=== bjf is now known as bjf-afk [00:13] robert_ancell - so, this crash might be a gconf issue [00:14] when an application calls gconf_client_add_dir for the first time, there should eventually be a call to CORBA_ORB_init [00:14] chrisccoulson, right [00:14] but this never gets called again because it assumes it remains initialized [00:14] even after shutdown_orb is called [00:15] in gconf_get_config_listener() [00:15] thats the interesting bit [00:16] anyway, i'm going to have to stop looking at this for tonight, else i will never sleep [00:16] it will take me the whole night just to wind down ;) [00:16] chrisccoulson, sleep on it! :) [00:17] i should comment on the bug report really, else i will just forget it all by morning ;) === asac_ is now known as asac [02:23] hmm, that seahorse bug sounds like a fun bug compiz used to have with gconf [06:57] i need some help.... i installed nvidia driver NVIDIA-Linux-x86-190.42-pkg1.run manually for GeForce 7300 LE since Hardware Drivers show empty driver... the installation is succefull and i can play 3D game... but i can't have compiz started ??? [06:57] btw this is for Ubuntu 9.10 [08:30] Good morning [08:34] good morning pitti [08:36] hey didrocks [08:39] mvo, good morning. Should software-center trunk still work on 9.10? [08:42] mpt: yes [08:43] mpt: is it not for you? [08:43] mvo, no, I get "NameError: global name 'os' is not defined" whenever navigating to an application screen [08:43] File "/home/mpt/hacking/software-center/softwarecenter/distro/aptutils.py", line 36, in get_release_date_from_release_file [08:43] mpt: oh, give me a sec [08:43] mpt: I fix that [08:45] mpt: r420 [08:46] mvo, that's changed the error: it's now "NameError: global name 'datetime' is not defined" [08:47] mvo, File "/home/mpt/hacking/software-center/softwarecenter/distro/Ubuntu.py", line 122, in get_maintenance_status [08:48] mpt: heh :) let me switch to a karmic machine to test [08:48] good morning desktopers [08:49] mvo, on another topic, I don't see a 1.0 branch on the Branches page. Should there be one? Or is lp:ubuntu/karmic/software-center enough for that? [08:49] mpt: the karmic branch is the 1.0 branch, we can create a explicit name if you prefer that [08:50] good morning seb128 [08:50] hey chrisccoulson, how are you? [08:51] good thanks - it's nearly the weekend :) [08:51] how are you? [08:51] mvo, I don't mind, just as long as there's somewhere to apply critical 1.0.x fixes if there are any (and to turn on the angled path button by default, perhaps) [08:51] * seb128 wonder if chrisccoulson ever sleep, he's up after everybody and already there in the morning too [08:51] * seb128 wonder if chrisccoulson ever sleep, he's up after everybody and already there in the morning too [08:51] chrisccoulson, good thank you [08:51] heh, i don't sleep very much ;) [08:51] although i have been to bed before 2am several times recently [08:53] mpt: please try r421 [08:54] mvo, that works, thanks :-) [08:55] "Launchpad will be going offline for maintenance in five minutes. " [08:55] urg [08:55] mpt: is this https://wiki.ubuntu.com/SoftwareCenter?action=AttachFile&do=get&target=1.0-available-category.jpg still desired? or will that change again for 2.0 ? otherwise I give it a short at implementing it now [08:55] * mvo wants something more exciting than bug triage [08:56] mvo, launchpad is going down so you will be forced into it ;-) [08:57] mvo, there is plenty of merges and updates too [08:58] mvo, I'd give that a 60% probability, though I need to discuss the layout with ivanka [08:58] mvo, there are other things that are both more certain and probably more exciting :-) [08:58] mpt: what do you have in mind ? [08:58] mpt: for 60% I won't start [08:59] fair enough -- one moment [08:59] seb128: FYI, versions.py failed over night (for some mystical reason index/lucid/var was removed completely) [08:59] running now [09:00] hey pitti [09:00] ok thanks [09:00] grrrr at robert_ancell [09:00] hard to make him things he has not interest in [09:01] he started full steam on 2.29 but doesn't seem to care about merging on debian [09:01] let's wait for him to join tonight ;-) [09:02] mvo, there's this [09:02] you promisied "more exciting" [09:02] mvo, or this [09:02] oh, foo [09:03] The intersection of {more exciting} and {fully specified} is mainly Clutter animations [09:03] at the moment [09:03] and [09:04] mpt: is https://wiki.ubuntu.com/SoftwareCenter?action=AttachFile&do=get&target=1.0-installed-home.jpg (as a treeview) still the plan? [09:05] mpt: what will it do when packages are activated? still a treeview? [09:06] mvo, that's subject to the same 60% certainty -- if we change how we present departments, the "Installed Software" section will change to look much more similar to the "Get Software" section [09:06] mpt: ok === kklimond- is now known as kklimonda [09:16] was there an announcement for LP going offline? [09:17] (or perhaps it just got lost in my inbox) [09:17] I didn't read one [09:18] i can't find one either [09:18] it seems we never get announcements any more [09:41] morning [09:42] lut huats [09:43] hello seb128 ! [09:44] cassidy, hey [09:44] cassidy, should we sync telepathy-gabble 0.9.2 in lucid? [09:45] ie is that the right version for this cycle? [09:45] seb128, this version still have regressions comparing to the 0.8.x branch but hopefully that should be fixed for the Lucid release [09:46] hum [09:46] seems a "no" ;-) [09:46] thanks [09:46] and we'll probably use new feature from this version in Emapthy 2.30 so I'd say to go for it [09:46] well [09:46] tbh, the main regression is the proxy support which has always be broken so... [09:46] "should be fixed" is not good for a lts [09:46] oh, LTS, right [09:46] we need to aim for something which will be stable by lucid time [09:46] not something which might be in the middle of unstable serie [09:47] you can keep 0.8 for now and I'll let you know when we think 0.9 is good enough to replace it [09:47] thanks [09:47] 0.8.x is still well maintained as that's the version shipped with Maemo 5 [10:08] cassidy, is empathy supposed to unhide the list on the same workspace it was? [10:08] or on the current workspace? [10:09] when you click on the notification area icon [10:09] using compiz [10:09] ahah, good question. IIRC we uses gtk_window_present() whose behaviour is unclear [10:09] if that makes a different [10:09] and depends of your WM [10:09] ok [10:09] (I know that sucks) [10:10] dunno what pidgin do but that works [10:10] it open the list on the workspace where you are [10:10] the empathy behaviour is driving me crazy [10:10] I work on several workspace and the list never opens where I'm working [10:10] we have lot of troubles with that; seems lot of WM are broken [10:10] seb128, I really want some WM guru to take a look [10:10] seb128, do you have that in house? :) [10:10] seb128, because IMO it is *all* WM that are broken [10:11] and each version is broken in a different way [10:11] I just tried and the contact list appear on the active desktop [10:11] (metacity) [10:11] the best wm guru we have there is Amaranth [10:11] rb/pidgin does ugly hacks to position the window [10:11] dunno what hack they do but that works mostly [10:11] you could maybe copy some of those [10:11] seb128, some help on this issue would be really appreciated [10:11] alright [10:11] seb128, they reposition manually the window on the desktop and the position/size [10:11] noting that for sprint and UDS [10:11] which should be useless [10:12] but we don't have wm gurus either [10:12] we'd prefer to have the bugged code fixed properly (in Empathy, WM, ...) than adding more hack [10:12] that's not going to happen though [10:12] afaik it works fine with metacity [10:12] cassidy, that's what I keep saying for years... but really that need WM guru to look into. I don't know enough how it works [10:13] it's an user story fail for empathy meanwhile [10:13] I'm prepared to be told empathy is doing something wrong, but I would like to know the real reasons, not just pull hacks to make it work by hiding the real defect [10:13] cassidy, the issue is that as you said the gtk function behaviour is not defined [10:13] and the gtk issue is open for years [10:13] we already tried different hack and they always fix one case and broke others [10:13] I appreciate you don't want to have hacks but meanwhile your software look buggy compared to others [10:14] tbh I almost never had any problem using ubuntu with metacity and its compositor [10:14] Zdra, the main issue is that the gtk call behaviour is not defined [10:14] but other users report problems even in the exact same configuration... that's crazy [10:14] ie should present bring you to the dialog, bring the dialog to you or just raise it [10:15] it's let to the wm to decide right now [10:15] so with some wm you will get the dialog on your workspace [10:15] some others will change the active workspace [10:15] some others will make it claim for attention in the taskbar [10:16] seb128, the biggest problem reported by users is that the window position is reset to left/top corner after unhide the window [10:16] there is that issue [10:16] seb128, and that is not supposed to be WM-dependent, the position really should be remembered [10:17] and the fact that using compiz it doesn't show it on the right workspace [10:17] I will put it on the lucid list of issues so we look at it [10:17] seb128, for the move to desktop, give focus, etc, I can understand empathy should do it, because it is not guaranteed by gtk_window_present [10:17] but I think you will have to workaround in empathy [10:17] the gtk behaviour not being defined is the issue [10:17] and it will require spec changes and gtk changes [10:17] which I doubt will happen quickly [10:17] Zdra, did we try to use the same hacks as rb ? [10:18] cassidy, no [10:18] cassidy, I saw it and told "I don't want to be responsible of such a hack" [10:18] maybe that's something we should try [10:18] let's be pragmatic in some case hacks are required to make the user experience good [10:19] yeah :( [10:19] IIRC other app does gtk_window_hide() gtk_window_show() instead of gtk_window_present() to be sure it is forced to be moved to current ws [10:19] I will try to see what we can do for a proper fix there [10:19] but as said if that require xdg spec changes and gtk changes it will take a while [10:19] so meanwhile we will want to workaround [10:20] Zdra, seems to be a good way if that works [10:20] seb128, tbh the whole concept of window hiding in the status icon is not well supported in GNOME [10:20] seb128, note that even rb is not unbreakable, I already had window possition issues with it ;) [10:21] I think the best hack is in piding [10:21] pidgin [10:21] best in the sense of works the best, not the nicer [10:23] right [10:23] as said pidgin seems to work fine for users [10:23] ie we get no complain about it [11:44] Laney, hey [11:44] Laney, do you want to sru the new f-spot update? [11:45] to sru = to work on the sru changes [11:56] seb128: yeah, maybe tomorrow [11:56] Laney, ok thanks [12:28] chrisccoulson: would you have some time for another SRU? (bug 463353) [12:28] Launchpad bug 463353 in system-tools-backends "[users-admin] Modifying user password has no effect" [Undecided,Confirmed] https://launchpad.net/bugs/463353 [12:29] sorry for delegating, SRU management and UDS planning pretty much eat all my time these days [12:29] or seb128? [12:29] pitti - yeah, i can take a look at that [12:29] do you want to assign it to me? (i'm just about to go for lunch) [12:30] pitti, I will let chrisccoulson look at this one, I have only today left as work day before sprint and uds [12:30] chrisccoulson: thanks muchly; will do [12:38] asac: around? [12:39] mvo, he's on vac for the week [12:40] *pff* [12:40] seb128: thanks :) [12:40] mvo, you're welcome [12:40] mvo, what you mean by *pff* is *slacker* right? ;-) [12:40] seb128: I just wanted his feedback on the "XB-Restart-Required idea [12:40] seb128: yeah [12:40] :P [12:40] what is that? [12:41] oh [12:41] you want to be able to know in advance what will need a restart [12:41] * seb128 switches channel for the discussion === MacSlow is now known as MacSlow|lunch === tseliot1 is now known as tseliot === MacSlow|lunch is now known as MacSlow === hggdh_ is now known as hggdh === mac__v is now known as mac_v [15:04] urg [15:06] pitti, there is 2181 i386 crash bug in the retracer queue [15:06] * seb128 look to the amd64 one [15:06] bonjour seb128 what's the matter? [15:06] hey rickspencer3 [15:06] rickspencer3, just 2181 crash bugs in launchpad which have not been retraced yet [15:07] on i386 [15:07] *sigh* [15:07] seb128 - they're not all seahorse-agent crashes are they? [15:07] ;) [15:07] chrisccoulson, could be, we will know when they will be retraced... [15:07] chrisccoulson, still looking for bugs? [15:08] seb128 - i don't think we're going to get away without fixing the seahorse-agent crash - the crash reports are still flying in quite quickly even with apport disabled by default [15:08] seb128 - did you have any bugs in mind? [15:08] chrisccoulson, bug #442130 [15:08] Launchpad bug 442130 in gvfs "Custom fstab causes duplicate entries for partitions in Places" [Low,Triaged] https://launchpad.net/bugs/442130 [15:08] chrisccoulson, we should add a bug pattern for the seahorse crash [15:08] chrisccoulson, that would block bug filing pointing to the open bug [15:09] yeah, that might be a good idea [15:09] seb128 - i'll take a look at that gvfs one later as well then [15:10] chrisccoulson, thanks [15:10] 892 amd64 crash bugs to retrace [15:10] seb128 - are you not here next week? [15:10] chrisccoulson, I will be traveling on monday and then be on us time [15:10] and sprinting [15:11] ah, ok [15:11] I will probably be there but on different hours [15:11] it's going to be quiet in here next week [15:11] and not as much as I usually do [15:12] i'll probably have a baby by the time you arrive back from the US. when she decides that she wants to arrive! [15:13] oh right [15:13] you are going to have short nights for other reasons soon ;-) [15:13] you are going to have short nights for other reasons soon ;-) [15:13] ups [15:13] chrisccoulson, when we stop seeing you everyday, we will assume the baby came :) [15:14] lol [15:14] i think i should just forget about sleep entirely! [15:14] sleep is overrated [15:14] that's what i keep telling people ;) [15:14] but sometimes i feel like i could sleep at work [15:15] probably not the best place to fall asleep ;) [15:15] hehe [15:15] * kenvandine was replacing the brakes on my car at midnight last night... i would rather have been sleeping... but if i am hacking on stuff i would prefer that to sleep :) [15:18] kenvandine - i take it you have a source of light to be able to work on your car that late at night? [15:18] i can't do anything to mine at the moment, as it's dark when i get back from work [15:20] garage [15:21] i need a bright portable light though [15:21] the overhead lights aren't great for working under fenders :) [15:21] * kenvandine has some busted up knuckles today from not being able to see well === bjf-afk is now known as bjf [16:11] seb128: ugh [16:11] seb128: does it keep breaking on the same bug, or does it make progress? [16:11] pitti, it makes progress [16:11] pitti, retraced 50 bugs between the 2 recent restart [16:12] ok, and today's crashes were due to the LP rollout, I guess [16:12] I'm suprised we keep getting so many crash bugs [16:12] is apport still on in karmic? [16:13] pitti, they crash every hour [16:13] gateway errors [16:13] or permission errors [16:20] seb128: what's the bug # for the unclickable button in --view mode? [16:23] Laney, bug #448162 [16:23] Launchpad bug 448162 in f-spot "sidebars pictures are not clickable" [Low,Triaged] https://launchpad.net/bugs/448162 [16:23] thanks [16:23] testbuilding [16:24] thank you [16:48] rickspencer3: do you know whom to subscribe to a bug report about "Add Service Tag and support URL in System Monitor"? [16:48] rickspencer3: (someone from OEM?) [16:49] yes, bfox [16:49] thanks [16:50] eep I'm subscribed to a blueprint for a UDS session [16:50] Amaranth: I just subscribed you, yes [16:50] I'm not going to be there :P [16:50] thought you were interested in following/commenting on the blueprint [16:50] sure [16:51] cube by default == no way :) [16:51] what blueprint? [16:51] Amaranth: if you aren't, please unsubscribe again and accept my apology about the spam [16:51] oh, set of effects to use [16:51] Amaranth: see, that's why I subscribe you :) [16:51] cube by default == yea baby [16:51] also, flaming cursor [16:51] rickspencer3: it looks ridiculous on slower hw, though [16:51] rickspencer3: we'd have to switch to 4 workspaces by default too [16:51] pitti, don't fall into the troll [16:51] hehe [16:51] see, the subscription was useful :) [16:51] thanks seb128 [16:52] ;-) [16:52] * pitti ignores the flamewar^Wbikeshed^Wdiscussion and goes on with planning [16:52] There are some things we should change though... [16:52] for example, the video plugin is completely useless [16:52] I thought the "flaming cursor" part was a bit of give away [16:52] rickspencer3: I liked that part [16:52] Amaranth, anything we can remove to speed up load time? [16:53] rickspencer3: You can test my branch to speed up load time and see if that helps :) [16:53] sweet [16:53] tbh, I would like to turn on the cube if possible [16:53] but sounds like there will be some serious issue [16:53] otherwise if we strip plugins we consider worthless it could help a little too [16:53] less for it try to find [16:53] but when I show the cube, it gets people really interested [16:53] would it make a difference to put everything now on by default in an -extra binary [16:53] and have that not installed? [16:54] now -> not [16:54] maybe add Cube to the High Level of effects option? [16:54] seb128: it could help a little [16:54] Amaranth, a little only? what is all the time spent? [16:54] seb128: it'd at least make people not enable cracktastic things without thinking [16:54] compiz takes 10s if there is nothing to read too? [16:54] seb128: well most of the time seems to have been spent in that shell script... [16:55] otherwise it only loads information for plugins it has loaded [16:55] ok, let's see how it goes with your version first [16:55] so not shipping a plugin would save a stat call or whatever [16:55] on the other hand the ccp plugin loads information for every single plugin [16:55] it's weird that a piece of shell takes 15 seconds to run [16:56] but it uses protobuf so it's pretty fast and there isn't as much IO [16:56] compiz itself uses DOM + XPath to load XML for the plugins that are loaded [16:57] robert_ancell seems to have added something to compiz to tell what part is taking the most time, he might have a better idea of where any potential bottleneck actually is [16:57] at least it loads faster than gnome-shell ;) [16:59] gconf is a bottleneck on my desktop [16:59] 4 seconds [17:00] yeah, that'd be ccp [17:00] so not shipping as many plugins would help some there [17:00] since ccp loads the info for all of them [17:00] oh, i meant gconfd starting is the bottleneck on mine [17:00] oh :) [17:00] this is before compiz starts [17:02] wow, i wish my company would spend some money on some decent laptops [17:17] pitti, dpm: we have https://blueprints.edge.launchpad.net/ubuntu/+spec/community-lucid-kubuntu-translations-feedback-and-improvements and https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-kubuntu-translations I propose to replace them with kubuntu-lucid-translations [17:18] Amaranth: I did a profiling patch some tme ago for compiz, should be in bzr [17:18] mvo: oh, that was you [17:19] asac: can we have a kubuntu firefox session to look at the KDE firefox integration bits? [17:21] Riddell, that's fine by me, but I'd like to be able to add info to the spec, at least move the summary in the community blueprint to the desktop one [17:22] Riddell: WFM; I didn't really put a lot of meat into it, so we can just invalidate it [17:23] Riddell, also, the -lucid- name I think should be kept for consistency with other sessions on the Desktop track [17:23] seb128: I added SRU stuff and opened a Karmic task on bug 448162. We'll get it to Lucid by a sync. [17:23] Launchpad bug 448162 in f-spot "sidebars pictures are not clickable" [Low,Fix committed] https://launchpad.net/bugs/448162 [17:23] dpm: I'm planning on using kubuntu-lucid for all our kubuntu ones, I have scheduling powers now I'm told so we're partly a track to ourselves [17:24] Laney, thanks [17:24] Riddell: done [17:24] Riddell, ahaaa, then feel free to use the superpowers and move it to the Kubuntu track [17:24] thank you for triaging and chasing [17:24] pitti: what have you done? [17:25] Riddell: marking desktop-lucid-kubuntu-translations as superseded by community-lucid-kubuntu-translations-feedback-and-improvements ? [17:26] pitti: I wanted to make a kubuntu-lucid-translations the superseed them both [17:26] Riddell: ah, there's a third one; sure, please go ahead [17:26] there will be [17:28] you can also just rename the existing one, FYI [17:28] Riddell: ^ [17:28] (if everything else in it is correct, that's easier) [17:29] so I can, that's clever [18:00] pitti, ccheney` I was thinking about having a dedicated session on OO.o translations, but seeing that there is https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-openoffice already, do you think it might be a better idea to discuss translations in there as well instead? [18:31] djsiegel: ping... sent the mail.. sorry, for sending it late ... i forgot about it :( [18:31] oh, thanks mac_V! [18:31] n [18:31] np [18:33] :) [18:40] hmmmm, i'm really confused about how screensaver inhibiting works at all in totem [18:40] it calls Inhibit on the wrong dbus path [18:40] :-/ [19:11] chrisccoulson: perhaps that's the s3kr1t bus that works :) [19:37] chrisccoulson: if you look at gnome-screensaver, it actually registers an object at that path [21:04] mclasen - sorry, i went away [21:04] my gnome-screensaver only registers a path at "/" [21:04] :( [21:04] the other one doesn't show up in d-feet [21:05] ah [21:05] don't know why [21:05] mclasen - totem always gives me this warning: [21:05] WARNING **: Problem inhibiting the screensaver: Method "Inhibit" with signature "ss" on interface "org.gnome.ScreenSaver" doesn't exist [21:05] but the inhibit is registered [21:05] oh, interesting [21:06] and this is why totem isn't cleaning them up properly for me, because it *thinks* it failed to register it, and so doesn't save the cookie [21:06] i'm a bit confused why i seem to be the only person with this problem though [21:08] did you try with an another user or a guest session? [21:09] seb128 - one second, i will try that now [21:12] hmmm, that's wierd - it works ok from the guest account [21:43] well, i'm totally confused! [21:47] Ampelbein, hum, if totem is crashing that's a totem problem ... not true [21:48] ups [21:48] Amaranth, ^ [21:48] seb128: it's certainly not a compiz problem :) [21:48] Amaranth, and not a totem one either [21:48] seb128: totem or one of the libraries it uses (nvidia fail again?) [21:49] I just hate those bugs about video crashing when compiz is on [21:49] nobody wants to take responsability for those [21:49] heh [21:49] I would like to know where to reassign it [21:49] totem is wrong but xorg or compiz would bounce back [21:49] If it's not a crash in compiz I have no idea [21:49] so we are stucked with useless replies [21:50] see if you get can a backtrace from him or something [21:50] I did ask for one [21:50] but then again you can't tell with nvidia and X, only with nvidia OpenGL [21:50] (if the nvidia driver is the cause, I mean) [21:50] but I guess it will be one of those xerror due to limited video ressources [21:56] hey robert_ancell [21:56] seb128, hey [21:57] robert_ancell, thanks for the update, some notes though [21:57] always with the notes :) [21:57] 1, could you please do the merges on debian? [21:57] yeah, I knew you were going to get me on 1 [21:57] I know it's no fun [21:57]