=== robbiew is now known as robbiew_ [00:55] kenvandine, therE? [01:26] hyperair - are you still there? === asac_ is now known as asac [07:28] itorrey: i'd suggest you ask david for the color scheme and stuff.. ;) you'd always have people complaining of too brown , too orange :) [07:29] Thanks Vish. Should I reassign the ticket to david? [07:30] itorrey: no need to re-assign , but rather take directions from the design team [David is a part of that] rather from every user who complains ;) [07:30] itorrey: you can ping him when he is around [07:30] or mail him [07:31] Sounds good. Thanks vish [07:31] np.. [07:47] good morning [07:48] Good morning [07:48] * bryyce waves [07:51] robert_ancell: hey [07:51] pitti, hey [07:51] bryyce, didrocks: hey too [07:51] robert_ancell: lp-i> I suppose it's used in xubuntu and other derivatives, too, so it went into the general core set [07:52] hey robert_ancell and pitti :) [07:52] morning bryyce [07:52] I think this requires discussion with cjwatson first, whether these can be changed at all (I don't think it's that easy; it would break the separation between derivatives) [07:52] pitti, right [07:52] bonjour didrocks [07:52] red tape :) [07:52] robert_ancell: need some sponsoring? [07:52] pitti, no, seb128 did it earlier today [07:52] thanks though [07:54] heya robert_ancell and didrocks, hope you had a good weekend [07:56] bryyce, was travelling so short weekend, but I have a public holiday on this weekend so will be better :) [07:57] bryyce: nice and restful, thanks :) [07:58] hum, I don't understand why I have one hour of delay between my google calendar and evolution. Something with timezone surely, but… (the google calendar web interface have the right time in my timezone, not evolution) [07:59] oh ok, importing between calendar doesn't take into account timezones :} [07:59] and as my timezone on google calendar wasn't set… [08:01] hello [08:02] salut baptistemm [08:03] salut didrocks [08:16] bonjour baptistemm [08:16] baptistemm: I fixed versions.html, sorry for that [08:16] salut pitti [08:17] pitti: apport doesnt seem to be working for me in lucid [08:18] --verbose ? [08:19] * fagan give it a go [08:20] fagan: I meant, what is not working? [08:22] pitti: It was just missing simplejson [08:22] Im testing a machine upgraded from hardy [08:23] So there is some strange things going on [08:23] Package: python-launchpadlib [08:23] Depends: python-simplejson | python (>= 2.6) [08:23] hmm [08:23] perhaps that's wrong? [08:23] python 2.6 does come with a json module [08:23] but it's named differently [08:24] Well it works now [08:24] So it must be something with the python version jump that lost a couple of modules [08:26] fagan: it worked after you installed a python package? it might be a issue with the python tool we have :/ do you hvae the full logs? [08:27] Yep what logs do you want mvo [08:28] fagan: the stuff in /var/log/dist-upgrade/* should give a clue and probably /var/log/apt/term.log [08:28] and maybe /var/log/apt/history.log [08:30] http://pastebin.ubuntu.com/358920/ [08:31] Theres a lot of not installed [08:32] Thats the apt log for the update [08:36] what was the error when apport wouldn't run? the exact error message? [08:37] http://pastebin.ubuntu.com/358922/ [08:37] mvo: it should be in here [08:37] * mvo looks [08:38] Oh it wasnt [08:38] Damn [08:40] Give me a sec to report some bugs [08:41] hey there [08:41] hey seb128_ [08:41] salut seb128_ [08:41] hi fagan [08:42] hey didrocks [08:42] I need to debug g-s-d one day [08:42] it keeps crashing when switching screens between docks and non docked [08:45] mvo: oh and I reported a bug against the software center too https://bugs.launchpad.net/ubuntu/+source/software-center/+bug/509518 [08:45] Error: This bug is private [08:46] whoops I changed it to public [08:48] seb128_: btw, bug #509249 when you have some time :) [08:48] Launchpad bug 509249 in cdbs "make list-missing file part more remarkable in utils.mk" [Undecided,Triaged] https://launchpad.net/bugs/509249 [08:48] thanks [08:50] hey chrisccoulson [08:50] so pitti is quicker to fix bugs than me to report them again ;-) [08:50] pitti, thanks for bringing back the versions.html [08:51] hey seb128 [08:51] how are you? [08:51] a bit tired but good [08:51] you? [08:51] fagan: thanks, I have a look [08:51] hey seb128_ [08:51] did rocking job on g-p-m I see ;-) [08:52] hey mvo, how are you? [08:53] seb128_ - yeah, i'm good thanks. a little less tired today :) [08:53] seb128_: good! unhappy about the treeview, but otherwise great [08:53] mvo: hi... is the cause for Bug #50837 know? rather is it easy to fix? [/me wants to invalidate papercut task ;) ] [08:53] Launchpad bug 50837 in synaptic ""Mark All Upgrades" removes all previous markings" [Undecided,Confirmed] https://launchpad.net/bugs/50837 [08:53] fagan: thanks, did that happen during the upgrade? or when it was finished? [08:54] known* [08:54] mvo: when its finished [08:54] I have it now [08:54] vish: not very hard, not very easy [08:54] do the brightness notifications still work properly for everybody with the g-p-m update? that's something i can't test unfortunately [08:54] hmm.. ;) [08:54] fagan: oh, then please leave the system in that state [08:54] pitti: https://bugs.launchpad.net/ubuntu/+source/hal/+bug/509524 [08:54] Error: This bug is private [08:54] fagan: what is ls /var/cache/software-center showing? [08:55] xapian [08:55] seb128_ - would you object if i added NoDisplay to the gnome-screensaver desktop file, so that it doesn't appear in the session capplet? [08:55] fagan: and under that subdir? [08:55] mvo, what do you blame gtk for today? being complicated to use? having limitation? [08:56] i keep coming across issues like bug 509411 [08:56] Launchpad bug 509411 in gnome-screensaver "lock doesn't lock" [Low,New] https://launchpad.net/bugs/509411 [08:56] chrisccoulson, yes! [08:56] seb128_ - thanks [08:56] you're welcome ;-) [08:56] chrisccoulson, joke aside, the autostart one? [08:56] fagan: I think you need to mail me the full logs, searching in a pastebin is a bit cumbersome :) please tar czvf logs.tar.gz /var/log/dist-upgrade/* /var/log/apt/ ? [08:57] seb128_ - yeah, the autostart one [08:57] seb128_: I'm unhappy that its slow when not in fixed_height mode, the design team wants rows to grows dynamcially [08:57] chrisccoulson, seems to make sense to me [08:57] seb128_: we have that now, it looks great, but it makes everything sluggish [08:57] cool :) [08:57] mvo, :-( [08:57] seb128_: mind you, its *much* better than it used to be :) [08:57] mvo: What address do you want me to send them to? [08:58] lol [08:58] that's something I guess [08:58] but still, i need a "almost_fixed_height" [08:58] mode [08:58] fagan: mvo (at) ubuntu.com is fine [08:58] fagan: many thanks! [08:58] Cool [08:59] fagan: and the output of ls "/var/cache/software-center/xapian" too please - but then I should have most of what I need [08:59] (i hope :) [08:59] seb128_: you don't happen to have good connections to a treeview hacker that can give me that mode ;) ? [09:00] mvo, I'm sorry but no [09:00] mvo: nothing in there [09:00] * fagan suspects thats the problem [09:00] mvo, I managed to get the triple click bug fixed though [09:00] mvo, by Ccing the right people to the bug ;-) [09:00] fagan: yeah [09:01] seb128_: sweet [09:01] * mvo hugs seb128_ [09:01] seb128_: who fixed it? [09:01] mvo, but those guys are too busy to work on non bug fixes right now I expect [09:01] mvo, kris [09:01] he would be the right guy for what you need too [09:01] but I think he's too busy and that's probably not a priority to work on [09:02] seb128_: yeah, I guess - I will have a look at the code and maybe I get a idea what could be done [09:03] mvo: its fixed if you "sudo software-center" [09:03] So fine now [09:04] * bigon whisper sothing to seb128 about empathy beiing stuck in the new queu [09:05] bigon, why is empathy still going to new queue? we don't export libs anymore... [09:05] pitti: I think you forgot to push the last commit of your gdm branch (it's still UNRELEASED) [09:05] * fagan restarts to see if xorg is fixed by the update [09:06] Zdra, he added a -dbg binary [09:06] bigon, newed [09:06] didrocks: sorry, pushed [09:06] seb128_, ah ok :) [09:06] hey seb128_ [09:06] hey pitti [09:06] pitti: thanks :) [09:08] #ubuntu-x==epic win [09:10] mvo: no xorg error when I boot now :) [09:10] that is something [09:10] :) [09:20] Zdra: you highlight on empathy? :p [09:20] bigon, yep [09:26] bbl, trying my gdm-set-default-session without gdm running [09:27] didrocks, good luck [09:32] seems it didn't crash badely :) [09:33] seb128_: is the rgba stuff still coming for lucid? [09:37] plop [09:37] hello everyone ! [09:38] salut huats [09:39] hello didrocks :) [10:13] pitti: You take care of hal right [10:13] bad_fagan: I'm working hard to kill it for good, yes :) [10:14] pitti: I got this on a hardy to lucid upgrade https://bugs.launchpad.net/ubuntu/+source/hal/+bug/509524 [10:14] Ubuntu bug 509524 in hal "hald-probe-input crashed with SIGSEGV in posix_memalign()" [Medium,New] [10:15] * bad_fagan is fagan but in college [10:15] bad_fagan: dupe of bug 500723 [10:15] Launchpad bug 500723 in hal "hald-probe-input assert failure: *** glibc detected *** /usr/lib/hal/hald-probe-input: munmap_chunk(): invalid pointer: 0x080492f9 ***" [High,In progress] https://launchpad.net/bugs/500723 [10:15] Ah [10:15] it's on my TODO list [10:17] * bad_fagan logs off before his lecturer comes around :) [10:35] re [10:35] ok, xorg keeps crashing on guest session [10:35] enough playing with that for today [10:36] didrocks, just curious, does you gdm change fix any open bug or known crasher? [10:38] seb128: thx for the newing, if you still have some time could you have a look at https://bugs.edge.launchpad.net/ubuntu/+source/farsight2/+bug/508182? [10:38] Ubuntu bug 508182 in farsight2 "Please merge farsight2 (0.0.17-2) from debian unstable" [Wishlist,New] [10:39] I've other things I want to get done now but maybe mvo or pitti can sponsor this one? [10:39] can do [10:39] I will have a look later otherwise [10:39] pitti, thanks [10:40] seb128: no, just cleaning before submitting upstream (gtype can't normally be casted to gobject) [10:40] I've tried to contact cjwatson about my upload rights but got no answer :( [10:49] looks like I was too late :/ [11:01] pitti, http://people.canonical.com/~seb128/versions.html [11:01] ooh [11:01] the new box running hardy works again [11:01] so you got it working on dapper? [11:01] no, they moved people to a new box [11:01] Linux rookery 2.6.15.7 [11:01] oh, they did? [11:02] read your emails ;-) [11:02] lillypilly, nice [11:02] seb128: nice page :) [11:02] mvo, thanks ;-) not new, we use that for updates for a while [11:02] seb128: I have seen it before, but its still nice [11:02] seb128: ok, so I can retire my gnome-versions installation on my server? [11:02] mvo, you will notice g-t is ready for sponsoring [11:03] mvo, if you want to do one sponsoring upload ;-) [11:03] baptistemm: ^ so I'll remove my versions.html for good now :) [11:03] pitti, yes, thank you for providing us service for the time during! [11:03] my pleasure [11:03] pitti, so I will stop complaining it's broken every 2 days :) [11:04] seb128: hm, seems i missed that mail somehow; don't see it === seb128 changed the topic of #ubuntu-desktop to: Welcome to #ubuntu-desktop - home of the Desktop team - http://lists.ubuntu.com/mailman/listinfo/ubuntu-desktop/ | If you want to help out, check out https://launchpad.net/~desktop-bugs/+bugs?field.milestone%3Alist=21439 or http://people.canonical.com/~seb128/versions.html | For support please join #ubuntu [11:05] seb128: will do [11:05] kees: here? [11:05] baptistemm, to be fair to pitti it worked fine all this time, we just got 2 recent glitches [11:06] well, the "glitch" was me rm -r'ing workitems/desktop/lucid [11:06] since the WI tracker moved to the DC, and I forgot about versions.html [11:06] yeah, I'm not sure why it was in workitems dir [11:07] it seemed to fit [11:07] seb128: so ... about the treeview - I got something here that makes it fast for me and still able to grow individual rows [11:08] mvo, without breaking the rest of the world? ;-) [11:08] *well* [11:08] lol [11:08] works here on my box since ~5min, is that enough? but serisouly, what needs to be done to get it in? I'm not sure upstream will be supportive given that fixed_height is normally about fixed_height [11:10] mvo, well if you recommend to change upstream sementic, no way [11:10] we don't want gtk behaving differently on ubuntu [11:11] seb128: it will not break any semantics, just bend the rules. but I will have a further look into how it could be done in a way that upstream may accept [11:11] could you explain what it changes exactly? [11:11] in any case opening an upstream bug to start discussion is always a good idea [11:11] they might have valid point [11:11] or good suggestions to do what you want [11:12] or agree with the change who knows [11:12] let's try ;-) [11:12] seb128: it assumes all rows are equal (just like before). but on row-changed signal, it checks the size of the changed row (even in fixed_height mode) and updates the visible area [11:12] should that be slower? [11:12] shouldn't [11:13] yes, but the cost is marginal [11:13] because instead of querying "get_size()" for each row, its done once + re-validation of the handful of visible rows [11:13] I measured it and it was really really little AFAICS [11:13] I'm not strongly opposed to it [11:14] but I would like to see upstream discussion happening before [11:14] ok, I will try to bring it up upstream first and see what they say [11:14] just to be good citizens [11:14] and because they know their code and might have good points to make too [11:14] thanks [11:14] maybe we can come up with a "almost_fixed_height" mode or "curor_row_now_fixed_height" [11:14] what user visible issue are you trying to solve? [11:14] that would be my alternative suggestion, thread everything fixed height, the only exception is the seclted or cursor row === MacSlow is now known as MacSlow|lunch [11:15] seb128: software-center should grow support for dynamically expanding the cursor row with additional "info" and "install" buttons [11:16] seb128: this is done now, but it makes stuff slow, its not too bad, but unneeded, e.g. the pathbar code is no longer smooth etc [11:17] mvo, I see [11:20] seb128, it was only a joke [11:20] ;-) [11:25] seb128: Can we change the default laptop lid closing behavour for lucid? [11:26] Please :) [11:26] to be what? [11:26] back screen [11:26] At the moment it suspends [11:26] no [11:26] right [11:26] it's the less destructive way [11:26] some people apparently close the lid and put their laptop in their bag or something [11:27] and get very angry when they notice it didn't suspend and got not smooth moves with disks in use [11:27] Ah I suppose thats ok then [11:28] I completely forget about the default behavour and I close the lid a lot and forget to change the setting [11:28] Ah ill just have to remember [11:29] seb128: I do that, and I think upstream default is great :) [11:30] didrocks: Oh I keep forgetting to do the docs for quickly, ill try get it done this weekend [11:30] fagan: perfect, thanks. You can look at the WI tracker, there should be some for you :) [11:31] didrocks: WI tracker? [11:31] * fagan is bad with lingo [11:32] fagan: work item [11:32] Ah [11:37] Oh yay on there :) [11:37] * fagan was looking at something else [11:50] pitti - shall we change the gnome-screensaver lock-on-idle policy to the upstream default this cycle? (i think we mentioned it a couple of months back) [11:50] chrisccoulson: like, making locking on timeout and suspend consistent and on? [11:50] I'm all for this; seb128, WDYT? [11:51] pitti - yeah. i was just thinking of setting the default so that it locks when the screensaver activates (which is the upstream default) [11:51] it already locks on suspend etc [12:04] re [12:04] I was lunching [12:05] chrisccoulson, pitti: the change would be from what to what exactly...? [12:05] seb128 - currently, we change the upstream defaults to not lock the screen when the screensaver activates [12:05] rational? [12:05] ie why did we do that? [12:05] but i was suggesting to use the upstream default, where the screen locks when the screensaver activates [12:06] i'm not sure why we made the change in the first place [12:06] pitti - any ideas? [12:06] no idea, it's been like that for ages [12:06] apparently someone didn't like it, or so [12:07] I've no strong opinion either way [12:07] with the recent grumblings from users about keyring access and being able to view keyring passwords in plaintext, i think we should be locking the screen by default [12:07] I like the current behaviour, but it might make sense to lock boxes when people walk away for a while [12:08] but I tend to be around and move the mouse to get screen back when the screensaver kick ine [12:08] in [12:08] which is probably not a common usecase ;-) [12:09] I'm fine going with upstream default [12:09] it's the "home desktop computer" use case [12:09] but then that's the same issue with suspend/resume [12:09] that's why I'm for consistency [12:09] right [12:11] yeah, i don't mind either way. personally, i turn locking on normally. but then, i lock the screen when i walk away even if there's nobody else around [12:11] perhaps i'm just paranoid ;) [12:11] well I do lock the screen when I want it locked [12:11] otherwise I want to be able to move the mouse to see if somebody talked to me there [12:12] but again we will never match all usecases that's why that's a config option you can set ;-) [12:12] let's default to upstream and consistency [12:12] and see what user feedback is [12:13] seb128 - ok, i'll do that tonight then, along with the change to the desktop file too [12:13] thanks [12:13] that will be a rebuild with the new libxklavier too ;-) [12:13] do you have any g-c-c change pending btw? [12:14] I will fix the typo I think and upload [12:14] so we get it rebuild with the new soname too [12:14] seb128 - i haven't got any g-c-c updates [12:14] pitti, btw any clue how to figure what triggers hal on a box? [12:14] chrisccoulson, ok [12:15] pitti, on my laptop all the hal binary start on boot [12:15] and the device input thing crash every time [12:15] binaries [12:15] rather ;-) [12:15] seb128: please check if it starts before or at gdm, or only at session start [12:15] if it's the latter, then I suspect your graphics card doesn't have xrandr brightness support [12:16] heh, i hope my g-p-m update didn't make it do that ;) [12:16] pitti, between xorg and gdm on the chart [12:16] chrisccoulson, no, it's doing that for weeks [12:16] btw, did anyone test their brightness keys with the new g-p-m? (ie, do the notifications still work ok). I don't have any hardware to test those [12:17] seb128: weird, I thought I fixed gdm last week [12:17] chrisccoulson: WFM [12:17] pitti - cool, thanks :) [12:17] chrisccoulson: oops, sorry, didn't test with 2.29.1 yet [12:17] will do later [12:17] pitti, let me reboot with today's updates [12:18] seb128: g-p-m is running in the gdm session, doesn't it? So I figure that triggers it === MacSlow|lunch is now known as MacSlow [12:25] re [12:25] pitti, http://people.canonical.com/~seb128/bootchart/seb128-laptop-lucid-20100119-6.png [12:25] see before gdm [12:26] I might have some old config in etc or something not sure [12:27] seb128: how do you generate a bootchart like that? [12:27] fagan, install bootchart and pybootchartgui [12:27] and look in /var/log/bootchart [12:27] Cool [12:27] it stores those images for each boot [12:28] I'm wondering why vino is using full cpu for 3 seconds for there [12:30] chrisccoulson, gpm notify-osd works fine on my mini config [12:31] seb128 - thanks for testing [12:31] np [12:31] chrisccoulson, it's slighty buggy in fact [12:32] it doesn't do the overshoot effect now [12:32] seb128: is the transparent windows gtk improvements fixed yet for lucid? [12:32] seb128 - thanks, i'll take a look at that [12:32] fagan, no, delay to next cycle [12:32] delayed [12:33] Awh [12:33] * fagan really wanted it :/ [12:33] I suppose you should intoduce something that big into an LTS [12:41] seb128: g-p-m is running in the gdm session, doesn't it? So I figure that triggers it [12:42] pitti, it does [12:42] seb128: try this: [12:42] sudo killall hald [12:42] pitti, but didn't you fix that? [12:42] killall gnome-power-manager [12:42] gnome-power-manager [12:42] does that start hal? [12:42] if so, check --verbose [12:43] it should say that no xrandr hardware is available, and fall back to hal [12:43] $ gnome-power-manager [12:43] ** (gnome-power-manager:3185): WARNING **: Either HAL or DBUS are not working! [12:43] ** (gnome-power-manager:3185): WARNING **: proxy failed [12:43] etc [12:43] yes it does [12:44] yes, that's normal [12:44] but it shoudl still work [12:44] seb128: ok, then it's working as intended [12:44] well all hal is started [12:44] seb128: I'll work on the input failure next; just need to go to the supermarket and for a haircut [12:45] oh? is hal started because of the backlight? [12:45] correct [12:45] ok good [12:45] let me know if you need details on the input issue later [12:45] same issue on my notebook [12:45] the netbook can do xrandr and doesn't start hal [12:46] right, same here [12:46] seb128: I can perfectly reproduce the input crash [12:46] just need to sit down and debug it [12:46] I'm a bit tight on time today, though [12:46] ok [12:46] have fun shopping and getting hairs cut ;-) [12:46] need to look a bit after my grandfather, and two meetings, too *sigh* [12:47] ok, good luck [12:47] I will not stop your longer to get going ;-) [12:51] pitti: test [12:51] pitti_: pong [12:51] (sorry for the noise) [12:53] * hyperair is envious of seb128's bootchart. [12:53] why do all the bootcharts i look at seem faster than the ones from my system? [12:53] it's slow... [12:53] * hyperair sighs [12:53] hey that's 45s [12:53] mine's 90s! [12:53] the mini is to desktop is 18 seconds [12:53] and still not done. [12:54] T_T that's so damn fast. [12:54] my login is particularly long for some reason [12:54] in particular, gnome-panel seems to like to chew my disk. [12:58] how slow is your disk? [12:58] you might want to try to clean gtk bookmarks, recently used, thumbnails [12:58] and see what difference it makes [13:03] chrisccoulson, seb128: I'm not sure locking the screen when screensaver is triggered is a good idea [13:03] what e.g. if you leave your computer open for friends to use it (choosing music...) [13:04] milanbv - the guest account helps there if you want to leave it unlocked for friends to use (the screensaver is disabled there) [13:04] hm, is that a realistic case [13:04] and rhythmbox inhibits the screensaver when playing music etc [13:05] also, totem inhibits it too [13:05] I mean: I'm using deezer.com to play music, and want my friends to be able to change it [13:05] I know, that's lousy because it's the only case that doesn't inhibit [13:05] well, that case wouldn't work as well. but then, you could just change the preference to not lock anyway [13:06] but the point is, people don't like their system to ask for password all the time [13:06] the upstream default is to lock the screen when the screensaver activates [13:06] I'd rather users keep their screensaver activating very quickly, and not lock the screen [13:07] e.g. my ss timeout is about 2 min - which would make typing the password everytime boring [13:07] it's one of those cases where it is impossible to please everybody [13:07] is there a precise rationale upstream to lock the screen automatically? security? [13:08] * hyperair prefers the screen locked automatically. [13:08] i assume it's for security [13:08] why? :-) [13:08] security [13:08] i do go walking away from my computer sometimes without locking, and i'd like it to lock itself if i don't come back fast enough. [13:09] I'm all for security, but most people I know lough at passwords [13:09] lough? [13:09] so they use weak password (birth date), and give them to everybody [13:09] *laugh [13:09] ah [13:09] some do that [13:09] i don't [13:09] milanbv - there was a lot of fuss recently from users over the ability to view plaintext passwords in seahorse once the keyring is unlocked, and one of the common complaints was that people could view your passwords if you walked away from your machine for a while [13:09] ah, I see [13:09] having the screen locked automatically is good security practise [13:10] milanbv: even if you do tell everybody your password, it'd be to people you know, right? [13:10] but users can still change the preference if they don't like it [13:10] milanbv: you're not going to announce to random strangers your password. [13:10] but ideally, we'd need a different (longer) delay then [13:10] the default delay is currently 10 minutes [13:10] blanking screen to save power should occur in 2min [13:10] that's fine imo [13:11] blanking screen != activating screensaver. [13:11] screensavers take up more power, if anything. [13:11] yeah, that's what I was wondering [13:11] OK, so we have blanking in 2 or 5 min, and locking in 10min? [13:11] with this, I'd be fine [13:11] yeah, gnome-power-manager throttles the screensaver when you're on battery power for that reason [13:12] blanking time shorter than locking time is not possible with the current architecture [13:12] chrisccoulson: throttle how? i saw some code regarding it while poking around, but i didn't really understand what it did. [13:12] hyperair - it should disable animated screensavers [13:12] so you just get a blank screensaver on battery [13:12] so do we have one timeout setting, or two? and what are they? [13:13] chrisccoulson: ah i see. i disabled them anyway, so i never noticed. [13:13] milanbv: two timeouts. one in gnome-screensaver for locking/screensaver, and one in gpm for lowering the backlight slightly [13:13] OK [13:13] imo the one in gpm could perhaps be extended to completely turning off the backlight. [13:13] oh wait [13:14] I think we would need 3 of them :-p [13:14] there's one [13:14] "Put display to sleep after _______" in gnome-power-manager [13:14] isn't that the one you're looking for? [13:14] milanbv - gnome-screensaver activates when the session goes idle (indicated from gnome-session) [13:15] gnome-power-manager blanks the screen some period after it goes idle [13:15] yeah, is it possible (or is it the case now) to put something like gpm in 2 min, ss in 10 min? [13:15] so that the sreen is blank quickly, and the lock only occurs after a longer period? [13:15] chrisccoulson: speaking of screensaver, i've noticed a lot of screensaver unpleasantries in karmic. stuff like it activating while i'm playing a video. i wouldn't have minded as much if it'd notice me shaking my mouse furiously while it was fading, but it'd take its time to fade before i can deactivate it [13:16] hyperair - there's been a few issues, but most of them should be fixed now [13:16] what media player do you use? [13:16] chrisccoulson: the thing is, they're not. [13:16] chrisccoulson: mplayer/gmplayer. [13:16] i've amended my mplayer config to --poke gnome-screensaver-command [13:16] but gmplayer doesn't have a heartbeat cmd [13:16] hyperair - mplayer doesn't support inhibiting gnome-screensaver [13:17] chrisccoulson: mplayer warps the mouse pointer or something in an effort to make the screensaver think the mouse moves around. [13:17] and gnome-screensaver-command --poke is broken in karmic [13:17] is --poke broken? i thought i saw a changelog entry abotu it being fixed [13:17] anyway i don't notice any issue while using mplayer now [13:17] hyperair - yeah, i rolled back the patch though, as it caused another issue [13:17] but gmplayer still has issues [13:17] ugh [13:17] what issue? [13:18] randomly popping up menus in openoffice [13:18] huh?! [13:18] oh god [13:19] a keypress huh [13:19] the change to gnome-screensaver copied some code from elsewhere to simulate fake keypresses with xtest [13:19] shouldn't gnome-screensaver-command use a dbus interface?! [13:19] good god. [13:19] this is really stupid. [13:19] hyperair - it does, but gnome-screensaver needs to reset the IDLETIME counter on the server somehow [13:20] chrisccoulson: and why can't it? [13:20] gnome-screensaver does no timing itself (there are no counters to reset there) [13:20] IDLETIME is reset by user activity [13:20] i see. [13:20] there is another way that is supposed to work, but is currently broken in karmic and lucid due to a xorg bug [13:21] ah hell [13:21] there's no nice way of doing it at the moment, which is why i rolled back the original patch [13:21] can't we just have it in place for those who don't have broken xorgs? [13:21] we all have broken xorgs though ;) [13:22] i don't. [13:22] there's no acceptable patch to fix the xorg issue yet [13:22] i use xorg-edgers. [13:22] the PPA, that is [13:22] hyperair: that's broken too [13:22] oh hell >_> [13:22] so what's this way that's broken in Xorg? [13:22] XResetScreenSaver? [13:23] yeah, that's broken [13:23] it doesn't reset IDLETIME anymore. that happened as part of a fix to sort out race conditions with dpms blanking [13:24] i see [13:24] sometimes i wonder if X isn't a buggy pile of shit [13:26] hyperair - the xorg bug is here: http://bugs.freedesktop.org/attachment.cgi?id=32398 [13:26] thanks [13:26] there is a patch, but, AFAIK there were some review comments on xorg-devel [13:27] what you just linked me was a patch [13:28] hyperair - oops: http://bugs.freedesktop.org/show_bug.cgi?id=25855 [13:28] Freedesktop bug 25855 in Server/general "Screensaver not disabled because of a XResetScreenSaver() regression" [Normal,New] [13:28] the review is here btw: http://lists.x.org/archives/xorg-devel/2010-January/004503.html [13:28] ah thanks [13:28] the comments look quite trivial. i'll chase that up if nobody pushes it forward, and see if bryyce will accept it in to lucid [13:29] and then i can fix gnome-screensaver to use it [13:29] yay [13:49] ok, I'm off to go to my grandpa, will take laptop/mobile with me to get online again [13:50] bbiab [14:05] chrisccoulson: if g-ss doesn't know how to reset X11's timer, then does that mean --inhibit will completely disable the screensaver? [14:05] even after uninhibited [14:08] hyperair - --inhibit will prevent the screensaver from activating [14:08] chrisccoulson: yes, but what about after the inhibit is removed? [14:08] then the screensaver can activate again [14:08] chrisccoulson: what happens then? if the timer expires while inhibited, then what happens? [14:09] hyperair - if there are inhibitors registered, gnome-session will not set the session idle, even after the idletime alarm fires [14:09] but if you remove all the inhibitors, it will go back to normal [14:09] normal how? [14:09] does the idletime alarm automatically reset itself? [14:10] idletime is an X thing, right? [14:10] bah, valgrind seems broken [14:10] it returns zillion of errors [14:10] hyperair - yeah, it's an X thing [14:10] seb128 - what are you debugging? [14:10] I'm trying to debug nautilus [14:10] which was working yesterday [14:11] but exit immediatly because > 1000 error now [14:11] ouch ;) [14:11] they're not real errors then? ;) [14:11] all in ==5869== at 0x4A5679F: __strlen_sse2 (strlen.S:106) === bjf-afk is now known as bjf [14:11] I'm wondering if glibc is buggy [14:12] hmmm, yeah, i don't know about that [14:19] tedg, hello [14:19] Good morning seb128 [14:19] seb128: I filed gnome bug #607447 (with demo and patch) [14:19] Gnome bug 607447 in GtkTreeView "treeview almost fixed_height mode" [Enhancement,Unconfirmed] http://bugzilla.gnome.org/show_bug.cgi?id=607447 [14:20] seb128: lets see :) [14:20] mvo, thanks [14:21] tedg, sometime the corner menus are empty there [14:21] tedg, is that a known issue? how would to debug? [14:21] seb128: There's a bug open, but it seems like no one can reproduce :( [14:22] seb128: The first thing would be to see if the services are running? [14:22] I got like 5-6 times this week [14:22] which ones? [14:22] seb128: Cool! (not that you're getting a bug, but that we might be able to debug it) [14:22] tedg: I confirm, I got it more than once [14:22] I don't have it right now but I will get soon again [14:22] seb128: The one with shutdown will be indicator-session-service and the other one is indicator-me-service. [14:22] I reboot the mini a lot for bootcharts [14:23] I have a suspicion that one of them is crashing. But, I don't understand why apport isn't throwing a ton of bugs my way if so. [14:30] didrocks, hiya [14:30] hey rickspencer3 [14:31] hey didrocks [14:31] ups [14:31] hey rickspencer3 [14:32] hi seb128 [14:34] seb128: but I still accept your "hey" :-) [14:37] I hope all is well with everyone in #desktop today [14:37] busy but good there ;-) [14:38] same here :-) [14:40] seb128: if I rename libclutk-0.3-doc to libclutk-doc, do I need to provide a dummy transionnal package (as that one only existed during the lucid dev cycle) or no dummy package and just provides/replaces/conflicts like http://paste.ubuntu.com/359042/? [14:40] c,r,p should be enough I guess [14:40] mvo, ^? [14:42] should be fine [14:43] seb128, mvo: thanks [14:47] fagan: hi, re bug #509518 - you do not have a /var/log/dist-upgrade/main.log ? [14:47] Launchpad bug 509518 in software-center "software-center crashed with DatabaseOpeningError in __init__()" [Undecided,New] https://launchpad.net/bugs/509518 [14:47] fagan: how did you do the upgrade "update-manager -d" ? [15:00] pitti, bryyce, tseliot should this blueprint be in a3 now? [15:00] https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-xorg-triaging-diagnosis [15:01] rickspencer3: there's some WIs left for a3 [15:01] rickspencer3: sorry but I have no idea [15:01] * tseliot hasn't worked on that blueprint [15:02] pitti, it's targeted for a2 atm [15:02] should we retarget for a3? [15:02] rickspencer3: doesn't matter much, but if you prefer that, please do [15:02] or does the work items tracker now pick up a3 work items? [15:02] rickspencer3: yes, the WI tracker doesn't care [15:02] pitti, I've lost track of how the work items tracker works :o [15:03] rickspencer3: we don't need to do that retargetting dance any more [15:03] ok === robbiew_ is now known as robbiew [15:39] can someone approve my emails to the ubuntu-desktop mailing list again please? :) [15:39] pitti: do you think it's jockey that blacklists the nvidia modules in /etc/modprobe.d/blacklist-local.conf ? [15:41] kees: I was looking for you [15:41] milanbv: hi, yup. I'm on US Pacific time. :) what can I help with? [15:42] still about users-admin changing passwords [15:42] ah yes. :) [15:42] to avoid all the problem we had with ecryptfs and keyrings, we will now change passwords via PAM, asking for the old one [15:42] admins will have to explicitly require forcing a new password without providing the old one [15:42] for the current user only, I imagine? [15:43] oh [15:43] er, ok [15:43] that's really better, else ecryptfs may be broken [15:43] yup [15:43] but we can't call pam_chauthtok() directly for any user [15:43] since we don't have the privileges required to access /etc/shadow [15:44] I think that's why gnome-about-me runs 'passwd' in the background for that [15:44] does that sound reasonable to you? or is there a better solution? [15:47] milanbv: can't you just make the same PAM calls that passwd does? [15:48] milanbv: you may want to ask slangasek on #ubuntu-devel about this, he knows PAM much better than I do. [15:48] kees - passwd is suid root though isn't it? [15:48] ;-) [15:48] I've discovered that yesterday! [15:49] heh, and we don't want users-admin to be suid root :P [15:49] well, to do any password changing, you'd need root privs. [15:49] either you're calling out to existing tools that have it, or you're writing your own PAM tool that is setuid [15:50] kees - how well do you know PAM? [15:50] yeah, I would have done that in the backends - but the general liboobs model doesn't really handle this well [15:50] only enough to keep from getting in trouble. :) [15:50] ie, would you know why /sbin/unix_chkpwd would exit with exit code 11? [15:50] tseliot: yes, jockey uses /etc/modprobe.d/blacklist-local.conf [15:51] tseliot: it would blacklist nvidia if you disable the driver in jockey [15:51] someone reported that they couldn't unlock their screen from gnome-screensaver, and their syslog shows that unix_chkpwd exits with that error code [15:51] chrisccoulson: isn't that segfault? [15:52] hyperair - no, thats signal 11 [15:52] pitti: ok, so it's that. We don't need that anymore as switching between alternatives is enough [15:52] oh whoops [15:52] it's returning 11 as the exit status ;) [15:52] tseliot: sweet, less custom code to worry about :) [15:52] * tseliot nods [15:52] tseliot: in fact, that's the behaviour of the standard KernelModuleHandler [15:53] so the nvidia one would need to not call that parent function any more [15:53] chrisccoulson: does /sbin/unix_chkpwd allow *any* use to check another user's password? [15:53] chrisccoulson: if /usr/include/security/_pam_types.h can be believed, it's PAM_MAXTRIES [15:54] kees - thanks. i'll have a look when i finish work and see if i can figure it out [15:54] pitti: aah, so it's this line: return KernelModuleHandler.enabled(self) [15:54] s/use/user/ [15:54] milanbv - i'm not sure [15:54] tseliot: no, not enabled() [15:54] tseliot: KMH.disable() [15:55] hmm, so I guess I'll copy/paste the code to run 'passwd' which is used by gnome-about-me [15:55] milanbv - unix_chkpwd only lets you check the calling users password [15:55] pitti: let me guess, XorgDriverHandler subclasses KMH: XorgDriverHandler.disable(self) [15:56] thanks [15:57] tseliot: right [15:57] tseliot: hm, and you do want that [15:58] tseliot: perhaps it's better to add a property "do_blacklist = True" to KMH, and set it to False for nvidia [15:58] tseliot: does it actively hurt to blacklist it? [15:59] pitti: currently we're getting something like this: https://bugs.launchpad.net/ubuntu/+source/jockey/+bug/509376 [15:59] Ubuntu bug 509376 in jockey "[lucid] nvidia-current failed to initialise" [Undecided,Invalid] [15:59] which defeats the purpose of having alternatives [16:00] "do_blacklist = True" should be fine [16:02] hyperair - can you recreate bug 509660? [16:02] Launchpad bug 509660 in gnome-power-manager "Unlock screen not shown when resuming from standby or hibernation" [Undecided,Confirmed] https://launchpad.net/bugs/509660 [16:02] tseliot: I see; yes, let's do that [16:02] chrisccoulson: lemme try. [16:02] pitti: ok, I think I can work on it, not that it's urgent... [16:03] hmmmm [16:03] it didn't work. [16:03] no i mean it came back on [16:03] hyperair - gnome-power-manager already sends a SimulateUserActivity message to gnome-screensaver on resume [16:03] weird [16:03] i'll need to find the exact conditions. [16:03] chrisccoulson: i know, i saw the code for it. i'm just not sure when it happened. [16:03] and gnome-screensaver contains the code to raise the dialog (if you run gnome-screensaver-command --poke from a script, then it raises the dialog) [16:06] hmm [16:06] weird. i can't remember where i've seen it [16:06] i'll try observing for the next few days [16:29] meeting time? [16:29] heya [16:29] hey [16:30] https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-01-19 [16:30] hi [16:30] let's see ... [16:30] hello all [16:30] didrocks [16:30] Riddell [16:30] who else? [16:30] tseliot, [16:30] o/ [16:30] * didrocks waves [16:30] tkamppeter, [16:30] hey [16:31] hi [16:31] ArneGoetje, [16:31] heh [16:31] hey [16:31] hi kenvandine .... I was wondering who I was forgetting [16:31] :) [16:31] hehe :) [16:31] dang it! [16:31] I totally forgot about conference attending *again* [16:31] * rickspencer3 kicks self [16:32] sorry for the rushed agenda, yesterday was supposed to be a US holiday, so I had minimal time to work on it [16:32] so let's just dive in, I suppose [16:32] kenvandine, partner update? [16:32] sure [16:33] DX will have weekly releases, because of the holiday the wiki page isn't complete yet [16:33] but it will be filled in today [16:33] OLS released 0.1.0 of libubuntuone, it is in the u1music ppa now... hopefully a new rb plugin soon... [16:33] i might try to get libubuntuone uploaded to lucid this week [16:33] not a high priority yet though, but sooner the better [16:34] libubuntuone is the gtk widgets for U1... in case anyone missed that last time :) [16:34] that's it for the partner update [16:34] kenvandine, will they have the ability to download the music that you purchased soon? [16:34] before alpha-3 :) [16:35] is there any detail somewhere about what sort of content will be there? [16:35] one of the rhythmbox guys asked about that yesterday [16:35] seb128, just jono signing the free software song, mostly [16:35] lol [16:35] seb128, not sure, but the plugin can browse if you set the right env variable :) [16:35] seb128, I believe these is a rather detailed spec on the wiki [16:36] rickspencer3, not sure if it gives details on the content [16:36] i am sure it is more than just jono :) [16:36] ok, thanks for that update [16:36] it's also jcastro [16:36] ;-) [16:36] lots of great stuff coming the desktop from both of those teams [16:36] we have an embarassment of richess concerning new desktop goodness ;) [16:37] moving on .. Riddell, Kubuntu update? [16:37] - alpha 2 done and in reasonable shape [16:37] - KDE SC 4.4 RC 2 coming today [16:37] - semantic desktop still blocked on virtuoso, kees promises soon [16:37] - kubuntu-notifications-helper now on the CD [16:37] that's all, short this week :) [16:37] ^ that fast? I only sent it to main today :) [16:38] pitti: well it's in the meta package and we've been testing it is more the point [16:38] right [16:38] so "now on the CD" is more specifically "now bound to be on the CD soon" [16:38] thanks Riddell [16:38] bryyce, x? [16:39] * Our signal rethrowing patch for apport seems to be the cause of recent xorg-server crashes. We've disabled it for now. [16:39] hi [16:39] did that significantly change in structure compared to karmic's? [16:39] :( [16:39] pitti, yeah [16:40] upstream moved all the signal handling around internally, so the patch had to be redone. but guess it is still not quite right [16:40] bryyce, is this part of the triaging/error recovery stuff? [16:40] rickspencer3, yeah, it is what catches crashes and collects backtraces [16:41] bryyce, I still get xorg display destroyed on new session [16:41] which itself was causing crashes [16:41] nice [16:41] bryyce, there are two questions for you on the wiki [16:41] * The diversions -> alternatives change was bumpy but seems to be working well now. [16:41] * tseliot nods [16:41] the first is straight form intel [16:41] Question from Intel: Lucid is still using 2.9.1. Intel already released 2.10 in 09Q4. Does Ubuntu plan to upgrade -intel 2.10 in Lucid [16:41] * Other than that, the usual bugs/patches/etc. [16:42] tseliot: congratulations for this, it was a huge change [16:42] yeah === Pici` is now known as Pici [16:42] :-) [16:42] okay, nice on that [16:42] rickspencer3, yes but the mesa/libdrm stuff had to go through first. 2.10 is on my todo list for a3 [16:42] bryyce, thanks [16:43] second one was concerning I guess 915 and lots of instability on 10v's after a suspend/resume [16:43] what's up with that? [16:43] powersave=0 should fix it [16:43] right, that is the powersave=0 thing [16:43] bryyce, right [16:44] well "fix" is a strong word tseliot [16:44] it seems upstream added this but it's quite buggy and all the distros are having to turn it off locally [16:44] s/fix/work around/ [16:44] ok [16:44] other than that, is intel solid? [16:44] graphics-wise anyway? [16:44] apw has that action [16:44] yes I think so [16:44] AFAIK Jesse Barnes partially reverted his changes [16:44] ok [16:45] bryyce, thanks for the update [16:45] tseliot, thanks too [16:45] is anybody else seing that xorg crash on new session? [16:45] bryyce + tseliot great job so far this release, lots of progress regarding graphics [16:45] it does it on user switch there on both my laptop and the mini10v [16:45] on the mini 10; rock solid on my laptop [16:45] lucky you ;-) [16:45] :-) [16:46] that's driving me crazy since it destroy my current session too [16:46] I get vt7 turned to a text vt [16:46] with Xorg still running [16:46] seb128, any error shown in your gdm log? [16:47] no, no error in neither Xorg or gdm logs [16:47] X is still running [16:47] let's talk after meeting if you want [16:47] seb128: text VT with cursor> I see that on shutdown sometimes [16:48] we'll have a mini 10v party at the sprint, and really get these puppies humming [16:48] I <3 my 10v [16:48] okay, moving on [16:49] ccheney, mozilla update? [16:49] finally got all of libsoup2.4 working except for the weird get_type code I haven't figured out yet, need to discuss with asac again to see what is going on with it [16:49] will also have OOo 3.2.0~rc2 uploaded today [16:50] assuming it doesn't keep failing to build, i think i worked all of its build errors out though [16:50] ccheney, were the build errors all KDE related? [16:50] actually not kde related for 3.2 it was due to all the changes from 3.1 to 3.2 in rules file [16:51] ah [16:51] its in the deb generation stage of the build now so i think it finally works [16:51] ok, in terms of mozilla ... [16:51] ite seems that you were hoping to have libsoup done two weeks ago, and then move on to epiphany [16:51] what is the status of the epihpany and such? [16:52] yea it kept growing bigger as i copied more things in and got some weird build errors a few times i had to get asac to help me investigate, now i have it all done except some code that has to change directly in glib and the get_type issue where it doesn't appear to actually have code [16:52] so still working on libsoup [16:53] there are a few gcc visibility settings in the code i am not sure if that is what is causing the weirdness [16:53] ccheney, do you feel that we are on track to accomplish this task in Lucid? [16:53] it should work pretty fast after i get the get_type parts working there are quite a few of them and they seem to all work the same way [16:53] ok [16:54] i feel that i don't understand glib/gtk/gnome well enough to copy the bits over and have work properly while also keeping OOo up to date [16:54] because the other libraries seem to be at least as involved [16:54] we have a contractor now that knows more about gnome, that could help with this? [16:54] ccheney, that is up to asac [16:54] ok [16:54] please discuss it with him [16:55] ok will do [16:55] if you feel the project is not on track, please ping me asap [16:55] ok [16:55] lets talk after meeting ccheney [16:55] so that we can steer the project as appropriate [16:55] (hi) [16:55] hi asac :) [16:55] ok, so ... [16:55] the glib modification i think should be done more familiar so i don't break all of hardy :) [16:55] pitti, any comments re release status? [16:55] er by someone [16:55] I did a quick update to https://wiki.ubuntu.com/DesktopTeam/ReleaseStatus earlier on [16:56] I haven't updated our RC bug lists, but it seems it's under control this cycle [16:56] good [16:56] my biggest worry for now is the boot speed stuff [16:56] I would like to discuss boot speed [16:56] I am quite worried as well [16:56] http://macaroni.ubuntu.com/~pitti/workitems/canonical-desktop-team-lucid-alpha-3.html didn't go down in the last 5 days, though [16:56] hmm weird [16:57] I think tseliot probably closed a few recently [16:57] I would have thought that with x stuff, we would have seen some [16:57] I mean, in practice, not on the whiteboard [16:57] ok, I'll follow up later [16:57] all: if you have completed work items, please note them as such on the white board [16:57] If we are above the trend line on Thursday, I will start postponing stuff! [16:57] bryyce: do you know whether there was discussion with the kernel team about the nouveau yay/nay? [16:58] pitti, I emailed them last week about this but haven't heard yet [16:58] ok, thanks [16:58] they've been sprinting though so assume they were busy with other matters [16:58] pitti, I will follow up on it this week [16:59] ok [16:59] so, boot speed [16:59] we've flatlined so far as I can see [16:59] so, seb128 and I discussed today some work this week to get us moving forward [17:00] I'm still waiting for today's image to appear on scott's charts [17:00] I re-enabled dailies yesterday and we have updated isos today [17:00] pitti, are you expecting a change? [17:00] yes [17:00] good [17:00] oh? [17:00] I fixed gdm/gpm to not start hal any more [17:00] which one? [17:00] and we got some new gnome components [17:00] oh, I though that was done previous week during the sprint [17:01] seb128: was after alpha-2 [17:01] Thursday evening [17:01] ah, ack [17:01] and we didn't have dailies since then [17:01] no daily since [17:01] anyway, I don't expect a dramatic change on the desktop itself [17:01] mvo: Nope upgrade manager -d isnt working at the moment for hardy [17:01] current une there is a bit over 17 seconds [17:01] we need to land the "start everything at once" bits and chrisccoulson's gconf dependency weakening === cypher__1 is now known as czajkowski [17:01] me was afk for a while there [17:02] I've been looking at start everything at once today [17:02] and I'm not convinced... [17:02] fagan, fyi ... in uor team meeting atm, will be done soon [17:02] and we need to see whether it's worth writing a new gconf backend [17:02] the charts with the change were slower than without [17:02] I need to look at that again [17:02] fagan: oh, its not - what was the error? [17:02] things start earlier [17:02] I looked into gconf for some 4 hours, with some experimental code [17:02] but fight each other [17:02] seb128 - i'm nervous about starting everything at once too [17:02] so it doesn't win us anything [17:02] but it can't be crowbar'ed, it needs an entire new backend written (with a faster storage than xml) [17:02] but, we can discuss after your meeting [17:02] mvo: It just doesnt offer lucid as an option for update-manger -d [17:03] Oh sorry rickspencer3 [17:03] hmmm [17:03] I'm concerned about spending so much efforts when we will get new techs next cycle which means dropping all that [17:03] well ... I feel we need more data [17:03] I mean for the gconf backend [17:03] and seb128 discussed this this morning [17:04] I will do some profiling on gnome-panel this week [17:04] pitti: I think someone (upstream) is already working on using a binary instead of xml for gconf [17:04] fagan: oh, I need to check why this is the case, I thought had enabled it [17:04] that doesn't solve the gconf issue though [17:04] thanks fagan! [17:04] chrisccoulson: we have to parallelize at least a bit; starting most bits 5 seconds into the session won't work [17:04] tseliot: dconf you mean? (with gvariant) [17:04] chrisccoulson, current it's 2 seconds in for me [17:04] ups [17:04] I'll continue to own gconf, and see to write a simple text based backend [17:04] pitti, currently it's 2 seconds there [17:05] pitti: it could be [17:05] but right [17:05] pitti - i agree, and there are parts we can start in parallel [17:05] but i think we need to be careful about making a small change to just start everything at once without understanding why some of the sequencing is there [17:06] ok [17:06] right, we have to do it early [17:07] I am having trouble seeing how this is all organized [17:07] like who owns what, etc... [17:07] rickspencer3: we have a work item list in https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-startup-speed [17:07] pitti, if I look to the work items, will that help me understand? [17:07] pitti updated the whiteboard some days ago during sprint [17:07] but nobody currently owns panel [17:07] rickspencer3, yes [17:07] basically I own "start everything early" [17:08] chrisccoulson's own the gconf delay bits [17:08] seb128, owns panel for now, i think [17:08] pitti owns the gconf speed issue [17:08] as I asked him to do some detailed profiling [17:08] and robert and I own gnome-panel [17:08] then we can pull out our secret weapon [17:08] one important change would be to start all applets at once [17:08] instead of spending a second for each single one [17:08] vuntz says it's supposed to do that [17:08] I will try to see why it's not working [17:08] right, I know; that's why I hope it's easy to fix [17:08] not looking forward working on bonobo ;-) [17:09] hehe [17:09] or on things using it [17:09] seb128, do we need to call out bonobo as a separate item, or is part of "panel"? [17:09] I'm a bit concerned about ressources we spend on bugs for a lts right now btw [17:10] not sure if meeting is the right time to raise that [17:10] like, too few? (right) [17:10] seb128, I understand your concern [17:10] seb128, FYI the OLS guys are working on a FAQ on the music service which includes information on the content available [17:10] but we this is important too [17:10] rickspencer3, let's have only gnome-panel for now [17:10] kenvandine, thanks [17:11] seb128, you mean bonobo work is part of gnome panel? [17:11] pitti, like I've no time to look at bug report or fix those [17:11] pitti, but I can by quick look see number of crashes and issues I don't want in a lts [17:11] pitti, and I fail to see when I will be able to work on those with the current workload [17:11] rickspencer3, yes [17:11] rickspencer3, they go together for that work I would say [17:12] we can still add some extra items after profiling work this week [17:12] seb128, ack [17:13] maybe I can spend some time on trying to fix important bugs? I would just need for pointers to see where to look at first :-) [17:13] I'll add a few to DesktopTeam/ReleaseStatus this week [17:13] didrocks, that or help on new version updates? [17:14] I'm doing most of those by myself this cycle [17:14] seb128: as you want. I think I can help too on updates, right [17:14] new GNOME again next week [17:14] ok, so the upshot is that didrocks will work on new versions this week so that seb128 can focus on gnome-panel [17:14] ok :-) [17:14] next week for new version [17:14] and otherwise, we will continue to work to plan defined on the blueprint [17:14] but that would be nice [17:15] it would let me some time to catch up on other things I planned to do this week [17:15] so I can look to gnome-panel this week [17:16] ok, so next meeting we should see some detailed profiling data from seb128 [17:16] :) [17:16] chouette! [17:16] ;-) [17:16] heh [17:16] any other business? [17:16] seb128: as regards the bug about new sessions, can you try gdm without patches 27 and 28 (both separately and together) please? [17:17] tseliot, will do [17:17] thanks [17:17] thank you [17:17] rickspencer3, no [17:17] ok, let's call that a wrap? [17:17] going ... [17:17] going ... [17:17] yes please ;-) [17:17] tseliot: 27 and 28? [17:17] * rickspencer3 taps gavel [17:17] thanks all [17:17] thanks rickspencer3 [17:18] thanks everyone [17:18] pitti: yes, they have to do with plymouth [17:18] thanks everyone ++ [17:18] tseliot: I mean, are these revisions somewhere? [17:18] thanks [17:19] pitti: sorry, I meant 27_save_root_window.patch and 28_plymouth_transition.patch in debian/patches (in gdm) [17:19] aah [17:19] yes, they're not bzr revisions [17:20] I'm offline for a bit to cd ~ and make some dinner [17:26] * seb128 kicks the wiki not responding === maxb_ is now known as maxb === MacSlow is now known as MacSlow|capoeira === gord_ is now known as gord [18:04] TheMuso`, Eastern Edition will have to be 1 hour late today [18:14] seb128 - the gnome-screensaver desktop file already has NoDisplay in it :) [18:14] no work to do there [18:14] hehe [18:15] pitti - i think we should add NoDisplay to the polkit-gnome-authentication-agent-1 desktop file too, to hide it from the session capplet [18:15] i've seen a few users disable that already, and then report a bug when everything that requires authentication breaks [18:22] chrisccoulson: sounds good [18:24] chrisccoulson: want me to do it? [18:24] pitti - yeah, if you've got time [18:24] if not, then i can do that later [18:24] hmmm, virtualbox is acting really weird :-/ [18:25] i've got 2 mouse cursors at different positions on the screen, tracking each other perfectly [18:25] and only 1 of them clicks on things [18:31] pitti, rickspencer3: so the switch session crash doesn't happen without splash... [18:31] ie when not using plymonth on boot [18:34] urgh, session switching is utterly broken here [18:34] it almost appears like the 2 VT's are overlayed [18:34] thats why i'm seeing 2 mouse cursors [18:34] i see bits of each desktop on the same screen [18:34] lucky you [18:34] when I try to open a new session there I get a text vt over my xorg [18:34] that is seriously strange... [18:35] so you're seeing similar behaviour? [18:35] I can see the mouse pointer on it though [18:35] I get a cursor over a text vt [18:35] with fsck log from boot usually [18:35] and switching to other vts doesn't make the xorg cursor hide [18:35] its a good job i'm not running lucid on my main desktop yet :) [18:36] if you only have one machine, how do you guys do active development on Lucid? virtualbox? [18:37] Nafai - i'm currently using virtualbox, although i think i'm going to invest in a second machine [18:37] virtualbox sucks for testing really [18:37] for desktop stuff I imagine so [18:37] Nafai, I do run lucid [18:37] sounds brave :) [18:37] i would run lucid if i were the only person using this machine [18:38] sounds like there are some cool things in the pipe for lucid [18:38] i tend to not run dev releases on my dev boxes since they break so often, generally just in a vm [18:38] ccheney - do you test openoffice in a vm? [18:38] yea [18:39] you must have some serious hardware ;) [18:39] the times i have tried running dev releases directly on hardware i spent more time trying to make it work than working on what i should be [18:39] heh, these new visual effects are quite funky [18:39] you only need ~ 2-4gb ram to run a vm [18:40] but yea my build box is fast [18:40] yeah, you'd need a good build environment for OO [18:40] core i7 3.6ghz (8 hardware thread) with 8gb ram [18:40] ccheney: nice! [18:40] i used to use a core 2 duo and it was a lot slower [18:41] seb128 - according to the output of ps, both my Xorg processes are in fact on the same VT [18:41] weird [18:41] only takes 1hr to do a OOo build (without l10n) [18:43] ccheney: so do you have pbuilders running on your monster box? [18:44] seb128 - http://paste.ubuntu.com/359153/ [18:44] funky :) [18:45] Nafai: i use a home grown chroot thing, easier to deal with since i have to modify debian dir so often, and pbuilder didn't exist when i first wrote the scripts [18:45] ah, makes sense [18:45] chrisccoulson: so for some reason the second xserver fails to start there [18:46] chrisccoulson: and I get a vt on top of xorg [18:46] all weird [18:46] chrisccoulson: but everything works fine if I don't use plymouth on boot... [18:46] seb128 - it is pretty weird. in my case, the second server starts, but on top of the VT i'm already using [18:46] i havent tried without plymouth [18:46] plymouth doesnt do anything for me anyway [18:46] just drop the splash argument in grub [18:47] ah, thanks. i'll try that in a bit [19:22] tedg, hey [19:22] tedg, I've reassigned bug #507975 to you [19:22] Launchpad bug 507975 in indicator-application "nautilus crashed with SIGSEGV in g_type_check_instance_cast()" [High,Triaged] https://launchpad.net/bugs/507975 [19:22] tedg, http://launchpadlibrarian.net/37927151/Stacktrace.txt [19:22] is the crash [19:22] I get the same with a small pygtk testcase [19:22] or with rhythmbox when closing it [19:23] that's in fallback case [19:23] ie when the user don't have the indicator applet [19:23] I guess many removed it in previous cycle [19:23] since it was pretty useless if you don't like to queue things there [19:23] which explain the number of duplicates [19:23] Hmm, okay. Maybe it's an issue with the callback. [19:23] let me know if you need details [19:23] Sorry, fallback. [19:23] I've a small testcase I can give you [19:24] Sweet. I'll add it to the test suite :) [19:26] tedg, http://people.canonical.com/~seb128/indicator.py [19:26] there is probably an easier way [19:26] but remove the message indicator [19:26] stop the message and application services [19:27] and run that [19:28] rickspencer3, can you take a note somewhere that we might want to make sure to add back the message indicators to config on upgrade? [19:29] rickspencer3, I expect many people found it useless and drop it but it's used for the applications too in lucid and upgrader will not know about that change [19:30] rickspencer3, or I'm not sure how you want to track such changes, work item, bug? [19:30] chrisccoulson: I filed bug 509798 as a reminder [19:30] Launchpad bug 509798 in policykit-1-gnome "set NoDisplay in .desktop file" [Low,In progress] https://launchpad.net/bugs/509798 [19:32] pitti, ^ [19:32] seb128: RC bug works fine for me, or a WI for final, as you prefer [19:33] ok thanks [19:34] tedg, let me know if you try the testcase [19:35] seb128: I was trying to get the test case dying under dbus-test-runner, but that wasn't working :-/ [19:35] I was just going to look at the code and see if the screwup was obvious real quick :) [19:36] Ugh, it was :( [19:41] good ;-) [19:41] tedg, you prefer things not easy to debug? ;-) [19:42] seb128: No I prefer that I didn't make stupid mistakes the first time. The saddest part is that I realized this mistake, wrote the function to fix it, and didn't change the signal handler. [19:42] seb128: If you want to distro patch it you can steal the diff off the merge request: https://code.launchpad.net/~ted/indicator-application/bug507975/+merge/17680 [19:44] * pitti slowly gets a theory of what makes this hald-addon-input thing die [19:44] hey TheMuso` [19:45] tedg, thanks [19:45] tedg, I will backport that now to stop having angry using opening nautilus bugs about copy crashing [19:46] seb128: Just tell users that they should do it right the first time, and not need more than one copy of a file :) [19:46] tedg, ;-) [19:46] * tedg hears that users love that. [19:46] tedg, they should not have removed the applet to start! [19:46] that will teach them ;-) [19:46] Heh, it's not a bug, it's a feature! [19:49] tedg, you need review for every commit? [19:49] seb128: Yes. [19:49] urg [19:49] I guess it helps to have good code but not to speed up things [19:50] Yes, it's also good to make sure that more than one person knows about the code. [19:50] We had a real issue with UNR when njpatel went on vacation last year :) [19:50] lol [19:50] I think we found the person that knew the most about the code was kwwii :) [19:50] urg [19:50] no comment ;-) [19:51] the code is made of svg icons right? ;-) [19:51] All of it. SVG and Javascript [19:51] We think it'll speed up boot time. You only need to load libxml and seamonkey. === pgraner` is now known as pgraner [20:42] seb128: just checking, gnome-system-monitor isn't in bzr, right? [20:43] pitti, it is [20:43] * pitti finds it then and will add Vcs-Bzr: [20:43] pitti, it's lp:~ubuntu-desktop/gnome-system-monitor/ubuntu/ [20:43] ah, got it [20:45] pitti, do you upload debbie's change? [20:45] pitti, sorry I forgot about this one, I had it on my todolist but we were in beta freeze... [20:45] seb128: yes; TBH I don't fully understand it, but I want to test it, and it's forwarded upstream [20:46] ok good [20:46] it doesn't seem complicated to me [20:46] the upstream bug has a better explanation [20:47] but I'd upload it as "harmless" [20:47] pitti, right [20:47] that's what I mean by doesn't seem complicated [20:47] it's very limited change [20:47] and I trust her for testing [21:13] have a good night [21:15] sleep well, didrocks [21:27] Hey djsiegel. [21:27] rickspencer3: np [21:27] hey TheMuso` [21:28] just wanted to get some info from you on the feasibility of using a webkit theme for empathy [21:28] and if you have any idea what gwibber is doing in this regard, as it also uses a lot of webkit [21:28] djsiegel: Really not sure where the version of webkit in Lucid stands WRT accessibility I'm affraid. [21:28] hmm [21:29] can you fill me in on the access. issues in general [21:29] ? [21:29] djsiegel: Let me try and get into lucid some time today, and I'll let you know. Lucid has been giving me a little grief. [21:29] sorry to hear [21:29] ok [21:31] cd [21:32] ergh, -EFOCUS [21:36] good night everyone === MacSlow|capoeira is now known as MacSlow [21:53] re [21:57] hi all. I wonder if anyone can give me a hand - I'm trying to update a yelp patch but get this error when running cdbs-edit-patch: http://paste.ubuntu.com/359242/ === cypher_ is now known as czajkowski === bjf is now known as bjf-afk [22:16] mdke - did you figure out the answer to your question>? [22:16] yelp uses quilt.... [22:17] chrisccoulson: not yet - but it sounds like you did... Guess I haven't done a yelp patch for some time :) [22:17] heh ;) [22:18] ok, I'll see if I can figure out how quilt works [22:18] thanks chrisccoulson [22:19] hey robert_ancell [22:19] you're welcome mdke [22:19] seb128 - i can't believe users are making such a big deal out of bug 501670 ;) [22:19] Launchpad bug 501670 in glib2.0 "g_set_prgname() called multiple times" [Low,Invalid] https://launchpad.net/bugs/501670 [22:20] chrisccoulson: me neither [22:20] seb128, hey [22:20] chrisccoulson: I think some got confused and think other crashes etc are due to it [22:20] robert_ancell, I'm going to take over gnome-panel profiling work this week [22:20] yeah, it seems that way [22:20] robert_ancell, did you have any work in progress on that? [22:21] robert_ancell, we said we would have some work done on profiling it for next week to show [22:21] * ajmitch waves to robert_ancell [22:21] seb128, no, nothing significant - it seems the big issue is loading everything in parallel [22:22] seb128, did you get my patch to log more info? [22:22] in series rather [22:23] robert_ancell, I think you did pastebin that before holidays but I'm not sure where it was [22:23] robert_ancell, do you still have it handy? can you email it to me? [22:23] * robert_ancell looking... [22:25] seb128, emailed [22:26] robert_ancell, thanks [22:26] robert_ancell, did you do any work on that since? [22:28] seb128, no [22:28] ok thanks [22:28] robert_ancell, I got launchpad-integration-cil to build and in main btw [22:29] robert_ancell, do you plan to work on some app patching for it? [22:29] we need a depends to keep it there ;-) [22:29] seb128, sweet, I'll patch f-spot, cheese, gbrainy [22:29] ok thanks [22:29] is there anything major I'm forgetting? [22:29] not using lpi? [22:29] yes [22:30] tomboy you can list too [22:30] banshee I guess [22:30] yes [22:30] The talk by Lucas Nussbaum about Debian QA was interesting - a lot of measuring the community by datamining [22:31] yeah, he gets quite some interesting stats [22:31] brb, dist-upgrade done, quick restart [22:35] re [22:35] robert_ancell, got your email thank you [22:42] hi, do you plan to integrate gnome shell for lucid+1 ? [22:42] hey feuloren, we will discuss that after lucid [22:42] we focus on lucid for now [22:43] seb128: ok ^^ I was thinking that the need for accelerated graphics was a big problem, with the live CD mainly, but ok [22:44] we do consider the lack for fallback for video cards which don't support acceleration an issue too [22:44] that's why we will need to discuss that after lucid [22:59] TheMuso`, can you give me another hours? [22:59] have a couple of things to knock out here, and need to update that wiki [23:00] robert_ancell, are you around for eastern edition today? [23:01] rickspencer3, is it now? [23:01] robert_ancell, well, it was supposed to be an hour ago [23:01] now I'm thinking one hour from now? [23:01] ok, I can do that, the network is a bit flaky but should be good enough [23:05] rickspencer3: Sure. [23:05] tx [23:09] So are there any small tasks that you guys have that I could use to get started in contributing to the Ubuntu desktop? I'm an experienced Python, C, C++ programmer, and have done some Debian packaging. === kermiac_ is now known as Kermiac [23:13] Nafai - if you want to do some packaging, there are currently plenty of updates to do for lucid: http://people.canonical.com/~seb128/versions.html [23:14] Nafai: there are lots of needs-packaging bugs in launchpad I think. you could probably poke at a few of them, or help getting packages updated for lucid, as chrisccoulson pointed at :) [23:15] and with 2.29.6 tarballs on the way, there'll be plenty to do over the next few days [23:15] oh, ok, 2.29.6 release is not until 27th [23:15] still a week away [23:17] chrisccoulson, right, tarballs next monday [23:17] seb128 - yeah, i just checked my calendar. i got a bit ahead of myself there ;) [23:17] with packaging, I could get away with just having a lucid chroot right? [23:17] (a little nervous upgrading my only machine to lucid just yet :)) [23:17] Nafai - i use a lucid pbuilder [23:17] ah, duh :) [23:25] be back after lunch [23:28] is anyone interested in porting transmission to the app indicator? [23:29] That sounds interesting [23:29] basically changing the use of a notification area icon to the new app-indicator api? [23:29] Nafai - are you familiar with the work which is happening in lucid? [23:29] yeah, that's right [23:30] just a little bit, read what jono has written about it [23:30] the transmission status icon looks relatively trivial to port, as it's quite a simple menu [23:30] chrisccoulson, Nafai porting apps to the app indicator is pretty simple [23:31] what is transmission written in? [23:31] jono - C [23:31] the status icon stuff in transmission is all contained in gtk/tr-icon.c in the transmission source [23:31] nafai, chrisccoulson, there is a C example on https://wiki.ubuntu.com/DesktopExperienceTeam/ApplicationIndicators [23:31] jono: thanks [23:32] its pretty much as simple as creating an appindicator object and putting the menus in there [23:32] check the Rhythmbox source code to see a good example [23:32] Is transmission on launchpad or should I just do apt-get source from a lucid install? [23:32] thats a chrisccoulson question :) [23:32] Nafai - the packaging is maintained in bzr (lp:~ubuntu-desktop/transmission/ubuntu) [23:33] cool [23:33] if you fetch that, and do "bzr bd-do" in the branch, it will download the tarball for you and put you in the full source directory [23:33] Nafai, also, if you have questions about writing the app indicator code, bratsche_ is a good person to ask and also #ayatana [23:33] awesome [23:33] jono: thanks [23:33] :) [23:33] excellent :) [23:33] sounds fun [23:33] thanks to your Nafai - ping me when you are done :-) [23:33] i need to do some app indicator porting too really [23:34] perhaps this will help me in applying for a job on this team, too :) [23:34] Nafai, it can't hurt ;-) [23:34] chrisccoulson, its pretty funky :) [23:35] jono - yeah, i'll probably do tracker and gnome-settings-daemon next week [23:35] chrisccoulson, that would be sweet :) === bratsche_ is now known as bratsche [23:35] now if I can only get the virtual box guest stuff working in my lucid vm [23:35] chrisccoulson, ping me when they are done :) [23:36] jono - will do. although, we're still discussing the packaging for the new version of tracker, so i don't know when we'll have that in lucid [23:36] hi jono [23:36] gotcha [23:36] hey rickspencer3 [23:36] hi chrisccoulson, Nafai [23:36] hey rickspencer3 [23:37] hi [23:47] chrisccoulson: ! awesome on the transmission patch. [23:48] chrisccoulson: I have a bug filed upstream and I've pinged the upstream a few weeks ago [23:48] jcastro - it's not done yet though ;) [23:48] chrisccoulson: that's ok, I can help you grease the skids when it's ready [23:48] cool, thanks :) [23:48] yay [23:54] chrisccoulson: do you have my email? [23:55] jcastro - did you send one recently? [23:55] chrisccoulson: no, I meant, do you have my email address so you can get a hold of me. :p [23:56] ah, yeah. i do :) [23:56] lol [23:56] finally got 3.2.0~rc2 uploaded and just saw debian uploaded 3.2.0~rc3, lol [23:57] well i'll get to see whether there are any new build bugs with rc2