[00:10] <YokoZar> brainstorm needs a "worst ideas" button
[00:14] <LaserJock> heh
[00:14] <pwnguin> if bash is any indicator
[00:14] <pwnguin> racist jokes
[00:15] <pwnguin> but i woudln't infer too much from worst ideas; the possible reasons someone might vote an idea down are wide and contradictory
[00:57]  * Hobbsee waves
[00:59] <LaserJock> hi
[01:26]  * Hobbsee dies, looking at excessive email
[01:26] <Hobbsee> i should actually read this...
[06:23] <pitti> Good morning
[06:26] <pitti> Hobbsee: tzdata> erk; we need to debug this; can you reproduce this? does it happen with install --reinstall the old version, too? (I didn't change any packaging)
[06:56] <dholbach> good morning
[06:57]  * pitti hugs dholbach
[06:57]  * RAOF never gets the hugs :`(
[06:58]  * dholbach hugs RAOF and pitti
[06:58] <RAOF> Yay!
[06:58]  * RAOF gives hugs.
[06:58] <pwnguin> so when is ff3 supposed to go final?
[06:58] <dholbach> afaik the theme was supposed to have changed - it somehow didn't change for me yet
[06:58] <dholbach> pwnguin: best to ask asac that
[06:58] <pwnguin> the theme looks like it changed a bit
[06:59] <pwnguin> the icons for left right and the like
[06:59] <pwnguin> but more importantly, google browser sync isn't published yet for ff3
[07:03] <dholbach> hrm..... seems that alsa is broken for me with 2.6.24-12
[07:04] <pwnguin> maybe it's time to investigate mozilla weave
[07:27] <pitti> doko_: I guess tzdata's javazic-fixup.patch should actually be applied? I'll add it to debian/patches/series
[07:28] <doko_> pitti: it is applied during the build
[07:29] <pitti> doko_: oh, argh, just see it in debian/rules
[07:32] <Ng> dholbach: bug 200338 :)
[07:32] <pitti> doko: do you plan to introduce the javazic binary to some proper package in Debian?
[07:33] <pitti> doko: (I wonder about forwarding this patch to Debian, but the sharutils .uu one is really nasty)
[07:34] <doko> pitti: Aurelian already knows. once we have OpenJDK in main, that can go away
[07:34] <pitti> ah, nice
[07:34] <dholbach> thank Ng
[07:37] <Ng> dholbach: it's quite impressive, the bug is 10 hours old and has 40 comments and 12 duplicates already ;)
[07:42] <dholbach> Ng: yeah it is - I guess bdmurray can come up with even more impressive bug reports :)
[07:48] <ubotu> Launchpad bug 200338 in linux-ubuntu-modules-2.6.24 "no sound hardy kernel 2.6.24-12 " [Critical,Confirmed] https://launchpad.net/bugs/200338
[07:52] <slangasek> oh, is *that* why sound wasn't working?  I thought it was just because I tried out -12 before lum was available
[07:57] <mdke> morning all
[07:57] <mdke> is about-this-computer going to be implemented by hardy, does anyone know?
[07:58] <mdke> or has it missed feature freeze?
[08:23] <dholbach> seb128: did you already see the bug-o-meter on  http://daniel.holba.ch/really-fix-it/ ?
[08:24] <seb128> dholbach: ah, nice ;-)
[08:25] <pitti> hi seb128
[08:26] <seb128> hey pitti
[08:26] <seb128> pitti: so, I've been looking at login speed a bit again and doing some stracing and graphs about the issue
[08:27] <pitti> bdmurray: do you plan a new p-lp-bugs upload soon?
[08:27] <seb128> pitti: jockey-gtk --check takes 11 seconds there! and it does that at every boot where I don't need it
[08:27] <crevette> good morning
[08:27] <pitti> bdmurray: once we have that, we should do an SRU for that, to fix the retracers, and scripts people run on gutsy
[08:27] <seb128> pitti: is there a way to cache the value? it seems to call a lot apt-cache and modinfo and that's sloooow
[08:27] <pitti> seb128: ah, good point
[08:28] <seb128> no cookie for mvo neither
[08:28] <dholbach> hey thekorn
[08:28] <pitti> seb128: I'll think about it
[08:28] <seb128> update-notifier call apt-check which takes 7 seconds on my laptop there
[08:28] <pitti> it should at least run in parallel with other stuff
[08:28] <pitti> seb128: caching should be no problem
[08:28] <seb128> pitti: it does, but that and update manager give a 5 seconds impact on a 25 seconds login time there
[08:29] <seb128> it adds to the load
[08:29] <pitti> 25 seconds? lucky you
[08:29] <pitti> seb128: right, understood; thanks for pointing out
[08:29] <seb128> pitti: could you do a graph of your laptop login?
[08:29]  * dholbach hugs super-seb128
[08:29] <pitti> seb128: how do I get a graph of session start?
[08:29] <thekorn> morning dholbach et al.
[08:29] <pitti> hi thekorn, how are you?
[08:30] <thekorn> pitti, I'm fine thanks
[08:31] <seb128> pitti: start a debug xorg session, run "strace -ttt -f -o login.log gnome-session", wait until having the desktop loaded, ctrl-C it, download http://people.ubuntu.com/~seb128/plot-timeline.py, run plot-timeline.py login.log -o login.png, copy the png somewhere
[08:33] <pitti> seb128: debug xorg session -> how?
[08:33] <pitti> just X from a VT?
[08:33] <pitti> and DISPLAY=:0 gnome-session from another VT?
[08:34] <seb128> pitti: it's called xterm session in the gdm login menu or something similar
[08:34] <pitti> ah, that
[08:34] <seb128> just any way you want which doesn't run gnome-session for you work
[08:34] <seb128> since you need to run it under strace
[08:36]  * seb128 hugs dholbach too
[08:37] <seb128> dholbach: do you have a box where login is slow? want to do a graph on it too? ;-)
[08:39] <pitti> hm, ctrl-c doesn't work, ctrl+alt+backspace does it :)
[08:39] <seb128> pitti: yeah, did the same yesterday, not sure why it doesn't work
[08:39] <pitti> wow
[08:39] <pitti> login.log -> 147 MB
[08:40] <seb128> strace is verbose ;-)
[08:40] <seb128> but that also means we do a lot of access at login
[08:41] <dholbach> pitti: my login.log is "just" 77,1MB
[08:42] <pitti> hm, I don't even have unusual stuff
[08:42] <dholbach> seb128: http://daniel.holba.ch/temp/login.png
[08:42] <pitti> to the contrary, I ꫭ
[08:42] <pitti> already disabled tracker
[08:42]  * pitti beats scim really hard
[08:43] <pitti> grrrrr
[08:43] <pitti> python /tmp/plot-timeline.py /tmp/login.log
[08:43] <seb128> pitti:  -o login.png
[08:43] <seb128> dholbach: thanks
[08:44] <dholbach> de rien
[08:44] <pitti> seb128: ^ that produces a 1400x34200 pixel graphics whihc is entirely black
[08:44] <pitti> python /tmp/plot-timeline.py -o /tmp/desktop-login.png /tmp/login.log
[08:44] <pitti> right, I used that
[08:44] <seb128> pitti: hum, weird
[08:44] <seb128> pitti: it should do something similar to http://daniel.holba.ch/temp/login.png
[08:45] <pitti> seb128: hm, I ran it on my desktop; let me re-run it on the laptop itself
[08:46] <seb128> dholbach: if you want to do some extra testing you can disable jockey-gtk and update-notifier, reboot and try again to see the difference
[08:46] <seb128> hey mvo
[08:47] <mvo> hey seb128
[08:47] <seb128> mvo: so I've been looking at the login speed, update-notifier runs apt-check on login which takes 7 seconds on my laptop, what can you do to make things better there? ;-)
[08:48] <mvo> seb128: woah, that is a long time. sure, we can delay the run of apt-check. could gnome-session not run some stuff in parallel? is u-n blocking other stuff from starting?
[08:48] <pitti> seb128: same result :/ png in eog is black, and firefox crashes when opening it
[08:48] <seb128> pitti: :-(
[08:48] <mvo> seb128: with what app do you measure the login speed btw?
[08:48] <dholbach> pitti: you have python-cairo installed?
[08:49] <seb128> mvo: pitti: start a debug xorg session, run "strace -ttt -f -o login.log gnome-session", wait until having the desktop loaded, ctrl-C it, download http://people.ubuntu.com/~seb128/plot-timeline.py, run plot-timeline.py login.log -o login.png, copy the png somewhere
[08:49] <pitti> seb128: cairo> installed
[08:49] <seb128> mvo: that gives graphs similar to http://daniel.holba.ch/temp/login.png
[08:49] <seb128> mvo: it's ran it parallele but those add to the load and that doesn't make things better
[08:50] <pitti> seb128: shouldn't an strace -ttt -e execve -o login.log give pretty much the same info?
[08:50] <pitti> (and -f)
[08:50] <seb128> pitti: it should, I didn't try though
[08:50]  * pitti tries
[08:50] <seb128> pitti: not sure why the graph things doesn't work for you
[08:52] <dholbach> arg.... gnome-session-save.... arg!!!!
[08:52] <funman> hi
[08:52] <dholbach> seb128: hum... I disabled jockey and update-notifier but it seems that update-notifier still runs
[08:53] <seb128> dholbach: maybe try to move the desktop away from /etc/xdg/autostart, I tried this way yesterday
[08:53] <dholbach> at least in the "start programs" thing
[08:53] <dholbach> ngh
[08:53] <dholbach> ok
[08:53] <seb128> but the GUI should work
[08:53] <seb128> I'll look at that later
[08:53] <dholbach> it's unchecked, but still it ran
[08:54] <funman> i notice a strange bug on a live cd session
[08:54] <funman> i'd like to get opinions
[08:54]  * dholbach restarts and tries again
[08:54] <funman> getcwd() returns NULL / ENOENT in cvs
[08:55] <pitti> funman: cvs, as in the revision control?
[08:55] <funman> but surrounding the code calling it with the same call to getcwd() (both before and after) shows that getcwd() works
[08:55] <funman> yes
[08:55] <funman> i suspected squashfs, but it fails also on ext3
[08:56] <pitti> seb128: http://people.ubuntu.com/~pitti/tmp/login-execve.log
[08:57] <seb128> pitti: still no luck with the graph?
[08:57] <pitti> seb128: ah, so the script takes the time between execve and SIGCHLD?
[08:57] <pitti> clever
[08:57] <pitti> seb128: trying with that reduced one
[08:57] <seb128> waouh
[08:58] <seb128> that's slow login
[08:58] <seb128> 85 seconds!
[08:58] <pitti> yeah, as I said
[08:58] <dholbach> seb128: http://daniel.holba.ch/temp/login2.png (without jockey and update-notifier)
[08:58] <pwnguin> maybe upgrade manager should sleep for five minutes on startup?
[08:58] <pitti> seb128: ah, png works
[08:59] <pitti> seb128: http://people.ubuntu.com/~pitti/tmp/login.png
[08:59] <seb128> dholbach: weird, they are still ran apparently there, you have lot of apt-cache and modinfo and dpkg calls
[08:59] <dholbach> no idea why *shrug*
[09:00] <seb128> pwnguin: 60 seconds would be enough
[09:00] <dholbach> I moved away the files and unchecked them in the gnome-session UI
[09:00] <dholbach> pitti seems to start a lot of applets and terminals and stuff too
[09:00] <dholbach> firefoxes too
[09:00] <pitti> I have two terminals
[09:00] <pitti> no firefox
[09:01] <pitti> and a terminal with mutt inside
[09:01] <dholbach> it says that on the .png file
[09:01] <pitti> applets> just deskbar, the rest is standard
[09:01] <dholbach> multiload?
[09:01] <pitti> dholbach: hm, weird; I might have a starter for it, but ffox doesn't start
[09:01] <pitti> right, I use multiload, too
[09:02] <dholbach> cpufreq-applet
[09:02] <seb128> pitti: you have a gap of almost 15 seconds after the cpuload applet load
[09:02] <seb128> cpufreq I mean
[09:02] <pitti> indeed
[09:02] <pitti> whoa
[09:02] <seb128> but it's not clear the gap is due to it, since things are ran in a async way you can't say which one is still running there
[09:02] <dholbach> cpufreq seems to call modinfo later on too
[09:02] <dholbach> (or might be)
[09:02] <seb128> another way to get a boot graph is too install bootchart
[09:03] <seb128> dholbach: no, that's jockey, things are async
[09:03] <dholbach> hmok
[09:03] <seb128> apt-cache and modinfo are jockey
[09:03] <seb128> apt-check and dpkg-query are update-notifier
[09:03] <lool> pitti: Heya; dunno who accepted elisa and pigment from binary new, but now there's likely a problem: we need pigment-python, and elisa-plugins-{bad,ugly,good} (all new sources)
[09:03] <seb128> I see pitti's disk is slow on those take a while
[09:03] <lool> pitti: Or should I be talking to whoever is on archive admin duty today (I guess slangasek)?
[09:04] <funman> http://pastebin.ca/936294
[09:04] <seb128> pitti: another way to get informations is apt-get install bootchart, mv /etc/init.d/stop-bootchart somewhere and run sudo ./stop-bootchart start one logged
[09:04] <pitti> lool: those were split out?
[09:04] <seb128> pitti: this way you get a bootchart until desktop loading
[09:04] <pitti> seb128: ah, nice
[09:04] <funman> i would say the bug is in glibc, but i can not read that code
[09:04] <lool> pitti: pigment-python yes, elisa-plugins-* not really
[09:05] <pitti> seb128: I'll enable autologin for that then?
[09:05] <seb128> pitti: that gives better informations on what takes how long and do lot if io
[09:05] <lool> pitti: elisa gained a plugin architecture and the main UI is in one of the packages (-bad)
[09:05] <seb128> pitti: yes
[09:05] <lool> (don't laugh)
[09:05] <Tonio_> hi there
[09:05] <pitti> lol
[09:05] <Tonio_> seb128: did you have to change something to network-manager-gnome recently ?
[09:06] <seb128> Tonio_: asac updated to a new version if that's the question, he maintains it, as an user no I didn't, why?
[09:06] <lool> it's inspired by the gstreamer package names, except you're forced to install -bad -- unless you want to run the elisa server though
[09:06] <Tonio_> seb128: knetworkmanager broke with latest updates. It fails to initiate wireless connection. nm-applet works
[09:06] <seb128> Tonio_: nm-applet has been updated to a new version too
[09:06] <seb128> I guess upstream did whatever is required to have those in sync and working
[09:07] <lool> pitti: If you work on accepting the new sources, then if you like we can discuss main promotion as well (if you need to discuss anything)
[09:07] <Tonio_> seb128: okay then I suspect I should investigate this with knetworkmanager upstream
[09:07] <Tonio_> seb128: libnl is probably the cause somehow
[09:07] <pitti> lool: I'll accept them to universe first anyway
[09:07] <lool> pitti: elisa at least starts for me in the new version now on Intel; didn't try with compiz running though
[09:07] <pitti> E: elisa-plugins-bad: not found
[09:07] <pitti> oh, experimental?
[09:07] <lool> pitti: It's uploaded already
[09:07] <lool> pitti: It should be in NEW
[09:08] <pitti> lool: ah, I thought I should sync it
[09:08] <lool> I uploaded them manually because the plugins-* are still in Debian NEW
[09:08] <lool> Basically the Debian ftpmasters took the same shortcut as Ubuntu's
[09:08] <pwnguin> maybe if you optimized your PATH ;)
[09:08] <pitti> lool: and python-pigment?
[09:08] <lool> pitti: Also
[09:09] <asac> Tonio_: for me knetworkmanager works
[09:09] <lool> In fact it used to replace python-pigment, but now it doesn't anymore so it's a new source with new binaries
[09:09] <lool> (python-pgm is the new name
[09:09] <lool> )
[09:09] <asac> Tonio_: though i cannot rule out that it needs an update because of new networkmanager 0.6.6
[09:10] <lool> pitti: New: pigment-python 0.3.3-1~hardy1 elisa-plugins-good 0.3.4-1~hardy1 elisa-plugins-ugly 0.3.4-1~hardy1 elisa-plugins-bad 0.3.4-1~hardy1
[09:10] <lool> It's a long chain of build-deps on build-deps now    :-/
[09:11] <lool> (elisa-plugins-* need python-elisa from elisa which needs python-pgm from pigment-python which needs libpigment from pigment)
[09:11] <lool> And naturally the lib changes SONAME frequently   :-/
[09:12] <Tonio_> asac: it works ? no issue connecting to any wireless network ?
[09:12] <Tonio_> asac: I can't initiate any connection, but it works when used together with nm-applet
[09:12] <Tonio_> asac: I'll ask other people, maybe that's just a local issue
[09:13] <asac> Tonio_: strange ... haven't tested today, but tested before and after the upload
[09:14] <Tonio_> asac: hum, do you have nm-applet running when you use it ?
[09:14] <Tonio_> asac: if yes, you won't see the issue since they'll start the connection together, and nm-applet finally succeeds
[09:14] <asac> Tonio_: not when i tested this
[09:14] <asac> Tonio_: yep i know. i will test  in a few
[09:14] <Tonio_> asac: hum........ okay I'll keep investigating then
[09:15] <asac> (to be sure)
[09:15] <asac> i think i didn't test with 0.6.6 ... just with 0.6.6~rc2
[09:15] <asac> but there where no significant changes
[09:15] <Tonio_> asac: looked to me more a problem with libnl than with network-manager itself
[09:15] <asac> Tonio_: point is that nm-applet 0.6.5 still works as well
[09:15] <asac> thats why i doubt that knetworkmanager is broken
[09:15] <Tonio_> hum yep, well the codebase is completly different
[09:16] <\sh> hmmm...how stole my sound?
[09:16] <Tonio_> maybe the code is simply broken with libnl1 or something....
[09:16] <Tonio_> \sh: general issue, looks like
[09:16] <asac> Tonio_: yes, libnl is a good idea to look at
[09:16] <asac> Tonio_: though i am not even sure why knetworkmanager needs libnl
[09:16] <asac> imo it should just talk to network-manager
[09:17] <\sh> Tonio_: cool :) I was wondering if my company workstation just lost its sound ability
[09:17] <Tonio_> asac: as I said, just ideas thrown to the air ;) I need to investigate more
[09:17] <Tonio_> also fonts changed recently no ?
[09:17] <Tonio_> "sans" of size 8 just looks strange....
[09:19] <asac> ArneGoetje: hey, did the thai font update ever happen? (just looking at not finished mail threads and found the one from thep i forwarded)
[09:19] <lool> pitti: Oh and you might have to wipe your .elisa to get the default set of plugins of the new version enable automatically -- if you already ran an older elisa version before that is
[09:22] <funman> ok .. cvs uses some rpl_getcwd() instead of getcwd()
[09:24] <funman> some autotools magic *sigh*
[09:33] <dholbach> \sh: <Ng> dholbach: bug 200338 :)
[09:35] <ubotu> Launchpad bug 200338 in linux-ubuntu-modules-2.6.24 "no sound hardy kernel 2.6.24-12 " [Critical,Confirmed] https://launchpad.net/bugs/200338
[09:38] <pitti> ArneGoetje: WDYT about assinging the scim switch key to a combination that's less likely to be hit accidentally?
[09:39] <pitti> ArneGoetje: maybe just ctrl+space or left+right ctrl, instead of shift+space?
[09:39] <dholbach> I use ctrl-space in liferea :-/
[09:40] <Mithrandir> pitti: I think we should use ctrl-alt-shift-f7
[09:41] <pitti> well, I guess it should still be *easy* to do, but deliberately
[09:41] <pwnguin> my keyboard doesn't have a right control ^_^
[09:42] <pitti> let's map it to 's' then
[09:42] <pitti> no uer need uch a letter
[09:42] <pwnguin> no, c
[09:42] <pwnguin> everything kan be spelled without it
[09:42] <pitti> pwnguin: but only for KDE users!
[09:43] <Mithrandir> pwnguin: at least in sensible languages, like Norwegian.
[09:43] <pwnguin> how bout tilda
[09:43] <pwnguin> nobody really uses it
[09:43] <pitti> ctlr+esc?
[09:43] <pitti> pwnguin: tilde == home directory
[09:44] <pwnguin> pitti: where/
[09:44] <Mithrandir> and aptitude search patterns.
[09:44] <Chipzz> pwnguin: that was one hell of a clueless remark :P
[09:45] <pwnguin> heh
[09:45] <Chipzz> not to mention that tilde is often used in urls
[09:45] <Chipzz> for user-homedirs
[09:46] <pwnguin> get rid of the ugly reset X key combo and use that ;)
[09:46] <pwnguin> it'll be hilarious trying to convince people that's really the combination
[09:47] <Chipzz> oh wait
[09:47]  * Chipzz slaps self for missing sarcasm :P
[09:48] <Chipzz> hrrrrm
[09:48] <Chipzz> what about caps-lock with some key? ctrl-capslock maybe?
[09:48] <Chipzz> then finally that bloody key will have an actual use :P
[09:49] <pwnguin> contrl alt \?
[09:49]  * Chipzz is actually surprised that keyboard manufacturers still put caps lock on keyboards these days
[09:49] <pwnguin> it's only gonna get worse as compiz gets more popular
[09:49] <pitti> Chipzz: it's a great Esc key for me :)
[09:49] <pwnguin> Chipzz: the thing is, everyone wants everyone ELSE'S capslock gone
[09:50] <Chipzz> pitti: yea, but you're hardly a normal user ;)
[09:50] <Chipzz> pwnguin: even when I do type something all caps, I still use shift
[09:50] <pwnguin> plus, i think unreal tournament maps capslock to scoreboard
[09:51] <Chipzz> yeah but I mentioned ctrl-caps lock
[09:51] <TheMuso> 5/c
[09:51] <mvo> pitti: is scim still enabled for you? the latest upload should have disabled it again for non CJK languages
[09:51] <Mithrandir> I don't have a caps lock
[09:51] <pitti> mvo: yes, it has never been disabled again for me
[09:51] <Chipzz> Mithrandir: hh kayb? :)
[09:51] <lool> Mithrandir: Put a brick on the shift key; the OS should cope
[09:51] <pitti> mvo: although I tried to disable it in the configuration dialo
[09:51] <pitti> g
[09:51] <Mithrandir> Chipzz: no, compose key.
[09:51] <pitti> mvo: but I didn't fiddle with it on my laptop, and it's still enabled there, too
[09:52] <Chipzz> Mithrandir: ah, sun (UNIX) keyboard?
[09:52] <pitti> mvo: maybe I need to reset some gconf keys or so?
[09:52] <mvo> pitti: I guess you did logged out/in again since the package was updated?
[09:52] <Mithrandir> Chipzz: no, thinkpad.
[09:53] <pitti> mvo: yes, often
[09:53] <mvo> hrm, not good
[09:54] <mvo> pitti: as a workaround (for now) you can run gnome-language-selector and disable it there in the checkbox
[09:54] <mvo> pitti: that is a good test for the code anyway :)
[09:54] <Chipzz> anyway, I think caps lock (with or without modifier) makes sense; it's used to "alter" the meaning of keys you press now anyway
[09:55] <pitti> mvo: btw, it should move from Administration to Preferences
[09:55] <pitti> mvo: wow, a reboot?
[09:55] <mvo> pitti: not really, a relogin is enough
[09:55] <pitti> mvo: not just logout/login
[09:55] <pitti> martin   28630  0.0  0.0      0     0 ?        Z    09:24   0:00 [scim] <defunct>
[09:55] <pitti> mvo: btw, I have had this broken process forever
[09:55] <pitti> what's the deal with it?
[09:56] <pitti> (since day 1 when we enabled it)
[09:56] <mvo> pitti: but strangely for new gdm defaults we need to reboot, it used to work to send it a reload, but that seems to be no longer the case
[09:56]  * pitti reboots, brb
[09:56] <mvo> pitti: no idea about this one, that must be something in the im-swich package
[09:57] <mvo> pitti: move to preferences because it does not require gksu anymore to start? what is our policy here, langpack install still requires gksu, so do we consider it something for preferences or more something for admin?
[09:57] <seb128> mvo: gksu = admin
[09:58] <mvo> seb128: its a bit unfortunate that it currently mixes system-wide stuff (langpacks) with user stuff (im-switch config)
[10:00] <pitti> mvo: right, but as long as normal users are supposed to use it, it shouldn't have X-KDE-SubstituteUID=true at least
[10:00] <pitti> since otherwise non-admins won't see it at all
[10:00] <pitti> yay sanity!
[10:00] <pitti> mvo: disabling scim worked
[10:01] <mvo> pitti: nice, when ArneGoetje is back, we should look at the root cause of the problem
[10:04] <mvo> doko: your python-central fix works, dapper->hardy lts upgrades with ubuntu-desktop work again :)
[10:07] <dholbach> pitti: can you please reject the five-a-day upload to Ubuntu I just did?
[10:08] <pitti> dholbach: is it NEW?
[10:08] <dholbach> yes, it should be
[10:09] <pitti> argh, my ssh hates me again
[10:09]  * dholbach gives pitti a big hug
[10:09] <pitti> dholbach: will do as soon as ssh works again or seb128 beats me to it
[10:09] <dholbach> thanks a lot, pitti and seb128
[10:11] <seb128> dholbach, pitti: done
[10:11] <pitti> merci
[10:11] <mvo> pitti: I would like to upload cupsys with a fix for #156634 (test if /var/run/cups is not available and if not, create it before the chown call. ok with you?
[10:11] <seb128> de rien
[10:11] <seb128> dholbach: five-a-day 0.17, that was it, right?
[10:11] <pitti> mvo: sure; please send the debdiff for me for committing into svn
[10:11] <dholbach> seb128: exactly
[10:11] <dholbach> seb128: I should have directly uploaded to 5-a-day-ppa instead :/
[10:11] <pitti> mvo: although, nevermind; MoM will mail it to me automatically
[10:11] <seb128> ok, upload rejected
[10:11] <dholbach> thanks :)
[10:12] <seb128> dholbach: ok ;-)
[10:12] <seb128> you are welcome
[10:27] <mvo> pitti: http://paste.ubuntu.com/5517/
[10:28] <pitti> mvo: hold on
[10:28] <pitti> mvo: you forgot a || true
[10:28]  * mvo holds
[10:28] <pitti> or a proper if..then
[10:29] <mvo> meh, right
does not exists -> does not exist</nitpick>
[10:30] <pitti> mvo: I wonder whether it should just not be chown'ed in that case, though
[10:30] <pitti> but it won't harm
[10:34] <mvo_> pitti: http://paste.ubuntu.com/5518 would be the alternative for unnedded chmoding, whatever style you like better
[10:35] <pitti> mvo_: I think I like the latter one
[10:35]  * pitti hugs mvo, thanks
[10:35] <mvo_> pitti: great, thanks for looking over it
[10:35]  * mvo_ hugs pitti back
[10:36] <mvo_> pitti: this is something that people asked for a SRU for feisty->gutsy too, how do you feel about it?
[10:37] <pitti> mvo_: if that's still relevant, ok for me
[10:38] <doko> mvo: good to know
[10:54] <\sh> guys, is lzma now a valid option for creating deb files?
[10:55] <pitti> \sh: yes
[10:55] <\sh> pitti: but it's not default, right?
[10:56] <pitti> \sh: right, you need to use dh_builddeb -- -Zlzma
[10:56] <\sh> pitti: cool :)
[10:58] <ogra_cmpc> grmbl ... gvfs renders my classmate unusable while developing ....
[10:59] <pitti> \sh: oh, and Pre-Depends: dpkg (>= 1.14.12ubuntu3)
[10:59] <ogra_cmpc> no, dear nautilus, you dont need to open a godammden 15M eating window for *every* mount while i'm working
[11:00] <\sh> pitti: so in all binary packages we need to add Pre-Depends: ?
[11:00] <ogra_cmpc> i wonder if the gvfs guys ever worked on systems where you only have 64M freely uable ram
[11:00] <pitti> \sh: in all that use lzma, yes
[11:00] <pitti> \sh: soyuz will reject built binaries which don't have it
[11:00] <\sh> pitti: ok..so it's not a good solution when we want to backport pkgs...
[11:00] <pitti> \sh: yes, unless you add some runtime tests to debian/rules and use some substvars magic
[11:01] <\sh> pitti: well, I wonder how much reduction we have when using lzma...are there any stats on it?
[11:01] <pitti> \sh: ia32-libs shrunk by more than 50% (41 MB -> 20 MB)
[11:02] <pitti> \sh: the entire alternate CD shrinks in the order of 200 MB
[11:02] <\sh> pitti: wow...
[11:02] <pitti> but it entirely depends on the particular package, of course
[11:02] <\sh> pitti: that sounds promising
[11:02] <pitti> \sh: wow indeed :)
[11:04] <YokoZar> pitti: Will we then default to lzma for Hardy+1 now that we know we have it?
[11:04] <YokoZar> I mean, no more need for pre-depends for Hardy+1
[11:04] <pitti> YokoZar: hopefully
[11:05] <pitti> we'll still need the Pre-Depends
[11:05] <pitti> at least until dapper goes out of support
[11:05] <pitti> (for backports, etc.)
[11:05] <pitti> but we should find a way to add it automatically
[11:05] <pitti> (pkgbinarymangler FTW *cough*)
[11:05] <YokoZar> pitti: Ahh, ok.  But we won't expect people to upgrade Dapper->Hardy+1
[11:05] <pitti> right
[11:05] <pitti> but we might do source backports from intrepid to pre-hardy
[11:06] <YokoZar> Would we also move lzma out of universe?
[11:06] <pitti> we did, it's in main of course
[11:07] <YokoZar> Err, I meant for Gutsy.  Or does that not make sense since backports can depend on universe packages anyway
[11:10] <pitti> YokoZar: we won't build lzma packages for pre-hardy
[11:15] <Riddell> hmm, no calc, anyone know what's happening with openoffice.org-l10n-en-us ?
[11:36] <TheMuso> pitti: Has lzma for live CDs been considered for future releases? Or is there too much of a performance hit?
[11:37] <ogra_cmpc> TheMuso, that would mean we'd have to patche the kernel in a way upsytream doesnt like
[11:38] <ryanhaig1> im trying to rebuild nautilus without tracker integration on gutsy, i have downloaded the source but I dont know what to change to disable tracker, is it a command line option when using debuild or do i need to change a file?
[11:38] <TheMuso> ogra_cmpc: oh ok.
[11:38] <ogra_cmpc> and lzma has a immensely higher ressource hunger so having a whole squashfs in lzma will be a resource hog
[11:39] <pitti> TheMuso: AFAIK there are working plans for lzma squashfs
[11:39] <ogra_cmpc> you trade performance against comression rate
[11:39] <pitti> TheMuso: no idea about its status, though
[11:39] <ryanhaig1> is this not the right channgel to ask the question?
[11:41] <MacSlow> Greetings everybody!
[11:43] <pitti> hi MacSlow
[11:43] <MacSlow> hey pitti
[11:46] <ogra_cmpc> pitti, do i miss any kind pof approval for the classmate-tools package ? seems it built fine but the publisher apparently doesnt copy the binary to a.u.c
[11:47] <pitti> ogra_cmpc: it's in binary NEW (see https://edge.launchpad.net/ubuntu/+source/classmate-tools/0.2-0ubuntu1)
[11:47] <Hobbsee> pitti: i'd changed /etc/timezone myself, to a valid config, iirc, and it barfed.  should be reasonably easy to reproduce
[11:47]  * Hobbsee found a better way to do it, but still
[11:48] <pitti> Hobbsee: with dpkg-reconfigure tzdata or tzconfig, or manually?
[11:49] <pitti> Hobbsee: i. e. maybe /etc/timezone and /etc/localtime were out of sync for you?
[11:49] <Hobbsee> pitti: i think i'd tried tzconfig, but it appaered to do nothing.  they would have been out of sync, yes.
[11:50] <Hobbsee> dholbach: any chance you can add thta you have to join the 5-a-day LP team to the instructions if you haven't already done so?
[11:51] <dholbach> Hobbsee: it was there all the time
[11:51] <Hobbsee> dholbach: oh, so multiple people have missed it?  at least 3?
[11:51] <Hobbsee> (another one was asking about it in #lp today)
[11:51] <dholbach> doko and you are all I know about, 110+ made in their successfully :)
[11:51] <dholbach> right
[11:52] <dholbach> I can make it BOLD if you like
[11:52] <Hobbsee> right.  apparently i'm just dumb then.
[11:52] <dholbach> that's not what I said - one point on the list can be missed easily
[11:52] <dholbach> I'll make it bold - safe is safe :)
[11:52]  * Hobbsee looks at the page again
[11:53] <doko> slomo_: will you sync gstreamer + plugins again for hardy?
[11:53] <Hobbsee> oh
[11:53] <Hobbsee> dholbach: right, i'm dumb.
[11:53] <dholbach> no worries :)
[11:53]  * dholbach goes out for lunch - see you guys later
[11:53] <Hobbsee> dholbach: i suspect that everyone' slooking at the steps that actually have a code box underneath them
[11:54] <dholbach> Hobbsee: yeah, that makes sense :)
[11:54] <Hobbsee> more <br>'s may help there
[11:55] <Hobbsee> oh wow.  i somehow feature on that list
[12:13] <seb128> TheMuso: hi, will you do the a11y GNOME updates?
[12:28] <pitti> seb128: what does "time gtk/jockey-gtk  --list" print for you? (real time)
[12:29] <seb128> pitti:
[12:29] <seb128> $ time /usr/bin/jockey-gtk --list
[12:29] <seb128> real	0m5.777s
[12:29] <pitti> seb128: could you replace /usr/lib/python2.5/site-packages/jockey/detection.py with http://people.ubuntu.com/~pitti/tmp/detection.py and check again?
[12:30] <seb128> pitti: do you know how to disable the caching?
[12:30] <pitti> it should print the same drivers, but *much* faster
[12:30] <seb128> I know you can do it by echoing something somewhere
[12:30] <pitti> seb128: which caching?
[12:30] <seb128> the disk cache
[12:30] <seb128> because
[12:30] <pitti> seb128: the jockey slowness is 95% due to inefficient data structures, and only 5% due to modinfo/apt-cache calls
[12:30] <seb128> ok
[12:30] <doko> seb128: will you upload gnome-menus before beta?
[12:30] <pitti> seb128: although I already optimized apt-cache show calls
[12:30] <seb128> $ time /usr/bin/jockey-gtk --list
[12:30] <seb128> real	0m4.955s
[12:31] <pitti> ugh
[12:31] <seb128> doko: yes, GNOME 2.22.0 tarballs flying today
[12:31] <pitti> seb128: maybe try to remove the .pyc?
[12:31] <doko> ok, thanks
[12:31] <seb128> pitti: no, that's still the old version, with hot cache
[12:31] <pitti> ah
[12:31] <seb128> trying the new one now
[12:32] <seb128> pitti:
[12:32] <seb128> $ time /usr/bin/jockey-gtk --list
[12:32] <seb128> real	0m1.648s
[12:32] <seb128> much better ;-)
[12:32] <doko> seb128: deskbar-applet alacarte  as well?
[12:32] <pitti> hm, here it dropped from 5.5 to 0.9, but still
[12:32] <seb128> doko: deskbar-applet yes, alacarte likely but I can do an upload if they don't roll a new tarball
[12:32] <pitti> seb128: might be even faster with the .pyc then
[12:33] <seb128> doko: do they just need a rebuild?
[12:33] <pitti> seb128: ok, thanks; I think I can speed it up even further, but that's already something
[12:33] <doko> seb128: yes, just trying to avoid unnecessary rebuilds
[12:33] <seb128> pitti:
[12:33] <seb128> $ time /usr/bin/jockey-gtk --list
[12:33] <seb128> real	0m1.012s
[12:33] <seb128> pitti: with the pyc
[12:33] <pitti> \o/
[12:33]  * seb128 hugs pitti
[12:34] <seb128> doko: ok, will upload those
[12:34] <seb128> pitti: so 5.7s to 1.0, that's good ;-)
[12:34] <pitti> seb128: most of the time is spent on modalias pattern matching
[12:35] <pitti> seb128: i. e. testing whether a modalias matches any of the globs in /lib/modules/2.6.24-12-generic/modules.alias
[12:35] <seb128> pitti: can you cache that?
[12:36] <pitti> seb128: I could pickle the entire (huge) data structure
[12:36] <seb128> pitti: in fact is there any need to run jockey at all on a configuration which didn't change?
[12:36] <pitti> seb128: depends; there might be a new driver available
[12:37] <pitti> seb128: later, if we have remote driver DB check, I need to come up with something new
[12:37] <pitti> seb128: but for now I could trash the cache if there's a new kernel or new hw
[12:40] <seb128> pitti: right
[12:40] <seb128> pitti: if the code is hard to optimize we can also try to delay the startup by 60 seconds or something
[12:40] <pitti> seb128: that would indeed be better
[12:41] <pitti> seb128: eventually, with network lookups, it'd might run a minute
[12:41] <pitti> seb128: does gnome-session start jobs in parallel?
[12:41] <seb128> yes
[12:41] <pitti> ok, so it is just slow because it actually takes much CPU/IO
[12:41] <seb128> the issue is mostly the number of ios
[12:42] <pitti> not because the later session scripts are blocking on the previous one
[12:42] <seb128> if you log again with a hot cache it's blasing fast
[12:42] <seb128> like 5 seconds instead of 30 seconds on my laptop
[12:42] <pitti> right
[12:42] <lamont> seb128: it'd be wonderful if gnome-session allowed me to say "and wait for this to exit before launching anything else"
[12:43] <pitti> seb128: hm, can I specify a delay in the .desktop file, or do I need a --sleep n option in jockey?
[12:43] <seb128> lamont: would be nice but it's not designed for that right now
[12:43] <lamont> yeah
[12:43] <seb128> pitti: you need a --sleep for now
[12:43] <pitti> seb128: ok; that's easy to add
[12:43] <seb128> pitti: but we might want to consider for next cycle a way to delay startups and speak with upstream about it
[12:43] <lamont> I've been resolving that since warty with a shell script and dpkg-divert
[12:44] <pitti> seb128: Exec= is shell, or an actual command?
[12:44] <pitti> seb128: i. e. would Exec=sleep 60; jockey-gtk --check work?
[12:44] <pitti> s/command/program/
[12:49] <seb128> pitti: I think it's a command, not sure though
[12:49] <seb128> I've to go for lunch now but I'll have a look after that
[12:50] <pitti> lunch -> me 2
[13:26] <Amaranth> making gnome-session start compiz first and finish starting it before continuing with everything would probably speed it up quite a lot
[13:27] <seb128> Amaranth: why?
[13:27] <Amaranth> seb128: because it wouldn't have to deal with copying so many pixmaps into textures right at start time
[13:28] <Amaranth> although that's something that should be faster (almost free) as drivers get better
[13:29] <PecisDarbs> does daily livecd still goes in gdm login loop?
[13:29] <Amaranth> PecisDarbs: you're an ati mobility user?
[13:30] <PecisDarbs> well, yes, but this loop also appeared on non-ATI systems too
[13:31] <PecisDarbs> Xorg starts succesfully, but when gdm counts secs and tries to login, it fails after some time and starts again
[13:31] <Amaranth> ok then, not something i need to deal with :)
[13:31] <PecisDarbs> :)
[13:31] <Amaranth> on the ati that'd be your driver crashing Xorg as it tries to start compiz
[13:32] <PecisDarbs> btw, live cd finally launches Xorg for Radeon Mobility successfully, thanks. Before that it screwed up frequencies up.
[13:32] <ryanakca> dholbach: can you break your lock on the 5-a-day branch? http://pastebin.ca/936470
[13:33] <dholbach> ryanakca: done - thanks for letting me know
[13:33] <ryanakca> dholbach: cheers :)
[13:33] <dholbach> ryanakca: I think you could have broken it yourself by just running      bzr break-lock      in the branch
[13:33] <dholbach> but I'm not 100% sure
[13:33] <dholbach> thanks in any case :)
[13:40] <pitti> mvo: hm, according to you https://bugs.edge.launchpad.net/ubuntu/+source/scim/+bug/199030/comments/18 should be wrong?
[13:41] <mvo> pitti: it used to be a global setting, but after some discussion it was decided it should be per-user
[13:44] <tseliot> mvo: have you reviewed EnvyNG?
[13:51] <zul> slangasek: pin
[13:51] <zul> er ping even
[13:59]  * jdong kicks scim
[13:59] <jdong> I don't know what I'm pressing by accident, but my keyboard randomly switches to arabic
[13:59] <jdong> as amusing as that is, it does get a bit irritating
[14:00] <seb128> jdong: ctrl-space?
[14:01] <jdstrand> hi pitti! how are you?
[14:01] <jdong> seb128: ah, that's it
[14:02]  * jdong unbinds the key
[14:02] <ubotu> Launchpad bug 199030 in scim "Can't close SCIM" [High,Confirmed]
[14:02] <mvo> tseliot: not yet, sorry
[14:03] <ion_> Also shift-space
[14:03] <tseliot> mvo: no problem, BenC has to review it too
[14:19] <pitti> seb128: jockey autostart delay of a minute committed
[14:19]  * seb128 hugs pitti, thanks ;-)
[14:29]  * persia comments that ALT+` is a common alternate combination to open the IME for keyboards lacking a sensible 半角／全角 key for Japanese locales, if there is still interest in finding an alternative keystroke combination (shift-space is especially annoying when one has several IME-specific keys)
[14:51] <ArneGoetje> persia: shift+space was the default key combination in kinput2 for japanese IME
[14:52] <persia> ArneGoetje: Yes, I know.  I turned that off too :)
[14:52] <ArneGoetje> persia: I don't linke it either.
[14:52] <ion_> Alt+` would be a much nicer default.
[14:53] <persia> Still, when one has a 半角／全角 one ought to be able to use it, and Alt+` seems to be the current standard choice for switching when one doesn't have special keys, for IMEs in common use by the average computer user.
[14:53] <ArneGoetje> persia: and the debian package of scim also doesn't use it... so, I don't know who made the patch and why, but I suggest that it'll be changed. BUt we need to have some comments from japanese users here I huess...
[14:53] <ArneGoetje> guess
[14:54] <persia> ArneGoetje: I suppose it could be raised at the LoCo meeting tomorrow, if there is time...
[14:54] <ArneGoetje> persia: that wold be an idea, yes
[14:55] <pitti> glatzor: hi!
[14:56] <pitti> glatzor: WDYT about changing p-distutils-extra to not merge po files all the time?
[14:56] <persia> ArneGoetje: On another note, I'm still catching up on email backlog, but was it ever determined who could negotiate with IPA to help set an appropriate license?
[14:56] <pitti> glatzor: it creates a lot of clutter in VCS, and at least I usually want to run it manually anyway
[14:59] <ArneGoetje> persia: not that I know of. I just made my comment but I don't feel to be in a position to negotiate with them
[15:00] <persia> ArneGoetje: Is that something you can escalate, or should I be chasing someone else?
[15:01] <ArneGoetje> persia: I thought Jun Kobayashi has contact with IPA...
[15:01] <ArneGoetje> persia: this is not only a Ubuntu specific issue...
[15:02] <glatzor> pitti: servus, you are my prime user :) Feel free to change it. I am mostly using bzr-buildpackage, so I don't get any clutter
[15:02] <pitti> glatzor: ah, that avoids it? nice
[15:02] <glatzor> pitti: it builds in a different folder
[15:03] <pitti> glatzor: ok, thanks; I'll think I'll fix it anyway and introduce a 'setup.py merge-po' command?
[15:03] <persia> He does, but at the 2008 Spring Tokyo Open Source Convention, he reported that his discussions were left that they would like to discuss with someone representing Ubuntu internationalisation.  Do you feel that it should be delegated to Kobayashi-san, acting on behalf of Ubuntu?
[15:03] <glatzor> pitti: got idea
[15:03] <glatzor> good idea
[15:03] <glatzor> :)
[15:07] <ArneGoetje> persia: I'm not sure... if they want to talk to someone from Canonical, that would indicate that they really want to keep their self crafted license and just adjust it so that we can redistribute it as a partner package or so... but I'm of the opinion, that they should scrap their self-made license and adopt a available open source license... as I have never done such license negotiations before, I'd like to have some comments/advice from so
[15:09] <persia> ArneGoetje: Well, I think it is somewhat complicated by the differences in the meaning of copyright in Japan and other places.  Japan is civil law, but not as flexible as many European jurisdictions.  Many standard open-source licenses are technically illegal.
[15:10] <ArneGoetje> persia: see, I have no idea about that... so, I'm definietely not the right person to do the negotiations then... ;)
[15:10] <persia> (not in that they violate the law, but in that they indicate actions not explicitly permitted by the law)
[15:10] <persia> ArneGoetje: OK.  Maybe there is a licensing contact with whom Kobayashi-san can coordinate?
[15:12] <ArneGoetje> persia: hmm... I don't know who does this kind of things... Business department?
[15:12] <mvo> pitti: could you name it "setup.py update-po" ? that would be more consistent with the makefiles usually found in po/ ?
[15:12] <persia> ArneGoetje: I've no idea at all.  Last person I asked suggested I contact you :)
[15:13] <ArneGoetje> persia: not for licensing though. ;)
[15:14]  * persia suspects it was a reflex CJK-fonts -> ArneGoetje response
[15:14] <ArneGoetje> persia: probably
[15:15] <dholbach> I doubt the business department can help with that - the archive admins deal with source code licenses the most
[15:16] <ArneGoetje> dholbach: anyone you can suggest?
[15:16] <persia> Any archive admins feel like coordinating with ubuntu-ja to help convince a Japanese government agency to release fonts in a DFSG manner?
[15:17] <dholbach> ArneGoetje: anybody on https://launchpad.net/~ubuntu-archive/+members might know
[15:18] <pitti> mvo: sure, I don't particularly mind
[15:22] <mvo> thanks
[15:35] <geser> pitti: Hi, please give-back: packagekit-gnome. Thanks
[15:46] <wasabi> So is policykit going to slowly supplant gksudo?
[15:46] <Mithrandir> geser: given-back.
[15:50] <Amaranth> wasabi: that's the idea
[15:50] <Amaranth> at least at first
[15:51] <wasabi> I really like it.
[15:51] <wasabi> Except when I logged in just now, it prompted me for my password... to cahnge timezones.
[15:51] <wasabi> With no indication about WHY it wanted to change timezones.
[15:51] <wasabi> And then the timezone got changed to the wrong one. :)
[16:28] <luisbg> dholbach, ping
[16:30] <dholbach> luisbg: pong
[16:45] <pitti> geser: done
[17:03] <Diablo-D3> hey all
[17:03] <Diablo-D3> is it me, or has alsa broken on the newest kernel in hardy?
[17:07] <stdin> http://launchpad.net/bugs/200338
[17:07] <ubotwo> Launchpad bug 200338 in linux "no sound hardy kernel 2.6.24-12 " [Critical,Fix released]
[17:08] <_MMA_> hahahah. Just started a new install because I thought I miffed something. :P
[17:08] <Diablo-D3> thank god I left -11 installed
[17:11] <ion_> Nice punctuation! In the bug report, i mean! That is, the one about the sound problem!
[17:13]  * _MMA_ sees it most often from European kids and just doesn't get it.
[17:13] <ArneGoetje> I'm a bit puzzeled with LP bug 199592, as I cannot reproduce it at all. Neither on the current Live CD image (from today), nor on my local system with the latest packages installed... Can anyone else reproduce the reported crash? Aparently it should be a common issue, as there are so many duplicate bug reports for it... :-/
[17:13] <ubotwo> ArneGoetje: Bug 199592 on http://launchpad.net/bugs/199592 is private
[17:29] <pitti> cjwatson, TheMuso: any chance you could test the newer b43-fwcutter from Debian? (bug 199994)
[17:29] <ubotwo> Launchpad bug 199994 in b43-fwcutter "Use version 4.80.53.0 of Broadcom's proprietary driver. " [Undecided,In progress] https://launchpad.net/bugs/199994
[17:30] <adrian15> Hi... Is there any Ubuntu installation - GRUB setup developer round here? Or anyone who can point me to its source code? I want to know if you use device commands to contemplate the Windows chainloading of Ubuntu's grub partition boot sector codes.
[18:06] <slangasek> zul: pong
[18:08] <zul> slangasek: hi, a couple of things I notcied that there is samba-2.0.28a in the svn tree I was wondering when its going to be in unstable, and there is a whole bunch of ebox modules that is stuck in new I was wondering if you can get those in the archive
[18:09] <slangasek> zul: I don't have an eta for samba 3.0.28a in Debian, I'm leaving that up to Christian; the NEW processing I'm working through this morning, yes
[18:10] <zul> slangasek: thanks..
[18:14] <shaya> is there a problem w/ sound in feisty's kernel?
[18:14] <shaya> I mean gutsy's
[18:15] <shaya> snd: Unknown symbol unregister_sound_special (and the like)
[18:15] <ion_> Yes. A fix is being built.
[18:15] <slangasek> you mean hardy's... :)
[18:15] <shaya> right
[18:15] <shaya> yes
[18:16] <Pici> shaya: see topic in #ubuntu+1
[18:16] <shaya> ah, ok
[18:17] <slomo_> doko: yes
[18:17] <doko> slomo_: thanks, last d/b-d on python-xml in main
[18:18] <slomo_> doko: don't worry, tomorrow early in the morning :)  did you take a look at my patch? my python is not good and it works but that's all i can say about it ;)
[18:20] <doko> slomo_: I don't care as long as you can still read the docs ;)
[18:20] <ion_> themuso: Re: your change in initramfs-tools 0.85eubuntu27, wouldn’t cp -al suffice instead of ln -f || cp -a?
[18:26] <ion_> themuso: Sorry, i remembered how -l works incorrectly.
[18:30] <LaserJock> anybody know if shipit is going to ship Ubuntu Alternate CDs for hardy?
[18:30] <LaserJock> I haven't heard anything yet
[18:33]  * _MMA_ waves at LaserJock but doesn't know the answer.
[18:34] <LaserJock> pfft, lotta good you are ;p
[18:35] <_MMA_> Yay! I'm useless!
[18:39] <mario_limonciell> _MMA_, why?
[18:40] <_MMA_> mario_limonciell: Because of my inability to answer Jordan. ;)
[18:41] <mario_limonciell> oh I thought you were meaning because of the sound fiasco atm :)
[18:41] <LaserJock> mario_limonciell: no, he must have an answer to all of my questions to feel any sense of usefulness ;-)
[18:43] <_MMA_> Yes. It's tied to my self-worth.
[18:44] <mario_limonciell> well some people have "mommy issues" or "daddy issues".  you've just got issues with independence.  i guess that's alright
[18:44] <LaserJock> "LaserJock issues"
[18:45] <mario_limonciell> step 1 is admitting you've got a problem.  we'll work on it at UDS
[18:45] <_MMA_> Or Jordan and I will work on it if I get him to Raleigh. ;)
[18:46] <LaserJock> lol
[19:15] <ar3ac> latest update give me troubles with my sound card
[19:16] <_MMA_> ar3ac: Known on Hardy. Use old (-11) kernel.
[19:16] <ar3ac> ok thanks
[19:17] <ar3ac> i'm reading the launchpad now
[19:18] <ar3ac> even quicker workaround:
[19:18] <ar3ac> sudo cp /lib/modules/2.6.24-11-generic/kernel/sound/soundcore.ko /lib/modules/2.6.24-12-generic/kernel/sound/
[19:18] <ar3ac> sudo depmod -a
[19:18] <ar3ac> sudo modprobe snd-hda-intel (or whatever you need)
[19:18] <ar3ac> and sound works again :)
[19:31] <slangasek> ogra_cmpc: why does classmate-tools depend on the desktop metapackages?  that seems an odd way to arrange it
[19:32] <ogra_cmpc> slangasek, its a tray icon it needs a system tray and an xdg capable desktop
[19:32] <slangasek> huh, ok
[19:33] <mario_limonciell> surely there are saner ways to install xdg compatible desktops?
[19:33] <ogra_cmpc> might be
[19:33] <slangasek> mario_limonciell: saner ways to request an xdg-compatible desktop, I would think
[19:33] <slangasek> but no matter
[19:34] <mario_limonciell> well but doing it this way, what if someone was to want to remove say ubuntu-desktop because they were maybe removing evolution since they didn't need it.  wouldn't classmate-tools break then?
[19:35] <slangasek> it would uninstall
[19:35] <nxvl> is there any ubuntu-drivers channel?
[19:36] <mario_limonciell> exactly, so it really should be depending on the packages that provide the system tray, like gnome-panel, pypanel, etc then
[19:37] <cjwatson> nxvl: you mean like the ubuntu-drivers team in Launchpad, or like kernel drivers?
[19:37] <nxvl> cjwatson: i mean the irc channel where the people who take care of the driver are, where i can find someone to report the problem with my sound card
[19:37] <nxvl> or/and to point me where i can report it
[19:37] <nxvl> :D
[19:38] <_MMA_> nxvl: -12 hardy kernel?
[19:39] <zdzichuBG> heh, fedora people also lost sound with latest kernel upgrade ;)
[19:39] <_MMA_> nxvl: If so, known issue. Will be fixed. Use -11.
[19:40] <cjwatson> nxvl: #ubuntu-kernel; if it's sound, they're already on it
[19:40] <mario_limonciell> nxvl, yeah it's been quite reported.  you can look at http://launchpad.net/bugs/200338 if you'd like to see about it
[19:41] <nxvl> _MMA_: yep, that's it, when i upgraded it to -12 problems started :D
[19:41] <nxvl> thanks all for his reponses!!
[19:42]  * nxvl HUGS all
[19:45] <LaserJock> cjwatson: do you know if Ubuntu Alt will be available on shipit for Edubuntu?
[19:50] <cjwatson> LaserJock: I don't, sorry; I can ask
[19:51] <cjwatson> LaserJock: the relevant people have gone home, though, so send me mail to remind me?
[19:51] <LaserJock> cjwatson: k, I just wondered as we had some questions in #edubuntu and I didn't really know what to say
[19:52] <_MMA_> Anyone know if samba is not working in Hardy? (might just be me)
[19:54] <candrews> https://bugs.launchpad.net/ubuntu/+bug/199754 I'm attempting to solve that issue (packaging mod_auth_cas). I have a package made, but when I load the module into apache, I get: Can't locate API module structure `authcas_module' in file /usr/lib/apache2/modules/mod_auth_cas.so: /usr/lib/apache2/modules/mod_auth_cas.so: undefined symbol: authcas_module
[19:54] <candrews> Can someone give me a hand to finish up this package?
[19:56] <slangasek> _MMA_: I know I just uploaded it yesterday, what's up?
[19:56]  * lamont wonders if slangasek gets annoyed by sync requests prior to dinstall even running...
[19:57] <_MMA_> slangasek: Well trying to browse with "Network" doesnt work. And after installing samba, smbfs and setting up fstab, the mounts dont come up.
[19:57] <slangasek> lamont: no, I just sit on them in favor of using my time more wisely :)
[19:58] <slangasek> _MMA_: is this a recent behavior change for you?
[19:59] <_MMA_> slangasek: Well, because using "Network" previously threw an error, I didn't jump to Hardy. After testing dailies, and saw the error was gone, I did a clean install on my main box thinking all was fine.
[20:00] <_MMA_> slangasek: Ive set things up the same way since Warty. Im unsure what's different.
[20:00] <slangasek> _MMA_: what's the server that you're trying to mount from?
[20:01] <_MMA_> slangasek: Ubuntu-Gutsy running samba.
[20:01] <lamont> slangasek: heh. ok
[20:01] <lamont> slangasek: it'll be a sync request for bind9 1:9.4.2-5 for hardy.
[20:02] <slangasek> _MMA_: oh, curious then.  The smbfs kernel driver is deprecated now in favor of cifs and mount.smbfs is now a thin compat layer on cifs, but that shouldn't normally be a problem for samba<->samba
[20:02] <slangasek> _MMA_: are you using credentials files?
[20:02] <_MMA_> slangasek: No. I have the user/pass defined in the fstab line.
[20:03] <lamont> feh.  for that matter, it's easier to file the sync request after packages.d.o updates
[20:03]  * lamont adds to tomorrows list
[20:03] <slangasek> _MMA_: I think I'd need to see the fstab line (minus the password) to diagnose, then
[20:04] <_MMA_> slangasek: Sure. 1 sec in PM.
[20:32] <wasabi> I just ran into another app that failed on Ubuntu 64 because of missing winbind nss/pam modules.
[20:32] <wasabi> (32 bit app)
[20:32] <wasabi> Alas.
[20:32]  * wasabi finds existing bug report from a year ago.
[20:41] <candrews> I'm attempting to package mod_auth_cas, I have everything done, except when I try to load the module into apache, I get this error: /usr/lib/apache2/modules/mod_auth_cas.so: /usr/lib/apache2/modules/mod_auth_cas.so: undefined symbol: authcas_module
[20:41] <candrews> Can someone help?
[20:41] <_MMA_> candrews: Packaging help is best in #ubuntu-motu.
[20:41] <candrews> thank you - I'll ask there.
[20:41] <_MMA_> np
[20:45] <slangasek> wasabi: what app, OOI?
[20:47] <TheMuso> ion_: The idea is to hard link, rather than copy. We only copy if the hard link fails.
[20:49] <ion_> themuso: Yes, and i misremembered that cp -l behaved just like that. :-)
[20:49] <ion_> themuso: Instead, it just fails if linking fails.
[21:10] <slangasek> glatzor: packagekit-gnome: "refreshing the cash"? :)
[21:12] <glatzor> slangasek: :)
[21:12] <slangasek> glatzor: s/cash/cache/, if that wasn't evident
[21:13] <glatzor> slangasek: I am currently uploading new packages to the ppa of the packagekit team.
[21:14] <glatzor> slangasek: the packages in universe are quite "rough"
[21:15] <slangasek> heh, ok
[21:55] <TheMuso> seb128: Yes, I'm on a11y stuff, which is mostly sponsored uploads.
[21:56] <seb128> TheMuso: right, that was not the case when I asked some hours ago ;-)
[22:16] <blueyed> Does somebody know if apt is supposed to resolve the "virtualbox-ose-modules" dependency of virtualbox-ose correctly (bug 188579) or if there's a workaround to give more hints for apt?
[22:16] <blueyed> It's fairly critical, because often wifi fails with the (new) -386 kernel.
[22:17] <blueyed> https://launchpad.net/bugs/188579
[22:19] <soren> blueyed: If I apt-get install virtualbox-ose-modules, apt wants to install   virtualbox-ose-modules virtualbox-ose-modules-2.6.24-12-generic virtualbox-ose-modules-generic
[22:19] <soren> blueyed: That looks correct, doesn't it?
[22:32] <shaya> the updated kernel for hardy that fixes sound taints the kernel
[22:32] <shaya>  snd: no version for "unregister_sound_special" found: kernel tainted.
[22:34] <blueyed> soren: yes, that looks correct. But if you apt-get install virtualbox-ose it's more likely to do not so..
[22:39] <soren> blueyed: http://paste.ubuntu-nl.org/59198/
[22:48] <blueyed> soren: are you on amd64 maybe? Do you have virtualbox-ose-modules-386 installable?
[22:49] <soren> blueyed: Ah, good point.
[22:49] <blueyed> I've thought about changing order as a (particular) workaround already, but that's not the root cause, so..
[22:52] <alefteris> could someone please point me to the package/code that is setting up the keyboard layouts (depenting on the language selection at boot) in the live cd and also the installed system?
[22:52] <tjaalton> alefteris: xkeyboard-config
[22:54] <blueyed> soren: http://paste.ubuntu-nl.org/59200/
[22:56] <soren> blueyed: Why is virtualbox-ose-modules a real package?
[22:57] <blueyed> soren: For automatic upgrades.. with only virtual packages they weren't upgraded automatically.. also the required(?) Replaces caused problems..
[22:57] <alefteris> tjaalton, will I be ale to define what layouts are loaded by default when a user selects for ex. Greek during the live cd boot? currentry in alpha 6 no latin layout is loaded, only the greek one. So its difficult to use the live cd or make the installation
[22:57] <cjwatson> alefteris: tjaalton is correct in that that's where the data comes from, but console-setup and xserver-xorg between them are what actually do the work
[22:57] <cody-somerville> The latest uploads seem to break sound on my laptop :/
[22:57] <blueyed> soren: Maybe depending on a particular kernel image in the meta packages would help?!
[22:57] <tjaalton> alefteris: I'm not that familiar with how it works, so can't answer :/
[22:57] <cjwatson> alefteris: console-setup has a scheme for dealing with that; please file a bug on https://bugs.launchpad.net/ubuntu/+source/console-setup/+filebug and we can deal with it
[22:58] <soren> blueyed: Only for users of -generic kernels :)
[22:58] <alefteris> ok thanks cjwatson and tjaalton
[22:58] <cjwatson> alefteris: interestingly, it's supposed to be dealt with properly already
[22:59] <cjwatson> alefteris: so there's a deeper bug somewhere and I'd definitely like a report so that we remember to investigate
[22:59] <mamefan> there's an ALSA bugfix that's been in the ALSA source for a while.  How does that get into ubuntu (it's still broken for me in the ubuntu packages)?
[23:00] <cjwatson> alefteris: please attach /etc/default/console-setup and /etc/X11/xorg.conf to the bug you file
[23:01] <amitk> mamefan: we are using alsa from upstream. Best way to track your problem is to file a bug in launchpad.net and point to the patch.
[23:01] <soren> blueyed: Maybe I can find a bit of time tomorrow to look into it. Right now, I need some sleep.
[23:01] <blueyed> soren: that would be great. AFAIR it also affects apparmor e.g.
[23:01] <cjwatson> alefteris: ... and /var/log/casper.log while you're at it
[23:02] <mamefan> amtik:  forgive my ignorance but I'm not really sure what that means.  I don't know what "patch" to point to.  I simply know that if I donwload the alsa-driver source from alsa and compile it on my machine it works.
[23:02] <alefteris> cjwatson, its difficult to issue them, without chaning the default live cd keyboard settings, as Im not able to type in latin, so I would have to add manualy the us layout before issuing those commands. is that ok?
[23:03] <cjwatson> alefteris: if you do it in GNOME, yes, that's fine
[23:03] <cjwatson> alefteris: failing that, a precise description of exactly what you selected at the CD boot menu will probably be enough
[23:04] <alefteris> ok i will get to it right away, thanks
[23:06] <YokoZa1> Does apt URL support multiple package downloads in a single link?
[23:06] <YokoZa1> Like can I do something like apt://package1+package2  ?
[23:33] <cjwatson> could somebody please provide a French->English translation of bug 199457?
[23:33] <ubotwo> Launchpad bug 199457 in openoffice.org "Impossible de filtrer les enregistrements dans un mailing" [Undecided,New] https://launchpad.net/bugs/199457
[23:33] <LjL> i'll try
[23:34] <cjwatson> thanks
[23:41] <LjL> done, as well as i could
[23:43] <cjwatson> LjL: thanks!
[23:44] <Riddell> pitti, doko: would either of you be able to MIR review bug 200799 before beta freeze?  it's just a widget style
[23:44] <ubotwo> Launchpad bug 200799 in kde-style-qtcurve "kde-style-qtcurve inclusion to the main repository" [Undecided,New] https://launchpad.net/bugs/200799
[23:51] <alefteris> cjwatson, here you are https://bugs.edge.launchpad.net/ubuntu/+source/console-setup/+bug/200803
[23:51] <ubotwo> alefteris: Error: Could not parse data returned by Launchpad: The read operation timed out
[23:52] <cjwatson> alefteris: thanks