though, sorry seb128 :)11:30
seb128didrocks, no worry, I hate you now but it's ok ;-)11:34
didrocksseb128: it doesn't change then? :-)11:34
kamstrupAmaranth: hehe, no. We're just lucky with the metadata :-)11:34
AmaranthRAOF, bryceh: I thought everyone knew nvidia's libGL caused lots of extra memory usage12:08
AmaranthKDE even has some complex hack to avoid this because they have so many GL apps12:08
AmaranthI doubt you'll get them to change, it has been like this so long I believe even KDE3 had a similar hack12:09
Amaranth(looking at bug 725434)12:09
ubot2Launchpad bug 725434 in cairo "Nvidia drivers lead to extra memory usage for each process using libGL" [High,In progress] https://launchpad.net/bugs/72543412:09
RAOFI certainly didn't know that until I looked.12:10
Sweetsharkwhats the story about gcj-native-helper? I had that as a dep in my last lucid release, but now I get "pbuilder-satisfydepends-dummy: Depends: gcj-native-helper which is a virtual package." in the pbuilder. Any hints?12:15
ricotzrodrigo_, hi :)12:33
rodrigo_hi ricotz12:34
seb128rodrigo_, can you add bug #740729 to your list?12:34
ubot2Launchpad bug 740729 in gnome-control-center "The indicators for gnome-control-center should be extended with icon/accessible descriptions." [Low,Triaged] https://launchpad.net/bugs/74072912:34
rodrigo_seb128, yes12:34
seb128I guess it's an easy one if you know what to do ;-)12:35
ricotzrodrigo_, great to have gnome-session ;), but there is a problem "default.list" shouldnt be installed12:35
rodrigo_ricotz, yes, that's not the only problem I've seen12:35
rodrigo_so still working on it12:35
ricotzrodrigo_, good, just wanted to point this out, good luck12:36
rodrigo_ricotz, thanks :-)12:36
rodrigo_ricotz, hmm, I merged it from debian, so it's wrong there also?12:36
ricotzrodrigo_, i am not sure how the handle desktop-file-utils12:37
seb128rodrigo_, Debian used to ship defaults.list in gnome-vfs but we ship it in desktop-file-utils, it's easier to get updated there12:37
rodrigo_ah, ok12:38
seb128rodrigo_, if they moved it to gnome-session we should just update it to not install it12:38
seb128we want to keep it in desktop-file-utils, it's a smaller source and either to update12:38
rodrigo_ok, I see12:39
=== MacSlow is now known as MacSlow|lunch
rodrigo_desrt, oh, you got 2 approvals from r-t for my patch?13:04
pittiSweetshark: hallo! would you mind to join #ubuntu-devel? all devs should be there, and it's easier for other teams to find people13:12
desrtrodrigo_: yup.  it's ready to go if you want13:13
desrtrodrigo_: but i think we should wait for ray to give his opinion13:13
rodrigo_desrt, well, let's wait a couple of days, for Ray's opinion13:13
rodrigo_ok, lunch then :)13:14
rickspencer3pitti, seb128 any idea who I talk to about the new dialogs and such for setting the tz in the clock indicator?13:17
Sweetsharkpitti: done13:17
seb128rickspencer3, mpt and mterry13:17
mterryseb128, hello13:17
seb128rickspencer3, do you have a specific issue? mterry fixed a stack of issues since the beta freeze13:17
seb128hey mterry13:17
pittiSweetshark: danke13:18
rickspencer3seb128, well, someone here at millbank showed me some wonky things in the tz picker part that was worrying them13:25
rickspencer3I'll follow up with mterry13:25
seb128rickspencer3, things like "can enter unknow location without getting feedback" are known and fix commited13:28
seb128rickspencer3, ok, feel free to raise the issue there if you want13:28
seb128there is like 15 bugs fix commited that will land after beta in natty13:29
=== bfiller is now known as bfiller_afk
=== chaotic_ is now known as chaotic
* bcurtiswx waves to room14:18
mterrybcurtiswx, hi!  :)14:19
=== MacSlow|lunch is now known as MacSlow
bcurtiswxhey mterry :)14:21
* vish wonders if PETA knows about the commands chrisccoulson recommends! Bug #745350 ;p14:29
ubot2Launchpad bug 745350 in firefox "while working on Firefox/Gmail, my cat sit down on keyboard " [Undecided,Invalid] https://launchpad.net/bugs/74535014:29
chrisccoulsonshhhhhh, don't tell them14:30
seb128rodrigo_, do you know if bug #707007 is still an issue in GNOME3?14:30
ubot2Launchpad bug 707007 in gnome-control-center "Monitor Preferences dialog buttons lacks conformity with rest of OS" [Low,Confirmed] https://launchpad.net/bugs/70700714:31
james_wchrisccoulson, nice tag :-)14:31
chrisccoulsonheh :)14:31
kklimondachrisccoulson: this bug is missing the lolcats tag ;)14:36
chrisccoulsonkklimonda, feel free to add it ;)14:36
kklimondaI would, I just can't think of a good comment :)14:36
chrisccoulsonheh :)14:37
chrisccoulsoni think we already used up all the available cat-related jokes ;)14:37
kklimondawell, I'm working on "cat got your tongue", but so far nothing has come out of it ;)14:38
dobeymvo: ping14:44
mvodobey: pong14:45
dobeymvo: hey, so i'm trying to use aptdaemon from C with dbus, but I am having a very hard time determining when everything is actually done.14:46
mvodobey: is your code available somewhere? what is the problem exactly? that the signals are not good? or timing issues with them?14:48
rodrigo_seb128, looking...14:49
dobeymvo: it seems that when i get the ExitStatus signal in PropertyChanged, the install/configure of the package has still not yet completed, and it takes several seconds afterward to complete.14:50
rodrigo_seb128, we don't have 'close' button anymore, just 'Detect displays' on the left and 'Apply' on the right, so I guess it doesn't apply anymore14:51
seb128rodrigo_, ok thanks, so not worth reporting upstream ;-)14:51
seb128that will just get autosolved next cycle14:52
seb128tedg, bug #745115, can you make anything from the stacktrace?14:52
ubot2Launchpad bug 745115 in indicator-application "indicator-applet-complete crashed with SIGSEGV in g_cclosure_marshal_VOID__POINTER()" [High,Confirmed] https://launchpad.net/bugs/74511514:52
seb128tedg, https://launchpadlibrarian.net/67607315/Stacktrace.txt14:52
seb128tedg, it's missing the first functions for some reasons but it might have enough infos to be useful?14:53
seb128or do you need details?14:53
* tedg clicks14:54
tedgseb128, Not sure, let me look at the code a minute.14:55
mvodobey: hmmmm, is your code available somewhere?15:00
tedgseb128, I don't think so.  It seems most likely to be an accessible string change (the only code that has really changed there), but I don't see anything obviously wrong with what it's doing :-/15:01
tedgseb128, There is a bug about that update not working though.15:01
seb128tedg, ok, let's see if we can get a better stacktrace then15:01
seb128tedg, thanks for checking15:01
dobeymvo: it's just on my hard disk at the moment, but i can commit/push i guess15:02
mvofwiw, unity consitently crashes for me (bug #742233) on my nviddia box15:02
ubot2Launchpad bug 742233 in unity "compiz crashed with SIGSEGV in nux::GraphicsDisplay::GrabKeyboard()" [Medium,Fix committed] https://launchpad.net/bugs/74223315:02
mvodobey: or just mail it, I haven't seen a problem like this yet15:02
seb128rodrigo_, can you check on bug #737539 as well?15:05
ubot2Launchpad bug 737539 in gnome-settings-daemon "[launcher] The "Low disk space" warning appears with title "gnome-settings-daemon"" [Low,Confirmed] https://launchpad.net/bugs/73753915:05
rodrigo_seb128, yes15:05
seb128thanks ;-)15:06
dobeymvo: ok15:06
seb128rodrigo_, oh and bug #740726 to match the g-c-c one you had earlier15:06
ubot2Launchpad bug 740726 in gnome-settings-daemon "Gnome-settings-daemon should set icon/accessible descriptions for app indicators." [Undecided,Triaged] https://launchpad.net/bugs/74072615:06
seb128rodrigo_, and I think I'm done with bouncing you bugs for today ;-)15:06
rodrigo_seb128, heh15:07
rodrigo_seb128, just keep bouncing, no problem15:07
rodrigo_seb128, ok, https://bugs.launchpad.net/unity-2d/+bug/737539 is fixed in 3.0, I guess we can just provide a patch for the 2.32 package15:07
ubot2Launchpad bug 737539 in gnome-settings-daemon "[launcher] The "Low disk space" warning appears with title "gnome-settings-daemon"" [Low,Confirmed]15:07
seb128rodrigo_, right, just do a merge request if you can, I will merge it15:08
rodrigo_hmm, well, now that I see the screenshot, I think it's a bug in unity15:08
rodrigo_the dialog has the correct title15:08
rodrigo_so not sure what it needs for unity launcher to show the correct one?15:08
seb128rodrigo_, ok, I guess it's similar to bug #740844 then, I've asked robert_ancell to have a look to that yesterday so let's wait for him to come back15:10
ubot2Launchpad bug 740844 in unity ""Authenticate" window shows in launcher as "Polkit-gnome-authentication-agent-1"" [Low,Confirmed] https://launchpad.net/bugs/74084415:10
rodrigo_seb128, right, looks it's the same issue15:10
rodrigo_it looks to me it's on apps that don't set the app name by not calling gtk_init15:11
rodrigo_that is, daemons15:11
seb128rodrigo_, right, see what macslow wrote on the polkit bug15:12
seb128"something like g_application_set_application_id(app, _("Authenticate")) should be used."15:12
seb128still it seems there is no reason why those need to be patched, the launcher should be able to get the title15:12
rodrigo_because there are lots of apps/daemons that would do that15:13
rodrigo_g-p-m I guess also has the same problem15:13
chrisccoulsoncyphermox, the firefox-stable PPA is fully translated now btw15:18
cyphermoxchrisccoulson, cool!, thanks15:20
ricotzcyphermox, hi :), do you think it is still possible to split libnm-glib2 in two packages (libnm-glib2 and libnm-glib-vpn1)15:24
ricotzcyphermox, while shipping these two libs in one package will cause problem with the transition to nm0.915:25
cyphermoxright, but we can still split them later once we prepare the packages for 0.915:26
cyphermoxricotz, I'll look into it15:27
ricotzcyphermox, there are already packages ;) and let people test them causes upgrade problems15:27
cyphermoxricotz, what do you mean?15:27
ricotzcyphermox, simply splitting them and make glib2 hard depend on glib-vpn1 should solve it without breaking anything15:27
cyphermoxricotz, yes, I get that, but upgrade problems?15:28
ricotzcyphermox, libnm-glib-vpn1 is still there and will be overwritten be the newer packaging15:29
ricotzone sec15:29
ricotzcyphermox, https://edge.launchpad.net/~ricotz/+archive/staging/+sourcepub/1606258/+listing-archive-extra15:29
dobeymvo: sent15:32
mvodobey: thanks, I have a call now, I look when that is finished15:33
dobeymvo: ok, thanks15:33
ricotzcyphermox, so i mean there are file conflicts while the current natty package ships two libs in one package15:33
cyphermoxricotz, you might be missing some further changes to make that work properly, no?15:34
cyphermoxricotz, I'm not against it, I just want to really know what's going wrong in your case because if I do the split, I'll likely fall in the same holes ;)15:36
cyphermoxricotz, from what I see in your package, looks to me like it would be the soname change that's causing you pain, not the fact that the package is split15:37
ricotzcyphermox, i dont think so, the nmglib library got bumped to 4 and also depends on the vpn1 lib, but many packages are linked againt the nmglib2 which cant be uninstalled15:37
cyphermoxthat's another problem... you need to rebuild those too15:37
ricotzyes, and this bumps are causing the problem while shipping two public libs in one package15:37
ricotzcyphermox, i know that the dependencies would need a rebuild, but i might be possible two let the different nmglib version co-exist15:38
cyphermoxoh, I see what you mean15:39
ricotzso this split would be actually a bugfix in my opinion ;)15:39
cyphermoxricotz, like I said, I'm not against it I just don't want to break things one month from release ;)15:40
ricotzand let the natty packaging hard depend on the vpn package would harm the current linking15:40
ricotzi know what you mean15:41
cyphermoxricotz, are you planning on rebuilding the VPN plugins too?15:42
ricotzcyphermox, not really, these packages are for testing gnome-shell15:43
ricotzand i was hoping since the vpn library version is the same it would work15:44
ricotzrelying on upstream abi stability15:44
cyphermoxso I'll get started on this, and let you know (shouldn't take long, and it's something I have to do anyway)15:45
ricotzso splitting the package is quite a proper fix though ;)15:45
ricotzcyphermox, alright, thank you15:45
=== skaet_afk is now known as skaet_
ricotzcyphermox, make the vpn dependency unversioned or (>= 0.8.4)15:46
cyphermoxricotz, it's part of what I had to do to reduce the delta with debian15:46
cyphermoxricotz, can you file a bug about it please?15:55
ricotzcyphermox, can do, so it will be the 525th bug ;)15:57
ricotzthese are many bugs :(15:57
pittikenvandine, seb128, didrocks, rodrigo_: so the current rebuild test showed some FTBFSes (libgtop2, gnome-games, glib-networking, bamf): can we split them amongst us?16:02
pitti(mostly multiarch related)16:02
didrockspitti: I can take bamf16:02
rodrigo_pitti, fine by me16:02
* kenvandine will take libgtop216:02
seb128pitti, there is a libgtop in unapproved and I wanted to sync on debian so I can do this one16:02
seb128kenvandine, ^16:03
pittirodrigo_: https://launchpadlibrarian.net/67639771/buildlog_ubuntu-natty-i386.glib-networking_2.28.0-1_FAILEDTOBUILD.txt.gz looks easy, do you want to try this one?16:03
* kenvandine takes gnome-games then16:03
pittiseb128: ah, thanks16:03
rodrigo_pitti, sure16:03
pittierm, and what do I fix then?16:03
pittiwell, I guess I'll wait for the next one then :)16:03
seb128pitti, there is a new version for the glib one in debian to sync, not sure if that would fix it16:03
pittirebuild is ongoing, after all16:03
ricotzcyphermox, https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/74576916:04
ubot2Launchpad bug 745769 in network-manager "Update packaging according to libraries" [Undecided,New]16:04
pittiseb128: unlikely (see log), but I guess it's worth fixing the FTBFS on top of the latest unstable one indeed (<- rodrigo_)16:04
pittikenvandine: I thought robert might want to take gnome-games16:04
kenvandinehe would be best16:04
kenvandinebut he isn't here now :)16:05
kenvandinehumm, our gnome-games packaging branch is out of date16:05
pittirodrigo_: if you have difficulties, please let me know and I can help you or take over16:06
rodrigo_pitti, seems easy from a 1st look, so trying, will ping you if it's not that easy :)16:06
pittikenvandine: perhaps robert forgot to push?16:06
pittirodrigo_: cool, thanks16:06
kenvandinemterry it appears :)16:06
* kenvandine gets it in sync16:06
mterrykenvandine, ?16:07
kenvandinethe latest bzr branch for gnome-games doesn't match what is in natty16:07
pittiseb128: https://launchpadlibrarian.net/67675825/buildlog_ubuntu-natty-i386.libgtop2_2.28.1-1ubuntu1_FAILEDTOBUILD.txt.gz16:07
pittikenvandine: https://launchpadlibrarian.net/67645831/buildlog_ubuntu-natty-amd64.gnome-games_1%3A2.32.1-0ubuntu3_FAILEDTOBUILD.txt.gz16:07
kenvandinemterry, it was ages ago now16:08
pittididrocks: https://launchpadlibrarian.net/67608122/buildlog_ubuntu-natty-i386.bamf_0.2.80-0ubuntu1_FAILEDTOBUILD.txt.gz16:08
kenvandinepitti, thx16:08
pittithanks guys, you rock16:08
seb128pitti, thanks16:08
didrockspitti: yeah, I've it opened, looking why the gio module isn't built16:08
rodrigo_pitti, hmm, my 1st thought was to remove the rm command in debian/rules, but it builds fine on x86_64, so is it because on i386 it doesn't generate the .a files?16:08
pittithe test rebuild is about half through, so we should expect some more; I'll follow it and take some more16:08
pittididrocks: gio recently changed the module dir for multiarch, didn't it?16:08
pittirodrigo_: they might have changed the library path due to the new multiarch directories16:09
didrockspitti: right!16:09
seb128dpm, hi, is there any way to see what the last unity template has been imported?16:09
pittirodrigo_: so there might actually be some more to it, like also moving the libraries to the right destination, etc.16:09
didrockspitti: should we upload now for post-beta or just pend the change in a vcs?16:10
pittididrocks: as you prefer16:10
pittididrocks: if it's likely that you'll get more changes post beta (like for bamf), vcs is probably enough16:10
didrockspitti: ok, thanks! ;)16:11
pittibut as we have a frozen queue, uploading is fine16:11
didrockslet's avoid an upload for now, I'll upload if we have no more change after beta16:11
nessitaseb128: ping16:12
seb128nessita, pong16:12
nessitaseb128: can I ask you about your last comment on bug #722485? in theory, you (as user) should not be able to reach the devices tab if you have no UBuntu One credentials (having credentials == having your device configured)16:13
ubot2Launchpad bug 722485 in ubuntuone-control-panel ""Value could not be retrieved" keeps appearing ont the control panel" [Low,Triaged] https://launchpad.net/bugs/72248516:13
seb128nessita, what device?16:13
nessitaseb128: your current computer16:13
seb128nessita, I've an account configured on my laptop, I'm syncing files16:13
nessitaseb128: so, what do you mean with "the device tab display this error in red which is not a real user friendly way to say I've no device configured"?16:14
seb128nessita, well, I though device would be i.e a phone16:14
rodrigo_so, anyone on i386 can run this command: pkg-config --variable giomoduledir gio-2.0 ?16:14
seb128nessita, I've an account and I'm connected, I'm syncing files on u116:14
nessitaseb128: are you connected to the internet?16:14
seb128nessita, but I don't have any other computer or device identificated16:14
seb128nessita, yes, "File Sync is up-to-date" it says16:15
didrocksrodrigo_: /usr/lib/i386-linux-gnu/gio/modules16:15
seb128nessita, and files uploads if I copy those to the UbuntuOne folder16:15
nessitaseb128: then, you're having any other kind of error. Could you please attach your log file to the bug report? (~/.cache/ubuntuone/log/controlpanel.log)16:15
rodrigo_didrocks, ok, on x86_64 it's /usr/lib/gio/modules, so that's the problem then16:15
didrocksrodrigo_: are you up to date? it seems it's /usr/lib/x86_64-linux-gnu/gio/module16:16
didrockslike for https://launchpadlibrarian.net/67612895/buildlog_ubuntu-natty-amd64.bamf_0.2.80-0ubuntu1_FAILEDTOBUILD.txt.gz16:16
dpmseb128, otp, will be back to you later16:16
rodrigo_didrocks, hmm, not really, haven't upgraded in a few days16:17
didrocksrodrigo_: I think you don't have the multiarch love then :)16:17
pittididrocks, rodrigo_: /usr/lib/gio/modules/ is still the fallback path for modules which haven't been converted to multiarch yet16:17
Amaranthah, lovely multiarch paths16:18
AmaranthI'll probably never remember them :)16:18
=== MacSlow is now known as MacSlow|break
didrockspitti: oh ok :)16:19
pittiAmaranth, didrocks, rodrigo_: FYI, the canonical library dir is /usr/lib/`dpkg-architecture -qDEB_HOST_MULTIARCH` now16:19
pittiin debian/rules /usr/lib/$(DEB_HOST_MULTIARCH) should also work16:20
pittikenvandine: ^16:20
didrocksthanks :)16:20
rodrigo_pitti, well, in this case I guess it's better to use the gio-2.0 pkgconfig variable, as this is the same used in the upstream source, right?16:21
pittirodrigo_: right16:21
rodrigo_pitti, testing now a build with that on x86_64, and will do it on i386 as soon as it's ok16:21
dpmseb128, on the admin page I see "2011-03-25 18:18:38.414867+00:00" as the last update - does that sound ok? I can't see anything on the imports queue , but imported templates are deleted from the queue history after a couple of days (https://translations.launchpad.net/ubuntu/natty/+source/unity/+imports?field.filter_status=IMPORTED&field.filter_extension=pot)16:21
dpmre: unity template16:21
pittirodrigo_: does the .pc have the correct multiarch already? (it ought to)16:22
seb128didrocks, ^16:22
seb128didrocks, dpm: ok, I guess it imports the upstream outdated one since the rules update is failing16:22
seb128dpm, thanks16:22
seb128dpm, https://bugs.launchpad.net/ubuntu/+source/unity/+bug/74577416:22
ubot2Launchpad bug 745774 in unity "the translation template update is broken" [High,Confirmed]16:22
rodrigo_pitti, not on my system, which is not up-to-date, but updating now16:22
didrockswill fix that just after bamf16:23
didrocksone pbuilder at a time ;)16:23
dpmseb128, ah, ok, thanks for the heads up. Yeah, it probably imports the non-updated pot in the tarball16:23
didrocksdpm: there is a variable issue, too bad we don't know which latest update contained a change though16:23
didrocksok, seems slangasek beat me for bamf while I was fixing it :)16:40
seb128so maybe I should wait to not duplicate work ;-)16:40
seb128pitti, ^ did you sync with steve to make sure we don't all dup work?16:41
pittiseb128: doko was discussing the general issues with slangasek, but I didn't ask Steve about particular packages yet16:41
pittiI did tell doko, though16:41
pittior, rather, he asked the desktop team to look into those16:41
seb128pitti, the upload robert_ancell did yesterday for libgtop seems to fix the build16:53
seb128so you can cross that one from the list16:53
pittiseb128: cool, thanks16:56
=== MacSlow|break is now known as MacSlow
rodrigo_pitti, lp:~rodrigo-moya/ubuntu/natty/glib-networking/fix-ftbf <- builds fine on x86_64 (with upgraded glib), so testing now on i386 and if it works there, I'll submit16:59
=== seiflotfy__ is now known as seiflotfy
dobeymvo: any luck?17:01
mvodobey: not yet, but the meeting only ended ~15min or so ago)17:03
dobeymvo: oh wow, ok. long call :)17:06
mvodobey: foundations call and then irc meeting :)17:07
dobeymvo: that's ok. i just finished lunch :)17:07
mvodobey: I don't see anything obvious, I would like to debug a little bit, but I need more of the code for it. I need to go for dinner soon, could you just mail me more (or point me to a repor?). I can check it out tomorrow morning then17:08
dobeymvo: ok17:16
rodrigo_pitti, ok, builds fine on both 64 and 32 bits, so https://code.edge.launchpad.net/~rodrigo-moya/ubuntu/natty/glib-networking/fix-ftbf/+merge/55588 is up for review/merge/upload17:21
* didrocks hugs pitti for dh_translations :)17:38
didrockspitti: however, what do you think about an env var to force intltool to be used? It doesn't here because we are using cmake and there is no intltool official support last time we checked with neil (we are abusing bzr export for this)17:39
didrockspitti: or maybe, I can base on  GETTEXT_PACKAGE, and assume we are using intltool with that17:40
pittire (sorry, was busy in a phone call)17:43
pittirodrigo_: yay you! I'll sponsor in a bit, thanks17:43
pittirodrigo_: ah, that looks great; would you mind sending the patch to Debian as well?17:44
pittirodrigo_: they will get multiarch soon as well, then it'll hit them as well17:44
pittiand it can already be applied as it is right now17:44
pittididrocks: GETTEXT_PACKAGE in po/Makefile isn't intltool specific17:48
pittididrocks: I'm fine with a variable to force intltool, but I wonder if we can autodetect it in this case?17:49
pittididrocks: i. e. should we always call intltool if we are using cmake/17:49
didrockspitti: I'm going to assume that for now if that's fine with you, I didn't see any support i cmake17:49
pittididrocks: as this is currently undefined, feel free to define it :)17:50
didrockspitti: heh, just doing it! ;) building unity to check it's updated17:50
kenvandinepitti, should girepository-1.0 go in the multiarch libdir as well?17:56
kenvandinei don't have anything there yet, but that seems weird17:56
pittikenvandine: hm, good point -- as they seem to be platform specific, I guess they should17:57
kenvandinei just wonder if that breaks anything17:57
kenvandinewe don't seem to have anything there yet17:57
=== bjf is now known as bjf[afk]
dobeypitti, didrocks: you can detect intltool from po/Makefile18:14
pittikenvandine: more importantly, does our current g-i actually look in the multiarch dir?18:14
pittidobey: does that exist with cmake?18:14
didrocksdobey still, cmake doesn't have po/Makefile18:14
dobeypitti: intltool doesn't support cmake18:14
didrocksdobey: that's what I told, see above18:15
kenvandinedobey, i am going to test that theory :)18:15
kenvandinepitti, ^^18:15
kenvandinedobey, sorry :)18:15
* kenvandine eats lunch first though :)18:15
dobeydidrocks: ok, well i'm confirming as intltool maintainer that we don't support it :)18:15
didrocksdobey: any plans btw? We are making ugly things in unity to workaround that ;)18:16
didrockspitti: can you have a look at rev 168 of pkgbinarymangler to see if I didn't screw everything18:16
pittididrocks: any chance you can write a little test for this?18:17
dobeyno current plans, no. i don't kno whow cmake works exactly, and generally avoid it if possible :)18:17
didrockspitti: oh sure, but little little then! :-) </kidding>18:17
didrocksdobey: maybe after natty we can have look at that together?18:18
pittididrocks: a test_dh_translations_cmake, similar to test_dh_translations_python()18:18
pittididrocks: hm, I wonder if we actually need the po/Makefile checks in build_pot18:19
dobeydidrocks: maybe. we can chat in budapest if you want18:19
didrocksdobey: that would be nice!18:19
pittididrocks: perhaps it would be sufficient to always call intltool-update with -g?18:19
pittididrocks: and just check if we have $domain18:19
pittididrocks: but the code looks good to me18:20
didrockspitti: I wanted to avoid widespreading the impact, hence this check18:20
pittididrocks: only nitpick: please fix the version number to be 94 :)18:20
pittididrocks: we've got test cases for everything, so I think cleaner code wins18:20
didrockspitti: oh oh ;)18:20
didrockspitti: adding the check and fixing that18:20
didrockspitti: ok, in that case, removing that18:20
pittiand also, I'd like to keep having tests for everything, as inevitably we'll get the n+first build system which we also need to support, and then we mustn't break the older stuff18:21
=== bjf[afk] is now known as bjf
* didrocks flushes && -e18:21
pittididrocks: and no didrock@localhost in changelog :)18:21
didrockspitti: agreed, doing that now18:21
pittididrocks: cheers18:21
didrockspitti: that will teach me working on my sandbox :p18:21
Sweetsharkpitti: Now here is a strange one: The build almost completes but on lucid a call to "dh_link -i" complains about a symlink, which it did not on maverick. Any idea why? The "-i" and calling it without any further parameter is not documented, so I am a bit confused by it.18:35
seb128tedg, hey18:36
tedgseb128, Howdy18:38
seb128tedg, will the file, close default menu go away in natty?18:38
seb128tedg, I'm asking to know if I still need to backport the gtk patch to get the menu translated or if I should not bother18:39
tedgseb128, It will for Unity, but stay for applet.18:39
seb128ok, so still worth backporting it18:39
cyphermoxricotz, I have packages ready for NM in my PPA with the changes we discussed18:51
rodrigo_pitti, what's the proper way to send a patch to debian?18:54
pittirodrigo_: http://www.debian.org/Bugs/Reporting is the canonical method18:55
pittirodrigo_: there's a script submittodebian which automates this18:55
pittiI haven't used it much18:55
rodrigo_ah, cool!18:55
kenvandinepitti, so the answer is no... GI doesn't look in the multiarch libdir19:00
didrockspitti: ok, rev 170 sould fix + add a test (which pass, unbelieavable \o/ ;))19:02
didrocksdinner time, bbl19:02
ricotzcyphermox, looks good and installs fine :) could you add a debdiff to the bug report19:05
pittididrocks: cool, thanks! bzr diff -r 167.. looks fine to me19:06
pittididrocks: just fixed the indentation in r17119:07
cyphermoxricotz, I'll upload that as soon as I can ;)19:08
ricotzcyphermox, great and hopefully it will be accepted19:09
didrockspitti: thanks! Should I upload or should we wait for tomorrow? (not sure if you have other pending changes planned). Next unity will build-dep on it19:20
cyphermoxricotz, just confirmed that the vpns work too19:22
desrtdoes anyone know darren's email address?19:33
ricotzcyphermox, great19:34
desrtthe man is ungoogleable19:35
pittididrocks: feel free to upload, I have nothing19:42
didrockspitti: ok, doing then. thanks :)19:43
=== bjf is now known as bjf[afk]
=== skaet_ is now known as skaet_otp
sirgadHey. Anyone here now how to configure gconf settings or similar on a LiveCD to offer default settings on boot?20:33
pittirodrigo_: sponsored20:35
Sarvattanyone have any idea why compiz doesn't even try to start up in a unity session with fglrx? http://paste.ubuntu.com/587508/ it just goes to a classic session instead, yeah I'm using the unity daily PPA20:49
Sarvattif I unity --replace it comes up fine, except I can't kill the gnome-panel20:49
Sarvattremove fglrx, the unity session works fine again20:50
didrocksSarvatt: run /usr/lib/nux/unity_support_test -p and check as well the exit code20:51
didrocksSarvatt: and the exit code?20:53
didrocksSarvatt: should be 0, something else means it doesn't work…20:54
didrockshence the fact that gnome-session fallback20:55
didrocksSarvatt: the test tool is changing starting tomorrow and rewritten from scratch, so hopefully, this will be fixed20:55
Sarvattsarvatt@sachiel:/etc/X11$ DISPLAY=:0 /usr/lib/nux/unity_support_test -p || echo $?20:55
didrocksyeah, that's why it fallback20:56
=== bjf[afk] is now known as bjf
Sarvatthrm, wonder if it's https://launchpadlibrarian.net/67681186/compiz_1%3A0.9.4git20110322-0ubuntu6%2Br2720%2B201103291330_1%3A0.9.4git20110322-0ubuntu6%2Br2723%2B201103301010.diff.gz21:06
Sarvattthe compiz check is whats failing21:07
didrocksSarvatt: no, because you can run compiz… and it was a match in the name21:07
didrocksSarvatt: so, the test tool has nothing "compiz" in its name21:08
didrocksSarvatt: anyway, as told, it's rewrittent and the new version will land tomorrow once beta released21:08
didrocksSarvatt: so, compiz is working for you from the ppa + fglrx driver?21:08
Sarvattyeah, unity too, just the session isn't starting it right21:09
didrocksSarvatt: yeah, so hopefully the new unity test tool will fix it21:12
=== skaet_otp is now known as skaet
cyphermoxanyone got a moment to look over https://code.launchpad.net/~mathieu-tl/ubuntu/natty/ubuntu-mono/icon-cache/+merge/55629 and sponsor it ? :)21:32
* didrocks waves goodbye, just too tired right now :)21:33
=== skaet is now known as skaet_afk
pittigood night everyone22:33
kenvandinerobert_ancell, ping23:02
robert_ancellkenvandine, hey23:02
kenvandineduring the multiarch rebuild gnome-games is one of the packages that fails to rebuild23:03
kenvandinei took a swing at fixing it... but didn't nail it yet23:03
kenvandinerobert_ancell, can you finish it off?23:03
kenvandinei pushed my changes to the ~ubuntu-desktop branch23:04
kenvandinedh_girepository: Could not find library libgames-support-gi.so.023:04
kenvandinemake: *** [binary-predeb-IMPL/aisleriot] Error 223:04
kenvandineis the current error23:04
robert_ancellkenvandine, sure.  man gi, vala, dso linking, multiarch - they're really trying to make our life hard!23:04
robert_ancellkenvandine, oh, I fixed that in the GNOME3 branch, I can do that23:05
kenvandinewith libdir set to $${prefix}/lib/$(DEB_HOST_MULTIARCH)23:05
kenvandineoh... then you can revert any changes i made :)23:05
kenvandinerobert_ancell, thx!23:05
robert_ancellit's a private library, so dh_girepository goes mental.  I have a patch that moves the typelib to a private location so it doesn't even notice it23:05
kenvandinerobert_ancell, out of curiousity, why doesn't the typelib get put in a separate gir package?23:05
=== skaet_afk is now known as skaet

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