[12:08] <YokoZar> haggai: ping?
[12:10] <sladen> sivang: grep name /proc/cpuinfo
[12:11] <sivang> model name      : Mobile Intel(R) Pentium(R) 4 - M CPU 1.80GHz
[12:12] <sladen> sivang: grep -c flags.*est /proc/cpuinfo
[12:13] <sivang> 0
[12:31] <pvh> Why is the resume= option needed for disksuspend not included in the menu.lst on laptops?
[12:32] <mdz> pvh: https://bugzilla.ubuntu.com/show_bug.cgi?id=5230
[12:33] <YokoZar> mdz: are you interested in signing my GPG key?
[12:37] <sladen> pvh: if mjg59's patches went in, you should be able to do it by echo'ing > /proc/...something
[12:38] <pvh> sladen: It's a bit messy at the moment, isn't it?
[12:38] <dholbach> does anyone have an idea, how i could debug a umount-of-a-fat32-partition-that-completely-kills-the-system?
[12:38] <pvh> sladen: Someone else was telling me to use /sys/power/state
[12:38] <pvh> sladen: And some of the documentation talks about /proc/acpi/event
[12:43] <dholbach> bye guys... i'm off to bed
[12:44] <Mithrandir> mdz?
[12:47] <sladen> pvh: (1) ACPI events (power-button-press).  (2) cause the system to enter the state (hibernate).  (3) onto the device selected (>/proc/power/resume)    Which bit do you feel is unneccessaryily messy?
[12:47] <pvh> sladen: Mm, no it's not messy in that sense. I don't mean to insult.
[12:48] <pvh> sladen: It's just that there are still lots of obsolete ways left in.
[12:49] <pvh> sladen: I think that a little bit of Hoary targeted documentation would resolve the situation.
[12:49] <sladen> pvh: I doubt anyone is insulted.  How do you feel it could be improved?
[12:49] <pvh> sladen: Once I've figured all the steps out, I might be able to supply a first-pass at that.
[12:50] <pvh> sladen: Unless there's already some good documentation out there that I've missed?
[12:51] <pvh> sladen: The wiki stuff is targeted to 2.4 kernels.
[12:51] <sladen> kinnison: your P4 appears to share the characteristics of sivangs;  does the -smi fallback fix^W work?
[12:59] <bluefoxicy> anyone wanna show me a walkthrough for building an official amd64 generic here?
[12:59] <bluefoxicy> uri or whatever
[12:59] <bluefoxicy> I tried on my own but i wound up with a kernel package which didn't actually install anything in /boot
[01:00] <bluefoxicy> (using make-kpkg)
[01:11] <lamont> bluefoxicy: apt-get install linux-source-2.6.10; dpkg-buildpackage
[01:11] <bluefoxicy> lamont:  nothing I have to touch around there?
[01:11] <lamont> don't think so...
[01:11] <bluefoxicy> i.e. to put out another release or another arch or such
[01:12] <lamont> assuming the kernel you want to build is one of the flavors found in the archive, dpkg-buildpackage will build it...
[01:12] <lamont> along with the others, mind you.
[01:12] <bluefoxicy> I'm trying to gauge if I'd be able to maintain grsecurity sources for amd64
[01:12] <bluefoxicy> I can patch the thing easy (2 misses, 1 is relavent 1 is just makefile EXTRAVERSION)
[01:12] <bluefoxicy> err, *kernels
[01:13] <bluefoxicy> pitti's got a few experimental ones up for x86
[01:13] <bluefoxicy> but none for amd64
[01:15] <AndyR> hi all
[01:15] <bluefoxicy> until ubuntu *cough* actually picks up grsecurity (possible?  maybe, maybe not), I've got two choices: build my own directly (I can do that), or build my own packages (I can't do that).  The latter provides a better learning experience; I can build kernels by hand with my eyes closed.
[01:16] <bluefoxicy> (hopefully so can everyone else here)
[01:19] <jdub> bluefoxicy: building kernel packages is easy, see kernel-package
[01:31] <sivang> I'm dead tired, good night freedom lovers! :)
[01:58] <Mitario> hi everyone
[02:12] <jdub> when using debmirror, will the target directory end up being $ROOT/{dists,pool}
[02:13] <jdub> or have dists/pool in it
[02:13] <jdub> directly
[02:20] <jdub> n/m, appears to put {dists,pool} directly
[03:36] <sladen> wow, this looks *nice*:  http://www.ubuntu-fr.org/
[03:53] <macewan> ubuntu frog?
[04:58] <zul> crunchy frog
[05:02] <lamont> sladen: looks very french. :-)
[05:04] <macewan> rather elegant layout. easy on the eyes & pleasant.
[05:12] <lamont> yes.  even with my poor-to-nonexistant french, it looks nice
[06:03] <tseng> looks nice besides the nipples-of-doom shot
[06:03] <wasabi> heh. somebody just gave me the crazy idea of putting .torrent files in apt
[07:07] <dholbach> morning
[07:27] <dholbach> hmmm, i thought my unmount-vfat--system-crash was related to autofs and i solved it that way.... but unfortunately - it's still there - hmmmmmmmm
[07:38] <sivang> morning all
[07:42] <dholbach> hi sivang :-)
[07:43] <sivang> dholbach: hey daniel
[07:45] <dholbach> sivan, how are you?
[07:50] <sabdfl> hi guys, what's our best answer for wifi network selection from the desktop?
[07:50] <sabdfl> in hoary
[07:52] <jdub> sabdfl: install netapplet
[07:52] <jdub> sabdfl: it's not on-by-default yet
[07:52] <sabdfl> jdub: is that what we plan to make on-by-default?
[07:53] <jdub> for hoary, yeah
[07:53] <jdub> it's only just been appletised
[07:53] <sabdfl> cool, thanks
[07:56] <dholbach> bye, i'm out... running
[07:59] <sabdfl> elmo: good work on the universe / multiverse sync
[08:23] <sivang> rehi sabdfl :)
[08:23] <sabdfl> hey sivang
[08:54] <dholbach> re
[08:55] <jdub> hrm
[08:55] <jdub> how do you get the installer stuff with debmirror?
[09:00] <jdub> aha, figured it out
[09:00] <jdub> sections should include main/debian-installer,restricted/debian-installer
[09:50] <jdub> daniels: around?
[09:58] <Treenaks> hm, tsclient displays a localized banner, using gtk_get_default_language() to get the locale
[09:59] <Treenaks> but the banner contains text, and gtk_get_default_language() returns the wrong part of my locale for that (i.e. not LC_MESSAGES)
[09:59] <Treenaks> Where's the bug? tsclient or gtk/pango?
[10:04] <Treenaks> I'd say GTK (it returns LC_CTYPE, not LC_MESSAGES)..
[10:08] <dholbach> re
[10:15] <dholbach> would anyone sponsor me a bluefish-1.0 upload?
[10:17] <jdub> daniels: http://www.cse.unsw.edu.au/~chak/linux/c400.html#xfree86
[10:18] <amu> dholbach: url?
[10:19] <dholbach> amu:    deb-src http://ubuntu.gplan.info hoary main    (although i should change main to universe or something) :-))
[10:22] <mdz> morning
[10:23] <sivang> morning mdz 
[10:30] <fabbione> morning
[10:31] <sivang> morning fabbione, how do you feel?
[10:32] <fabbione> trashed
[10:32] <fabbione> yesterday they managed to drag me out of the house with a really fancy excuse... and organized a surprise bachelor party
[10:33] <Treenaks> what was the excuse? ;)
[10:33] <fabbione> installing ubuntu sparc in a telco datacenter
[10:33] <Treenaks> LOL
[10:33] <fabbione> i couldn't resist the amount of nice hardware they put in the list
[10:35] <Treenaks> jdub: a link from that page (http://linmodems.technion.ac.il/pctel-linux/welcome.html) would likely make my winmodem work :) thanks! :)
[10:36] <Treenaks> (hm, no, 2.4 only)
[10:37] <infinity> fabbione : That's some excuse. :)
[10:39] <dholbach> fabbione: you have an idea, what could cause  umount -ing a vfat partition to kill the whole system? or how i could debug it?
[10:39] <fabbione> dholbach: #5431
[10:39] <dholbach> oh ok *has a look*
[10:41] <dholbach> funnily enough, if i did it in the console and stopped gnome-related processes, i could umount it flawlessly
[10:41] <dholbach> but both times (according to lsof) no processes had open files on it
[10:43] <fabbione> dholbach: the "open files" are in the kernel
[10:43] <fabbione> you don't see them in lsof
[10:43] <dholbach> aha... ok 
[10:45] <fabbione> mdz: confirmed.. it works perfectly with growisofs :-)
[10:46] <mdz> fabbione: yes, someone should mail its author about the silly name
[10:48] <fabbione> or we can just ln -sf growisofs irockondvds
[10:48] <jdub> is growisofs generally better than cdrecord?
[10:48] <fabbione> no correction..
[10:50] <jdub> aha
[10:50] <jdub> bong
[10:50] <jdub> Package: sl-modem-daemon
[10:50] <jdub> Depends: libasound2 (>> 1.0.8), libc6 (>= 2.3.2.ds1-4), debconf, sl-modem-modules-new | sl-modem-source (>> 2.9.6-1) | kernel-image-2.6
[10:50] <jdub> *kernel*-image
[10:51] <Treenaks> jdub: scary
[10:51] <sivang> Treenaks: that's a link to a webserver on my former uni :) how did you come by it?
[10:51] <Treenaks> sivang: the page jdub pasted had a link to it :)
[10:51] <sivang> Treenaks: ah cool. *searching the backlog*
[10:54] <jdub> hrm
[10:54] <Treenaks> slmodem is scary
[10:54] <jdub> can you use globs/regexps in Depends package *names*?
[10:55] <jdub> mdz, fabbione: ^
[10:55] <Treenaks> jdub: linux-image-2.6.[0-9] + ? ;)
[10:55] <fabbione> jdub: put down the pipe man...
[10:56] <jdub> fabbione: i know :-)
[10:56] <jdub> but
[10:56] <mdz> jdub: BONG
[10:56] <jdub> we don't have a metapackage that is cpu independent
[10:57] <jdub> so i'm stuck depending on linux-image-386 and -686 and -k7 and -386-smp and -686-smp and -k7-smp and any future ones :)
[10:57] <fabbione> hmm
[10:57] <fabbione> Package: linux-image-2.6.10-3-386
[10:58] <fabbione> Provides: linux-image, linux-image-2.6
[10:58] <fabbione> that's all you need
[10:58] <fabbione> and it's there already
[10:58] <jdub> och
[10:58] <jdub> good point
[10:58] <jdub> thank you fabio
[10:58] <jdub> it is not even late :)
[10:58] <fabbione> so you can just Depends: linux-image-2.6 | linux-image-whatever-real-package
[10:59] <jdub> do i have to | it with a real package?
[10:59] <fabbione> yes
[10:59] <fabbione> or the other way around
[10:59] <fabbione> check with lintian
[11:07] <infinity> The other way around.
[11:07] <infinity> Though, it's pretty sick to be depeding on a kernel image at all.
[11:08] <infinity> jdub : Why is that necessary?
[11:13] <jdub> infinity: alsa driver in 2.6 instead of driver module provided
[11:14] <infinity> Yeah, I read the package description.  I'm still not convinced it's necessary. :)
[11:14] <infinity> Especially since "linux-image-2.6" could have anything builtin, including no sound support.
[11:26] <infinity> ARGH.
[11:26] <infinity> How does one stop CVS from doing keyword expansion?
[11:26] <infinity> It just destroyed a massive patch.
[11:29] <Treenaks> infinity: I think it was something with -k and cvs admin
[11:29] <Treenaks> 1 mom
[11:30] <Treenaks> +
[11:32] <Treenaks> infinity: https://www.cvshome.org/docs/manual/cvs-1.11.19/cvs_12.html#SEC97
[11:32] <Treenaks> https://www.cvshome.org/docs/manual/cvs-1.11.19/cvs_12.html#SEC100 even
[11:38] <infinity> Thanks.
[11:42] <infinity> Oh, thank god it's recoverable.  I didn't want to redo that patch..
[11:51] <dholbach> hi seb128
[11:51] <seb128> morning
[11:51] <sivang> seb128: morning
[11:57] <dholbach> jdub, seb128: gtkmm2.4-2.5.5 just hit debian experimental :-D
[11:59] <seb128> yeah, I know
[12:06] <seb128> jdub: what's going on with beagle ?
[12:06] <jdub> got a prelim package without any depends :)
[12:07] <jdub> haven't spent much time on it since last weekend
[12:07] <seb128> k
[12:08] <seb128> jdub: and about *mm, is that ok to sync 2.5.n ?
[12:09] <mvo_> jdub: when is feature freeze? at monday or after monday :) ?
[12:09] <jdub> how much stuff will that break in universe? :)
[12:09] <jdub> mvo_: wednesday
[12:11] <mvo_> jdub: ah, misread the schedule. thanks
[12:11] <jdub> yes, current format makes that confusing
[12:11] <seb128> jdub: gtkmm is in main, but since that's a GNOME stuff, for the freeze ... the breaking part is probably not an issue, the is not a ton of rdepends and dholbach has runned a part of them and that's fine
[12:11] <jdub> seb128: ok
[12:12] <jdub> seb128: (i am being abruti about C++ :-)
[12:12] <seb128> jdub: should I mail matt/you about this ?
[12:12] <seb128> ah ah
[12:12] <jdub> nah, go ahead
[12:13] <seb128> k, thanks
[12:13] <dholbach> jdub: what does abruti mean? :-)
[12:14] <jdub> dholbach: it's french for meathead.
[12:14] <jdub> kind of
[12:14] <jdub> seb was being very mean to me one day ;)
[12:15] <dholbach> abruti... sounds... erm ... nice :-)
[12:18] <jdub> oh man
[12:19] <jdub> hrm
[12:24] <amu> jdub: any new from Luis? he wanted to gimme some docs/artwork about gnome, looks like he's lost in bermuda ;) 
[12:25] <amu> s/new/news
[12:25] <jdub> i'll nudge him
[12:26] <jdub> interesting
[12:26] <jdub> anything playing through alsa plays fast
[12:27] <LarryT> hi people :)
[12:27] <jdub> unless it's playing through dmix
[12:27] <jdub> then it plays at normal speed
[12:27] <Kamion> jdub: depending on kernel images at all has always generally been considered BONG, except for kernel modules specifically compiled against a particular module ABI
[12:27] <jdub> Kamion: yeah? hmm.
[12:27] <jdub> Kamion: should i just take out all the kernel related depends?
[12:27] <jdub> seems like a very bong package
[12:28] <Kamion> I would, but I haven't looked at the package :-)
[12:28] <jdub> that said, everything but the driver seems to be Free
[12:28] <jdub> a no advertising clause is not considered non-free is it?
[12:28] <amu> hmm daily-live has still keyboard problems, no @ and | if i choose german locale, xorg say "UNKOWN"      
[12:28] <LarryT> someone could tell me if gparted will be on ubuntu , pleae ? :)
[12:29] <Kamion> LarryT: nobody's been working on it in Ubuntu. However, since Debian recently did a big parted upgrade, it might come into Debian soon; if so, there would be a case for us syncing it into universe.
[12:29] <LarryT> may be i am not on the right channel ? ...
[12:30] <amu> jdub: thx
[12:32] <dholbach> LarryT: i'm working on it
[12:32] <LarryT> there you're back :)
[12:32] <Kamion> ah, I wasn't aware dholbach had been doing stuff
[12:32] <dholbach> LarryT: but i want people to test it a bit, before it gets included and people jump down my throat because their partition was eaten :-)
[12:38] <amu> Kamion: tried remaster the ppc, as you suggested with a hfs.map from CVS, CD still does not boot :(      
[12:38] <Kamion> I'll have to try the instructions out myself later
[12:40] <amu> Kamion: thx
[12:48] <dholbach> but LarryT was right... gparted on the liveCD would be *NICE*
[12:50] <dholbach> so i probably should start a big call on package testing :-)
[01:12] <lifeless> can we get fuse in the official kernels ?
[01:12] <lifeless> or at least a built module ?
[01:16] <sladen> hostap would be groovy too;  then I could have wifi back on this laptop
[01:46] <sivang> I'm back
[01:49] <jdub> ahr! all my sound is playing fast
[02:05] <Mithrandir> does python -c  'import locale; locale.setlocale(locale.LC_ALL, ""); print locale.getlocale()' do anything useful for any of you?
[02:05] <Mithrandir> (generating a backtrace is not considered useful.)
[02:08] <sivang> ('en_US', 'utf')
[02:08] <Mithrandir> bah, stupid python.
[02:08] <sivang> this is my default one, should I list me all the ones I have?
[02:08] <Mithrandir> no, no need.
[02:08] <mvo_> Mithrandir: 'C'
[02:09] <Mithrandir> I found out what the problem is, and it's that python is dog stupid, having a hard-coded list of locales.
[02:13] <sivang> bah
[02:47] <daniels> jdub: solved problem; the video RAM hack has been in for ages so it will give you however much you ask for, and we have DRM too
[02:48] <jdub> daniels: didn't seem to work on avdd's
[02:48] <jdub> daniels: btw, i noticed that neither my laptop nor his loaded the appropriate drm kernel module
[02:49] <jdub> mine has never loaded it, it seems
[02:51] <daniels> i915?
[02:51] <daniels> should get loaded when needed
[02:52] <daniels> hoary-only, btw
[02:52] <jdub> his is an i830, mine's an i855
[02:52] <jdub> the Device>VideoRam setting didn't work
[02:52] <daniels> bongtasmic
[02:52] <daniels> using hoary?
[02:53] <jdub> but he got this awesome letterbox video mode
[02:53] <jdub> yeah
[02:53] <daniels> hmmmmm
[02:54] <jdub> he'll be over on wednesday, and possibly here earlier
[02:54] <daniels> haven't heard any other reports of that
[02:54] <daniels> and, I mean, I have an i855 ;)
[02:54] <jdub> does yours always load the i830 module?
[02:54] <amu> *hpf* shit, uploaded a source-packaged to debian 
[02:54] <mjg59> jdub: For 855, the  correct module is i915
[02:54] <jdub> (does it always unload it when X quits?)
[02:54] <mjg59> With Xorg
[02:55] <mjg59> For 830 as well, come to think of it
[02:55] <jdub> oh
[02:55] <jdub> right, xorg change
[02:55] <jdub> i see
[02:55] <daniels> jdub: mine always uses i915
[02:55] <daniels> jdub: and nah, it won't unload it
[02:55] <daniels> only load
[02:56] <jdub> ok, he must have i915 going
[02:57] <amu> elmo: ping
[03:02] <bob2> daniels: did you see the vberestore thread on lkml?
[03:04] <aj> elmo/lamont/anyone: ping? :)
[03:06] <daniels> bob2: nope
[03:06] <daniels> don't read lkml
[03:06] <bob2> ah
[03:06] <amu> daniels: i've still trouble with the liveCD, german keyboard, no @ and | is possible   
[03:07] <daniels> amu: sounds like you have pc104 instead of pc104
[03:07] <daniels> er, instead of pc105
[03:07] <daniels> bob2: when was it?
[03:07] <amu> xorg.config say 105
[03:07] <daniels> amu: really?
[03:07] <amu> yep
[03:07] <azeem> same here, pc105
[03:07] <daniels> bob2: ah, got it
[03:08] <bob2> daniels: hrm, yesterday I think
[03:08] <bob2> ah
[03:08] <bob2> dunno if it's got anything interesting, but it seemed like you might be able to fob some stuff off to the kernel now ;)
[03:13] <mjg59> bob2: Not as yet, sadly
[03:14] <bob2> dang
[03:27] <amu> daniels: XkbLayout says "UNKNOWN" 
[03:27] <daniels> amu: ?!?
[03:27] <daniels> is the debconf priority at that stage high, or critical?
[03:28] <daniels> i suspect we need to bump the question to critical if it's unknown
[03:28] <daniels> either that, or default to us/pc105 and hope for the best :\
[03:28] <daniels> anyway, I need to sleep.  'night all.
[03:28] <sivang>  night daniels 
[03:29] <daniels> amu: what language/location did you select?
[03:29] <amu> daniels: ..on a booted liveCD, i must check the debconf prio ..    
[03:29] <amu> daniels: german/german
[03:30] <daniels> amu: ok, i'll check it out tomorrow or tuesday
[03:30] <amu> daniels: ok, just let me know, if you need me to test something   
[03:31] <daniels> thanks dude, will do
[03:31] <amu> daniels: thx2 and sleep well 
[04:17] <lamont> aj: sup?
[04:23] <aj> lamont: irc.parisc isn't
[04:24] <lamont> aj: seems to be working for me... - see other window
[04:45] <sivang> Mithrandir: what are you using for your python hacking as in an IDE? vim/emacs/other? :)
[04:45] <Mithrandir> emacs
[04:46] <sivang> Mithrandir: (I am looking for something with object insight)
[04:46] <sivang> Mithrandir: cool, could you email me the python highlight parts of the config file? or is it some kind of a pluging that I have to install?
[04:47] <sivang> Mithrandir: (the one that tells which methods are available when issuing the dot :-)
[04:47] <Mithrandir> no idea, I just browse the docs with firefox.
[04:48] <infinity> object insight.  Bah.
[04:48] <infinity> That's crazy talk.
[04:49] <infinity> (but if you find a text-mode editor that does it well, tell me what it was?)
[04:49] <Mithrandir> it's hard to do with python, since it's so dynamic.
[04:50] <sivang> infinity: hehe
[04:51] <sivang> Mithrandir: how do you enable the highlighting in emacs for python (it works quite well for me on C and Perl)
[04:51] <Duck_busy> anyone working at the python 2.4 transition ?
[04:52] <Mithrandir> sivang: python-mode, possibly?  Also  (global-font-lock-mode t)
[04:52] <Mithrandir> in .emacs
[04:53] <sivang> Mithrandir: how do I enable it? (it's a major or minor mode_
[04:53] <sivang> )
[04:53] <infinity> Mithrandir : <shrug>... If visual studio can manage object insight in their crappy text editors, I'm sure some enterprising free software developer can make a text mode editor not suck at it.
[04:53] <sivang> infinity: we should try do something up, how hard can it be ? ;-) <hides>
[04:54] <Mithrandir> infinity: it's a lot easier to do in, say C++ than python.
[04:54] <infinity> I don't know about "hard".. But it'd be difficult to do it efficiently.
[04:55] <infinity> Visual Studio cheats by making you tell it ahead of time what you may or may not be linking to.
[04:55] <infinity> The "right" way would be to parse the source and determine what's being loaded, what it does, and how it's being used.  Which sounds.  Painfully inefficient.
[04:56] <sivang> infinity: right
[04:56] <sivang> :-/
[04:56] <infinity> Maybe I should extend ae to be the ultimate python editor. :)
[04:56] <sivang> infinity: ae?
[04:57] <infinity> sivang : My tiny editor of choice, cause I'm weird.
[04:57] <infinity> sivang : It was removed before woody.  Remind me to upload it before Sarge (and get it in Hoary universe.. Hrm.. I have how many days?)
[04:57] <sivang> infinity: till wedensday
[04:57] <sivang> ah, my crappie spelling...
[04:57] <infinity> Anyhow.  It's tiny.  It's featureless.  It's be a great base for adding worthless features like this, cause I wouldn't be stuck trying to slot it into the already-determined frameworks of vim and emacs.
[04:58] <sivang> infinity: right, what is it written in?
[04:58] <infinity> C.
[04:58] <sivang> infinity: superb :)
[04:58] <Mithrandir> infinity: you are _sick_
[04:58] <sivang> Mithrandir: LOLs
[04:59] <sivang> how it seems to me that all of the nice features an object has in python, would make it ideal to write such an editor in python , no? (let's not talk about the efficiency for the moment)
[04:59] <Mithrandir> sivang: have you ever used the monstrosity that is ae?
[05:00] <sivang> Mithrandir: I don't think so, does it follow vim/pico/nano/emacs bindings?
[05:00] <Mithrandir> well, you'd need to know the type of an object to introspect it.
[05:00] <Mithrandir> actually, you need the object itself.
[05:00] <sivang> Mithrandir: eh right..which mean instantiating it?
[05:01] <Mithrandir> you can't just do that.  Imagine a method foo(self, bar); how are you going to introspect on bar?
[05:01] <sivang> hrm right....:-/
[05:02] <Mithrandir> even introspecting on self can be hard enough, since something might have added properties to self which you use.
[05:02] <Mithrandir> not a good coding style, I agree, but very much possible.
[05:02] <infinity> Mithrandir : Oh, it's not that bad. :)
[05:03] <infinity> sivang : Keybindings are completely configurable, though it ships with some defaults people might be comfy with.  Does modal and modeless, etc.
[05:03] <infinity> Mithrandir : Doing it in a foolproof manner might be tough, but getting it "close enough" mightn't be too hard.
[05:04] <sivang> infinity: maybe to mere ease off the reference searching task...
[05:05] <Mithrandir> infinity: how would you introspect at bar above, then?  Analyze the call graph?
[05:06] <infinity> I could always embed python and use it to figure itself out...
[05:07] <sivang> infinity: is there a way to embed python in a python script?
[05:08] <infinity> Maybe.  I certainly hope not.
[05:08] <infinity> I suppose you could write a python python library that linked to libpython, but.  Uhm.  Why?
[05:08] <sivang> yes, that sounds, a bit how to say? strange?
[05:11] <Mithrandir> lamont: cyrus-sasl seems to be missing on amd64
[05:15] <Mithrandir> sivang: in general, for my emacs stuff just look at http://err.no/dotfiles/emacs
[05:16] <sivang> Mithrandir: thanks, this is a domain of yours?
[05:16] <Mithrandir> yes
[05:16] <sivang> wow :)
[05:24] <lupus_> seb128, can you update libdbus-cil for the new mono
[05:24] <seb128> why me ?
[05:25] <seb128> ask to one of the guys who work on it rather
[05:25] <tseng> lupus_: whats wrong with it
[05:25] <seb128> ie tseng :)
[05:25] <lupus_> dpkg: error processing /var/cache/apt/archives/mono-assemblies-base_1.0.5-2_all.deb (--unpack):
[05:25] <lupus_>  trying to overwrite `/usr/lib/mono', which is also in package libdbus-cil
[05:25] <lupus_> Errors were encountered while processing:
[05:25] <tseng> oh like that you mean
[05:25] <seb128> oh, there is a bug in bugzilla about it
[05:25] <tseng> has nothing to do with the "new mono" really.
[05:26] <lupus_> true
[05:26] <lupus_> let's say the new mono package :)
[05:26] <tseng> for now you can dpkg -i --force-overwrite /var/cache/apt/archives/mono-assemblies-base_1.0.5-2_all.deb
[05:26] <lupus_> lets 
[05:26] <tseng> and CC yourself on the bug.
[05:27] <lupus_> k
[05:27] <lupus_> thx
[05:28] <seb128> tseng: do you know if somebody has planned to update f-spot ?
[05:28] <tseng> i do not
[05:28] <seb128> k
[05:29] <tseng> havent met the maintainer yet, and i cant get on the pkg-mono list
[05:29] <tseng> as of yesterday
[05:29] <seb128> k
[05:29] <seb128> I'll ping the maintainer on the debian side first :)
[05:30] <tseng> alright.
[05:30] <sivang> seb128: do you know if there was any decisoin about removing the change UID field in users-admin "advanced" tab? There's a major bug report about this.
[05:30] <tseng> yay, got on the list now
[05:30] <seb128> sivang: I know, the bug is assigned to me :p
[05:30] <tseng> someone fixed mailman
[05:31] <seb128> sivang: I think we should just mask it from the UI
[05:31] <sivang> seb128: ok, that would be a very trivial patch to do, using glad etc, would you like me to do it and make a new pkg for you to review?
[05:31] <sivang> seb128: (i.e. a change only to the interface file)
[05:33] <seb128> yeah, that's probably a few lines changes
[05:33] <seb128> if you want to do it you're welcome
[05:35] <sivang> seb128: ok
[05:35] <sivang> seb128: tnx :)
[05:36] <seb128> thank you
[06:24] <T-Bone> Kamion: ?
[06:53] <Kamion> T-Bone: ?
[06:53] <Kamion> (here for about five minutes)
[06:56] <T-Bone> Kamion: did you change something to d-i lately?
[06:56] <T-Bone> yesterday's iso doesn't detect my CDrom drive anymore on ia64
[06:57] <Kamion> T-Bone: all my changes are mailed to hoary-changes, to which I'd expect developers to be subscribed to find out what's going on :)
[06:57] <Kamion> T-Bone: nothing springs to mind
[06:59] <T-Bone> i'm not subscribed to hoary-changes
[06:59] <T-Bone> i wonder what's going on. kernel modules seem properly loaded
[07:02] <Kamion> step through cdrom-detect's postinst I guess
[07:03] <T-Bone> Kamion: it's worth than that
[07:03] <T-Bone> off hand it looks like the kernel is fucked up
[07:03] <T-Bone> all ide modules are loaded, but it doesn't see anything on the ide bus
[07:16] <T-Bone> ok the kernel is fucked up
[07:16] <T-Bone> /proc/interrupts is mostly empty. The only thing working in USB!
[07:16] <Kamion> hmph, when my fiancee logs out of GNOME, a bunch of processes keep running, so she can't log in properly the second time round
[07:19] <Kamion> reproducible with a freshly created test user
[07:20] <sivang> Kamion: last upgrade?
[07:20] <Kamion> sivang: since a while, I think
[07:20] <ZorroBytes> Hi, what's the status of Java for ubuntu - does the license for java prevent it being bundled as an install? It comes with Suse 9.2 Pro default isntall
[07:21] <sivang> Kamion: I'm upgrading to see if it happens here
[07:21] <Kamion> most of them go away if I add the user to the audio group; with a user that I just created with 'adduser', the panel keeps running
[07:23] <sivang> Kamion: adduser doesn't apply all the needed groups I think, try to create a user in users-admin? 
[07:23] <Kamion> ZorroBytes: AIUI, main problem is that if we ship Sun Java, we can't ship any other Java
[07:24] <Kamion> sivang: oh I realise that pointy-clicky tools will add more groups and I've done that now by hand; however things should not break if users aren't in the groups they want
[07:24] <ZorroBytes> Kamion, good point, but in reality (I'm a java developer) it's the sun one that I install first off. Then if I want another version, say IBM's, I go get that myself.
[07:24] <Kamion> ZorroBytes: sadly, legal issues are even more real. :-)
[07:24] <ZorroBytes> but for the normal joe user out there, having sun installed as default wouldn't be too much of a problem?
[07:24] <Kamion> ZorroBytes: there's lots of stuff on the wiki though and jbailey is working on it
[07:24] <ZorroBytes> url?
[07:24] <Kamion> search for "Java"
[07:24] <ZorroBytes> ta.
[07:24] <Kamion> www.ubuntulinux.org/wiki/
[07:25] <Kamion> no, we're not going to install Sun Java by default, sorry
[07:25] <ZorroBytes> Got the page.
[07:25] <Kamion> there are a couple, there's JavaIntegration
[07:25] <ZorroBytes> that's the one I'm looking at at the moment :)
[07:26] <Kamion> anyway it's a long time since I used Java at all, so I'll bow out of the conversation now
[07:27] <ZorroBytes> ta for the info :)
[07:30] <Josephus> was there any change in gdm lately? I rebooted, and now gdm could not write my x auth file
[07:31] <Treenaks> Josephus: did you run some X program as root?
[07:31] <Josephus> possible yes
[07:34] <Treenaks> Josephus: that could be the problem then.. and don't run X programs as root :)
[07:36] <Josephus> ok i won't, but how can i get my X back? :)
[07:36] <Treenaks> remove .Xauthority from your ~, and this is really a #ubuntu question
[07:37] <Josephus> Treenaks: i thought it was because of some update
[07:38] <Josephus> But default cursor change was because today's update, as it was mentioned on ubuntu-devel list
[07:38] <sivang> Kamion: right
[07:38] <sivang> Kamion: well, upgraded I'll check around if it happens here
[07:55] <mxpxpod> Josephus: are they going to change back to the old default cursor?
[07:56] <sivang> Kamion: hrm, I can confirm that :-/
[07:57] <sivang> Kamion: only I couldn't even logout, but didn't any problem logging in
[09:07] <pitti> Good evening everybody
[09:10] <sivang> godd evening pitti :)
[09:10] <sivang> s/godd/good/
[09:27] <sivang> why smurfix ?
[09:27] <smurfix> I can IRC from here. But for some reason, ssh is totally impossible.
[09:28] <sivang> where are you at?
[09:28] <sivang> (if I may ask)
[09:29] <mdz> pitti: good evening
[09:29] <pitti> Hi mdz
[09:29] <mdz> pitti: do you have a reproducible test case for the udev bug?
[09:30] <smurfix> Tegernsee. Alps. The phone works, normally, but packet data doesn't like me here.
[09:30] <pitti> mdz: no, I cannot reproduce it at all on my machines
[09:31] <pitti> mdz: I tried to debug it on smurfix' machine, but did not come very far
[09:31] <smurfix> pitti: feel free to try again there, if necessary.
[09:32] <smurfix> I can also try to track it down -- next week, when I'm back home
[09:48] <mdz> pitti: I can reproduce it here
[09:48] <mdz> pitti: by uninstalling/installing hal
[09:49] <mdz> pitti: do you know the cause of hal.hotplug[18165] : SEQNUM is not set
[09:49] <mdz>  ?
[09:49] <pitti> mdz: I tried that several times (un/reinstall, down- and upgrades)
[09:50] <pitti> mdz: I don't know the cause
[09:51] <pitti> mdz: something seems to induce hotplug events without sequence numbers (which should normally be there)
[09:51] <pitti> but I have no idea what generates them
[09:51] <mdz> pitti: it is the hal postinst
[09:51] <mdz> the do_udev function
[09:52] <pitti> aha
[09:52] <mdz> that entire function should be replaced with a call to udevstart, I think
[09:53] <pitti> but events without a sequence number should just be executed immediately
[09:53] <pitti> so I don't see where there could be a feedback loop
[09:53] <pitti> why udevstart?
[09:54] <pitti> can udevstart be limited to block devices?
[09:55] <mdz> no, i don't think so
[09:55] <mdz> but why does it need to be?
[09:58] <pitti> mdz: well, just to avoid some redundancy, I guess
[09:58] <pitti> mdz: udevstart takes a considerable amount of time
[09:58] <mdz> pitti: dpkg-reconfigure hal reproduces the problem for me
[09:58] <mdz> pitti: modifying postinst to call udevstart fixes it
[09:58] <pitti> mdz: does it help to replace do_udev by udevstart?
[09:58] <pitti> ok
[09:59] <pitti> great
[09:59] <mdz> it is strange that the maintainer did this
[09:59] <mdz> it is hal's own hotplug hook which fails if SEQNUM is not set
[09:59] <mdz> I do not know what causes the loop, though
[09:59] <mdz> pitti: I think that udev should have an update-udev script
[09:59] <pitti> sjoerd: ^ any idea?
[10:00] <mdz> pitti: so that packages which install a udev config file can call it, to have the changes effected immediately
[10:00] <pitti> sjoerd: (about do_udev without sequence number)
[10:00] <pitti> mdz: hmm, good idea
[10:00] <pitti> mdz: however, what should this script do apart from udevstart?
[10:00] <pitti> just wrap it, in case it changes in the future?
[10:00] <sjoerd> pitti: shouldn't be a problem.. we can't fake sequence numbers from usersspace anyway
[10:01] <pitti> sjoerd: <mdz> it is hal's own hotplug hook which fails if SEQNUM is not set
[10:01] <pitti> sjoerd: maybe that is what is causing the feedback loop
[10:01] <sjoerd> hrm.. is that run on an udevsend..
[10:01] <sjoerd> oh.. does hoary have udev in hotplug mode ?
[10:02] <pitti> sjoerd: I think so
[10:02] <pitti> sjoerd: /etc/hotplug.d/default/10-udev.hotplug -> /sbin/udevsend
[10:02] <sjoerd> that's the same as in debian
[10:02] <mdz> pitti: yes, I think it would only call udevstart at this time
[10:03] <sjoerd> anyway, there is no problem if the hal program fails there
[10:03] <mdz> sjoerd: Ubuntu uses /sbin/udevsend as the hotplug helper
[10:03] <sjoerd> mdz: ah, right, so your udevsend actually causes the hotplug scripts to be run
[10:03] <sjoerd> which isn't the case in debian...
[10:03] <mdz> correct
[10:03] <mdz> I don't know why this results in a loop
[10:04] <pitti> hal -> udevsend -> this causes a hotplug event -> calls udevsend again?
[10:04] <sjoerd> the symlink pitti has is strange then ? 
[10:04] <smurfix> Yeah, something like that seems to happen.
[10:04] <mdz> no, the symlink is standard
[10:04] <mdz> udev has some loop detection which should prevent that
[10:04] <sjoerd> ah
[10:04] <mdz> otherwise this would happen with every hotplug event
[10:04] <pitti> smurfix: does replacing do_udev() with udevstart help for you, too?
[10:05] <pitti> mdz: right, but maybe because it does not have a sequence number (no kernel event), the loop detection fails?
[10:05] <smurfix> pitti: No idea at the moment
[10:05] <mdz> pitti: it's possible; I don't know how the loop detection works
[10:05] <pitti> mdz: it might rely on the same sequence number
[10:06] <sjoerd> guess there is no, way to tell udev to just recheck the permissions and naming and not call the hotplug scripts..
[10:06] <pitti> mdz: the only odd thing is, why I can't reproduce it
[10:06] <mdz>                 /* prevent loops in the scripts we execute */
[10:06] <mdz>                 if (strncmp(key, "UDEVD_EVENT=", 12) == 0) {
[10:06] <mdz>                         dbg("seems that the event source is not the kernel, just exit");
[10:06] <mdz>                         goto exit;
[10:06] <mdz>                 }
[10:08] <mdz> the strange thing is, if I run the do_udev function from a shell, it doesn't cause a problem for me
[10:08] <pitti> mdz: I pinged Md in #d-devel, I will ask him for update-udev
[10:08] <mdz> but it does when run from the postinst
[10:09] <mdz> it seems like a race
[10:09] <pitti> mdz: in #d-devel
[10:09] <mdz> perhaps related to the fact that it starts many udevsend processes in the background, and then restarts dbus
[10:09] <mdz> pitti: ok
[10:10] <mdz> why does it run them in the background, anyway?
[10:11] <sjoerd> no good reason that i remember
[10:12] <mdz> hmm, I can't reproduce anymore
[10:12] <mdz> after undoing my workaround
[10:13] <pitti> Heisenbug...
[10:13] <jdub> http://bugzilla.ubuntu.com/show_bug.cgi?id=4271
[10:13] <jdub> ^ i've been pinged about this a couple of times now
[10:14] <mdz> jdub: ...
[10:16] <mdz> jdub: it doesn't happen to me, so if you can reproduce it, you're in the best position to debug it
[10:20] <pitti> mdz: according to Md, running udevstart might have unintended side effects, so we should not do that in a postisnt
 currently your best option is running udevstart
[10:21] <jdub> mdz: i can't, have never used en_US
[10:21] <mdz> I certainly do
[10:21] <mdz> and if it were broken, I would have noticed and fixed it before Warty released
[10:22] <pitti> mdz: okay, I'll modify hal to use udevstart for now
[10:23] <pitti> mdz: after Md finds a better solution for udevstart, we can switch over to update-udev
[10:23] <mdz> pitti: or can we disable that code altogether?
[10:23] <pitti> mdz: well, it would cause hal not being able to read CD-ROMs etc.
[10:24] <pitti> mdz: wrong, not CD-ROms
[10:24] <pitti> mdz: but already existing USB devices
[10:24] <mdz> why?
[10:24] <pitti> mdz: hal isntalls udev rules to change group disk to group hal for USB devices
[10:24] <pitti> mdz: s/USB/removable/
[10:24] <pitti> mdz: so hal does not need to run in group disk
[10:24] <mdz> can we not do that in the udev package, instead of hal?
[10:24] <mdz> (using a different group)
[10:24] <pitti> mdz: so it can mess with removable devices, but not with hard disk ones
[10:25] <pitti> mdz: in the very first packages we indeed did this in udev proper
[10:25] <pitti> mdz: but after some discussion this was moved  to hal, by installign an udev rules script
[10:26] <mdz> I think that udev is the right place
[10:26] <mdz> removable devices should have different permissions regardless of whether hal is installed
[10:26] <pitti> mdz: to get rid of do_udev(), we need to move /etc/udev/rules.d/z_hal-plugdev.rules from hal to udev
[10:26] <pitti> mdz: that's easy to achieve
[10:26] <jdub> 08:24 < luis> jdub: 'install hoary CDs, choose en_US as locale'
[10:26] <jdub> 08:24 < luis> jdub: 'use liveCDs, choose en_US as locale'
[10:26] <jdub> 08:25  * luis is booting a liveCD at this very moment, will test again
[10:26] <mdz> changing permissions of devices based on whether a certain package is installed, seems fairly evil
[10:27] <mdz> jdub: seriously, I do that about 20 times per week, easily
[10:27] <pitti> mdz: okay, then I do that
[10:27] <jdub> mdz: i don't doubt you
[10:27] <jdub> mdz: just relaying
[10:27] <mdz> jdub: in which program is he seeing the problem exhibited?
[10:27] <jdub> gnome stuff
[10:28] <jdub> as noted in the bug
[10:28] <pitti> sjoerd: do you think this can be done in Debian, too? (setting device permissions in udev instead of hal)
[10:28] <mdz> I don't know which gnome stuff has different strings for en_GB vs. en_US
[10:28] <jdub> hrm, now that i look at /etc/environment on my machine, i get:
[10:28] <jdub> LANGUAGE="en_AU:en_US:en_GB:en"
[10:28] <jdub> and i totally shouldn't have en_US before en_GB ;)
[10:28] <jdub> mdz: nautilus 'wastebasket'
[10:29] <sjoerd> pitti: you should discuss that with md.. 
[10:29] <seb128_> you guys should all use fr_FR :p
[10:30] <sjoerd> pitti: but i don't think there is a generic solution and this one is basically temporary for hal 0.4.x..
[10:31] <mdz> jdub: Trash here
[10:31] <mdz> jdub: just tested with an array-4 live CD
[10:33] <Hwolf> Would it be possible to mark Openoffice.org2 as a subsitute for OO01?
[10:34] <pitti> mdz: hmm, but putting removable devices into the hal group even if hal is not installed seems a bit weird, too
[10:35] <pitti> mdz: we could revert the group back to plugdev
[10:36] <pitti> mdz: however, that would mean that normal users can repartition/reformat removable media
[10:36] <pitti> mdz: i. e. malicious software could circumvent permissions on USB hard disks
[10:39] <sivang> seb128_: hehe
[10:40] <kent> hmm, mono-assemblies-base is giving me problems. Can i solve it by hand, or will it be solved automatic with a future update? i cant seem to be able to remove the conflicting packages, it always wants to do apt-get -f install, but that one dont work, since its a conflict. 
[10:43] <seiya> hi guys
[10:43] <mdz> pitti: I can't think of a realistic attack scenario
[10:43] <seiya> do you know anything about website contest?
[10:46] <pitti> mdz: as I said, normal users can circumvent permissions on USB hard disks
[10:46] <pitti> mdz: the "users" in the sense of a human being can of course do this anyway
[10:47] <pitti> mdz: (by moving it to another computer)
[10:47] <pitti> mdz: but the user processes shouldn't
[10:48] <Hwolf> Guys: With my latest upgrade I lost the ubuntu-themed mouse pionters. Which package is responsible for those?
[10:49] <mdz> Hwolf: https://bugzilla.ubuntu.com/show_bug.cgi?id=6172
[10:49] <Hwolf> mdz, ok, couldn't find it.
[10:50] <Hwolf> btw, am I the only one who thinks a filled trash-applet looks like a clipboard?
[10:51] <tseng> yes
[10:51] <Hwolf> -c +s
[11:04] <seiya> jdub: are you going to show us projects submitted to ubuntu website contest?
[11:16] <bluefoxicy> update-notifier has been insisting that there are 7 updates available for the past 7 hours now
[11:16] <bluefoxicy> after installing them
[11:16] <geppy> The gstreamer-jack package is broken.  Is there anything that I can do to fix it?
[11:17] <tseng> geppy: how about gstreamer0.8-jack ?
[11:17] <tseng> gstreamer-* is 0.6, pretty old
[11:17] <geppy> tseng:  I have 0.8-jack
[11:18] <tseng> is that which is broken?
[11:18] <geppy> tseng:  In my experience, they've both been broken.
[11:19] <mdz> bluefoxicy: thanks for testing it; feel free to add your comments to the bugs already open about such issues
[11:19] <trulux> hi
[11:23] <jdub> geppy: probably worth talking to gstreamer upstream about it, more work has been done on the jack plugin since last release
[11:23] <geppy> jdub: Alright, thanks.
[11:25] <bluefoxicy> mdz:  right now I'm more interested in #5779, which now has a patch to correct the problem.  The patch didn't pop up till about 2 days ago so *shrug* it hasn't reached the maintainers.  No big deal, I'm too busy playing dragon warrior to care :)
[11:26] <bluefoxicy> mdz:  the update notifier bug seems to be #6071, so it's known.  *nod*
[11:27] <seiya_> jdub: hi
[11:30] <trulux> mdz: I've just came back and read the message on features freeze
[11:30] <trulux> mdz: is there anything I could do for getting some new SELinux stuff on Hoary before it gets freezed?
[11:30] <trulux> at least bringing updates to the packages
[11:30] <robertj> hey mdz, I've been making an effort to get my act together and actually learn enough to provide a patch to create an admin group for sudoers and have made some progress but am having difficulty testing my changes
[11:31] <robertj> mdz: sudoers is fine but the passwd package does its action during config and it's already configured...
[11:31] <robertj> running sh passwd.config throws up errors about templates and the like so I assume it's not something your supposed to do
[11:32] <bluefoxicy> o.o
[11:32] <bluefoxicy> what group does Squid run as
[11:32] <bluefoxicy> I need it running in its own group
[11:33] <bluefoxicy> ahh, cache_effective_group
[11:36] <sivang> Night everybody
[11:37] <robertj> should I take this one to d-d?
[11:37] <robertj> are they still pretty hospitable ;)
[11:38] <Duck_busy> sjoerd: coin ?
[11:39] <sjoerd> Duck_busy: pong
[11:39] <Duck_busy> sjoerd: :-)
[11:39] <Duck_busy> sjoerd: i want to talk about the python transition in hoary
[11:40] <Duck_busy> people r complaining the depends r wrong
[11:40] <Duck_busy> because debian pkg depends on python 2.3 and ubuntu now uses 2.4
[11:40] <Duck_busy> do you plan some mass rebuilding ?
[11:41] <sjoerd> Duck_busy: you need to ask the ubuntu people about this, not me
[11:41] <Duck_busy> ho, i though you were one of them
[11:41] <mdz> trulux: is there a list of what needs to be done?
[11:42] <jdub> seiya_: yes, there will be a poll soon
[11:42] <mdz> trulux: if you make a list in the wiki, we can discuss it, but it is very close to the freeze
[11:42] <sjoerd> Duck_busy: nope ;)
[11:42] <mdz> robertj: ubuntu-devel@lists is the place to discuss it
[11:42] <Duck_busy> sjoerd: who could i ask ?
[11:42] <robertj> mdz; thanks
[11:43] <Duck_busy> there's not ubuntu tag on people
[11:43] <Duck_busy> -t
[11:44] <seiya_> jdub:oh, pool - ok, thx for answer
[11:44] <sjoerd> Duck_busy: this channel seems like a nice target....
[11:45] <jdub> seiya_: pool! :)
[11:45] <seiya_> jdub: and soon means few days?
[11:45] <jdub> seiya_: oh, most likely less
[11:46] <seiya_> jdub: he he right poll :)
[11:46] <trulux> mdz: sure, I have a wiki page in hardened debian wiki, lemme dig for the link
[11:47] <crimsun> Duck_busy: most of those packages need at least a bump in debian/control:Build-Depends. I believe mdz talked about this issue a couple weeks ago, though I don't have a log of it.
[11:49] <Duck_busy> crimsun: yes, depends must be changed and build-depends too, is there a way to be sponsored as a debian maintainer, so as to have my pkg fixed quicker ?
[11:50] <mdz> Duck_busy: yes, these packages will be fixed in due course
[11:50] <crimsun> for my part, I've been sending diffs to MOTU for the python packages I've run across
[11:52] <mdz> crimsun: that's great; if a few more people would do a few packages each, it would be done very quickly
[11:52] <Duck_busy> mdz: can i come back here with fixed pkg to upload ? (understand ubuntu people r complaining to me and not you, as they see my name as the maintainer)
[11:52] <jdub> oh man
[11:52] <jdub> local lug list
[11:52] <jdub> bunch of people being banged up the sh*tter by consolehelper
[11:53] <jdub> most are completely confused
[11:53] <jdub> (multiple users logging in, exclusive permissions, etc)
[11:53] <jdub> one blames it entirely on gnome
[11:53] <mdz> Duck_busy: yes, definitely
[11:53] <jdub> eeeek
[11:53] <Duck_busy> mdz: thanks
[11:53] <Duck_busy> crimsun: thanks too
[11:53] <mdz> Duck_busy: or you can mail ubuntu-devel@lists.ubuntu.com if no one is here to upload
[11:54] <crimsun> Duck_busy: np
[11:54] <Duck_busy> mdz: roger
[12:01] <mdz> jdub: what is consolehelper?