[00:24] <smallfoot-> i want 2.6.25 plz
[00:25] <bryce> smallfoot-: you can download it from http://kernel.org/
[00:26] <Forgery> hey, is anybody here?
[00:28] <smallfoot-> bryce, then i have todo mad stuff like compile it
[00:28] <smallfoot-> Forgery, yes plz
[00:28] <smallfoot-> bryce,  cant someone compile it, make a .deb and put it in a repo
[00:28] <Forgery> hey
[00:28] <Forgery> i'm installing the new 8.04 and yea i'm a complete newb to this
[00:29] <Forgery> i'm installing it on an old box so it isnt the fastest thing in the world but it could run XP ok so it's not bad
[00:29] <Forgery> I get to the screen when I can see the bird as the wallpaper
[00:29] <Forgery> but there are no icons on the desktop
[00:29] <Forgery> the green cd light is flashing away but nothing is happening
[00:29] <Forgery> is this meant to happen?
[00:30] <laga> Forgery: how much memory does it have?
[00:30] <Forgery> erm
[00:30] <Forgery> 256
[00:30] <Forgery> i think
[00:30] <bryce> smallfoot-: 2.6.25 was just recently released and is not stable, so only people comfortable with compiling it themselves should be using it
[00:30] <Forgery> ooo somethings happening....
[00:31] <laga> Forgery: i'm not sure if 256M is enough
[00:31] <Forgery> ok the screen just turned off
[00:31] <Forgery> really?
[00:31] <Forgery> i thought if it could run XP
[00:31] <bryce> Forgery: I would suggest at least 384M, better with 512
[00:31] <Forgery> then it would have no problem with this
[00:31] <Forgery> ok let me check the memory
[00:31] <Forgery> i think i can still access as windows
[00:32] <Forgery> about to find out anyway
[00:32] <bryce> Forgery: I'd be surprised if you could run XP on 256M without performance issues
[00:32] <laga> Forgery: try the alternate disk maybe, that should install better
[00:32] <bryce> but I haven't used windows in a long time
[00:32] <bryce> Forgery: you may find Xubuntu runs easier on 256m
[00:32] <Forgery> ok
[00:33] <Forgery> if it's a case of getting a little extra memory then that isnt expensive really
[00:34] <Forgery> haha!
[00:34] <Forgery> 128 :$
[00:34] <jcastro> Forgery: note that ubuntu doesn't have icons on the desktop by default, so unless you have a usb key plugged in or something then not having icons on the desktop is normal
[00:34] <Forgery> i know but there was nothing on the screen, no task bar or anything
[00:34] <jcastro> oh oh
[00:34] <Forgery> although the 128m memory might explain it! haha
[00:35] <bryce> 128m??
[00:35] <Forgery> 128mb
[00:35] <bryce> ouch, yeah that'd be your problem ;-)
[00:35] <Forgery> haha that's funny
[00:36] <Forgery> ok anoother question i'd like to ask although it isnt specific to linux
[00:36] <bryce> also, like laga said, the LiveCD loads everything into memory, so on such a low memory system you should use the alternate cd
[00:36] <Forgery> is what's the best way to find if I have DDR or DDR2?
[00:36] <Forgery> i can only assume it's DDR on my old box, simply because its old
[00:36] <Forgery> but it's something I don't know how to dfind out
[00:37] <Forgery> is the best way to find the make of the motherboard?
[00:38] <bryce> yeah refer to your mboard manual
[00:38] <Forgery> so where do you find the name of your motherboard on it
[00:38] <bryce> Forgery: also please be aware this channel is more for development and not support, so you may want to ask these q's on #ubuntu
[00:38] <bryce> (see topic)
[00:39] <bryce> Forgery, sudo lshw
[00:39] <Forgery> ok, sorry guys, was just the first one i could find
[00:39] <bryce> no prob
[00:40] <Forgery> i'll go off investigate
[00:40] <Forgery> thank you for your help guys! much appreciated
[00:40] <Forgery> bye
[00:45] <FrankH> hi. what is the new package name for vim-full in 8.04?
[00:45] <ion_> apt-cache show vim-full
[00:47] <FrankH> tried that. doesn't return anything
[00:47] <FrankH> suggest vim-gnome
[00:47] <ion_> You might not have universe enabled. Also, please see the topic.
[00:49] <FrankH> universe is enabled. sorry if this is not a devel question.
[00:49] <FrankH> hoped that someone in here had the same issue
[01:47] <smallfoot-> when does 8.10 ibex start? when is the first alpha? ;)
[01:56] <nxvl> elmo: ping
[02:54]  * Twigathy waves
[02:56] <Twigathy> Can I request a version bump on a package? The initramfs-tools package has a bug which means dhcp-based nfsroot booting does not work. See debian bug#39514 and ubuntu launchpad bug# 221613. All that is required to fix is to bump from 0.85eubuntu36 to something 0.88 flavoured (Where the bug is allegedly fixed...)
[02:57] <CheGuevara> hardy has just been released
[02:57] <Twigathy> I'm working around this atm by specifying IP address / netmask etc. manually but it'd be nice if dhcp would work :)
[02:57] <CheGuevara> and hardy+1 not started yet
[02:57] <CheGuevara> a bit of bad timing :P
[02:57] <Twigathy> yeah, I know :)
[02:57] <CheGuevara> file a bug
[02:57] <Twigathy> Maybe I should become a beta tester - using weird things like nfsroot..
[02:58] <Twigathy> yep - have done so
[02:58] <niadh> Just wanna report a bug here, regarding hal not being restarted upon upgrade to hardy, spent hours working out a problem with banshee and ipods, only to discover hal needed to be restarted manually, I'm told I shouldn't have had to do this, so figured i would report it
[02:59] <Twigathy> niadh: https://bugs.launchpad.net :)
[02:59] <niadh> I may need some help about specifics to write, but I'll definatly submit it
[03:02] <wgrant> niadh: You're meant to reboot after upgrades....
[03:02] <wgrant> niadh: But I thought it was restarted unless you were coming from Dapper.
[03:02] <niadh> I DID reboot after the upgrade
[03:02] <niadh> And i upgraded from 7.10
[03:03] <wgrant> So you had to restart hal twice?
[03:03] <niadh> Just updated my system to all the latest patches via update manager, THEN ran the upgrade procedure. Yeah, system rebooted and then I had to manually restart hal
[03:07] <niadh> just to confirm the upgrade procedure DOES restart hal?
[05:09] <jdong> -rw-r--r--  1 jdong jdong  30M 2008-04-24 23:56 urlclassifier3.sqlite
[05:09] <jdong> asac: ^^ err I think the URL classifier thing is a security vulnerability. It's slowing down my browser's performance :D
[05:10] <jdong> asac: but more seriously, it's adding serious latency and CPU spinning to opening every URL. The file regenerates to around that size every few days.
[05:17] <Amaranth> jdong: what is it?
[05:17] <jdong> Amaranth: I think it's the phishing detector thing
[05:17] <jdong> Amaranth: but it grows at some annoying rate of 5MB/day
[05:18] <Amaranth> right, i'll disable that then
[05:18] <Amaranth> could this be why firefox randomly locks up and beats the crap out of my HD?
[05:20] <jdong> Amaranth: exactly
[05:20] <jdong> Amaranth: it's sqlite journaling activity
[05:20] <jdong> Amaranth: which enforces a sync()
[05:20] <jdong> on *every* *page* *load*
[05:26] <jcastro> jdong: asac is aware, as is upstream
[05:26] <jcastro> upstream is working on a fix
[05:27] <jcastro> it's bug 215728
[05:27] <ubotu> Launchpad bug 215728 in firefox-3.0 "High CPU Consumption" [Undecided,Confirmed] https://launchpad.net/bugs/215728
[05:33] <tjaalton> Caesar: you should be able to find it on launchpad
[05:34] <pwnguin> i wonder if this affects liferea
[05:34] <pwnguin> i think it uses webkit or something to render
[05:35] <wgrant> liferea uses xulrunner-1.9, pwnguin...
[05:40] <jdong> jcastro: ok, thank you!
[05:40] <warp10> Good morning!
[05:41] <pwnguin> wgrant: i dont know a whole lot about the liferea source code, but I do see this: http://liferea.svn.sourceforge.net/viewvc/liferea/trunk/liferea/src/webkit/
[05:42] <jcastro> jdong: the current plan is for mozilla to attach a patch to an lp bug, then we'll roll it out via ppa. If you could subscribe to the bug and help test when that's available it would help a great deal
[05:42] <wgrant> pwnguin: We use the xulrunner-1.9 backend.
[05:43] <pwnguin> ah, multiple backends, i see
[05:43] <jdong> jcastro: I'd be more than happy to :)
[05:43] <jcastro> jdong: <3
[06:22] <dholbach> good morning
[06:49] <pwnguin> oh dear, this has gotten way out of hand: http://np237.livejournal.com/17716.html
[06:50] <pitti> Good morning
[06:51] <dholbach> hi pitti
[06:55] <geser> Guten Morgen pitti
[07:21] <tjaalton> help.ubuntu.com still has 7.10 material
[07:21] <tjaalton> um, I mean no 8.04
[08:43] <mdke> tjaalton: yes, it will take at least a few days to get it ready
[08:44] <mdke> tjaalton: but if you file a bug report on ubuntu-doc you can follow its progress
[08:44]  * mdke gtg
[08:47] <tjaalton> mdke: ok, no rush :)
[08:53] <edugonch>  hello, do somebody have problems with Anjuta and glade, I start edit a widget but when I close Anjuta and then I open again and choose glade IDE Anjuta doesn't open
[08:55] <asac> jdong: do you have your .mozilla dir on a special filesystem?
[09:06] <james_w> I've got a couple of bugs to propose for SRU to Hardy. Do we have to wait for them to be fixed in Intrepid as normal, or can we propose them now? The one I have in mind will be fixed in Intrepid in the first run of the auto-sync.
[09:06] <pitti> james_w: that's fine
[09:06] <pitti> james_w: i. e. feel free to open SRU bugs now, we already did a few
[09:07] <james_w> great, thanks.
[09:07] <james_w> pitti: oh, and thanks for getting back on the ca-certificates SRU, you rock!
[09:07] <pitti> james_w: no problem; sorry for the latency
[09:08] <james_w> it's understandable.
[09:13] <fabbione> hey guys
[09:13] <pitti> hey fabbione, how are you?
[09:13] <fabbione> pitti: good thanks and you?=
[09:14] <pitti> I'm intrepid today :)
[09:14] <fabbione> pitti: eheheh..
[09:14] <fabbione> did they already opened the repo in lp?
[09:14] <pitti> no, not yet
[09:14] <fabbione> ok
[10:37] <seb128> hum, bug #215499
[10:37] <ubotu> Launchpad bug 215499 in nautilus "Nautilus not preserving timestamps" [Undecided,New] https://launchpad.net/bugs/215499
[10:38] <seb128> does anybody has an opinion if a standard copy should preserve timestamps or not?
[10:38] <seb128> I would tend to say that since cp doesn't do it by default nautilus shouldn't
[10:38] <cjwatson> sounds like cp vs. cp -a to me
[10:38] <seb128> right
[10:39] <seb128> but cp -a is not the default
[10:39] <seb128> I'm not sure it should be in nautilus either
[10:39] <seb128> but I'm no good argument out of "that's the standard unix way"
[10:39] <seb128> s/I'm/I've
[10:40] <cjwatson> seems like it'd be nice to have a way to do cp -a, though; timestamps are valuable information you might want to preserve in a backup
[10:40] <cjwatson> even if it isn't the default
[10:40] <soren> "can't fully use my laptop because of this"... Yay, proportions.
[10:41] <seb128> right
[10:41] <cjwatson> I don't have a strong opinion on whether it should be the default or not, although I know that I have a reflex of using cp -a unless I explicitly think "hmm, I don't need the timestamp"
[10:41] <seb128> I'll change that to a wishlsit to add an option doing that
[10:41] <seb128> I don't think changing the default makes sense
[10:41] <cjwatson> bug report says "whether cross-partition or the same partition" - is the behaviour the same for both cases at the moment?
[10:41] <seb128> or maybe a gconf key to change the default
[10:42] <realist> Perhaps a checkbox, that says "preserve timestamps/permissions"?
[10:42] <seb128> I've to check if cross partition behaves different
[10:42] <seb128> realist: where?
[10:43] <seb128> realist: you usually get no dialog when doing a dnd or ctrl-C, ctrl-V
[10:43] <realist> seb128: I don't know, I don't use nautilus. In a preferences dialog?
[10:43] <seb128> ok, an option then
[10:44] <soren> Where does LANG and LC_* get set when I log in on the console?
[10:50] <geser> soren: doesn't pam_env.so take care of it?
[10:51] <soren> geser: Oh, there it is. Yes, it does.
[10:51] <soren> geser: Thanks!
[11:15] <abli> Hi! I am having problems with consolekit on a fresh hardy install. After the default install I installed kubuntu-desktop. Now in a gnome session, running  ck-list-sessions prints "active = FALSE" even for the current session under gnome. As a result automounting usb thumbdrives break because only the active session has permission to do so (by default settings)
[11:16] <abli> any ideas why the active session is not set correctly or what should I check?
[11:16] <Riddell> works for me in KDE
[11:17] <abli> yes, works for me in kde as well. but not in gnome
[11:19] <Riddell> abli: using KDM?
[11:20] <abli> no. gdm
[11:20] <abli> I only installed kubuntu-desktop to get kde as a second option. I want to use gnome as default, and have the possibility of switching to kde
[11:22] <cjwatson> abli: works for me in GNOME, FWIW
[11:22] <cjwatson> ah, though I don't have kubuntu-desktop installed
[11:22] <abli> Yeah, I assume it works for everyone but me. The question is, what should be setting that flag so I can debug why it isn't set.
[11:25] <cjwatson> abli: normally that gets changed by consolekit when the active VT is changed to the one corresponding to the session
[11:26] <abli> cjwatson, ok. does consolekit write logs somewhere?
[11:27] <cjwatson> though I admit I don't know whether something else deals with it when the session is initialised
[11:27] <cjwatson> IME the best way to track what consolekit is doing is to attach a monitor to the D-BUS
[11:28] <cjwatson> it's a bit hairy ...
[11:28] <cjwatson> (p.s. I'm not a consolekit expert, I just know what I needed to figure out in order to implement support for it in openssh)
[11:29] <abli> ok, I'll try that
[11:29] <cjwatson> I suspect that d-bus messages won't be sent for consolekit's internal actions though
[11:30] <cjwatson> abli: I believe you can run the daemon with --debug
[11:30] <abli> ok, I'll try that, too :)
[11:30] <cjwatson> not sure whether you have to do anything special - you'll probably have to close all sessions and shut down the running daemon first
[11:31] <cjwatson> probably --debug --no-daemon too
[11:50] <thom> grrrr.
[11:51] <thom> Apr 25 12:50:13 madrigal kernel: [18896.521532] gnome-keyring-d[5102]: segfault at 7f511b0d5000 rip 42dce0 rsp 46e90d20 error 7
[11:51] <laga> thom: apport?
[11:52] <wgrant> thom: I saw a bug on that an hour or two ago.
[11:53] <thom> laga: unfortunately not
[11:53] <wgrant> laga: apport is off now.
[11:55] <laga> wgrant: i was suggesting that he enables it again to catch the segfault, but it's probably not necessary if there is already a bug report.
[11:55] <wgrant> Hm, maybe it was on the forum. I can't see the relevant bug.
[11:56] <wgrant> Could be bug #218434, I guess.
[11:56] <ubotu> Launchpad bug 218434 in gnome-keyring "gnome-keyring-daemon crashed with SIGSEGV" [Undecided,New] https://launchpad.net/bugs/218434
[11:58] <thom> yeah, pretty hard to tell
[11:58] <thom> i guess i'll just try again
[12:01] <cody-somerville> Why are my posts to ubuntu-devel being held for moderation saying I'm not a developer? :)
[12:01] <cody-somerville> I'm sending from my @ubuntu.com address.
[12:01] <wgrant> cody-somerville: Probably because you need to manually request the sync these days, I guess.
[12:01] <wgrant> I've seen a couple of other people complain about that recently.
[12:02] <cody-somerville> Is elmo the person I need to chat with?
[12:02] <highvoltage> cody-somerville: I think the admins of the list need to add you to a whitelist of some kind first
[12:03] <soren> Could someone let new virtinst into hardy-proposed, please?
[12:07] <seb128> soren: open a bug, subscribe ubuntu-sru, etc
[12:07] <seb128> soren: and wait for pitti to be back from lunch ;-)
[12:07] <soren> seb128: Already did.
[12:07] <soren> Ah.. I did not do that.
[12:07] <soren> :)
[12:07] <soren> (wait for pitti to get back from lunch, that is)
[12:08] <cjwatson> cody-somerville: I wonder if it only looks at direct members of ubuntu-dev - mostly people are members via the motu team these days
[12:08] <canburak> hi, I want to mirror just the latest release of ubuntu. hos is that possible? ~200GB is so much for me
[12:09] <cjwatson> canburak: use debmirror with --dist=hardy
[12:09] <cody-somerville> cjwatson, but the xml file thingy will return an aggregated list... so unless they're screen scrapping...
[12:09] <highvoltage> canburak: why would you want to mirror it, to act as a local mirror?
[12:09] <cjwatson> cody-somerville: I don't know - #canonical-sysadmin would be a good place to ask
[12:10] <cjwatson> highvoltage: local mirrors are very common and I hardly see the need to ask why :-)
[12:10]  * cjwatson has one himself
[12:10]  * highvoltage too :)
[12:10] <highvoltage> cjwatson: but the reason is important though, if you only want it for one arch being used in your office, you don't need to download the others
[12:11] <highvoltage> cjwatson: and he probably doesn't need to download the source packages either
[12:11] <cjwatson> true, best to read the debmirror manual page
[12:11] <highvoltage> *nod* canburak ^^^
[12:13] <canburak> highvoltage: no to act as a provate mirror. currently I don't have the required space for a local mirror
[12:13] <highvoltage> canburak: ah, then you probably just need i386 (or 64bit) binaries for main universe multiverse and restricted, which would be 20GB or so
[12:14] <canburak> i once had a country mirror for TR and nuked when i was in bad mood :)
[12:14] <highvoltage> canburak: but check the debmirror manpage, it's probably your best bet.
[12:14] <canburak> highvoltage: is it possible to use debmirror of the debian installation to mirror ubuntu?
[12:15] <highvoltage> canburak: yes, you might just have to use the newest available debmirror that already knows about hardy
[12:15] <canburak> highvoltage: my mirror is on etch
[12:16] <canburak> and you suggest to use (if available) debmirror from backports?
[12:16] <canburak> and you suggest to use (if available) debmirror from debian-backports?
[12:16] <wgrant> debmirror doesn't have to know about the suite, does it?
[12:34] <cjwatson> canburak: pretty much any version of debmirror should do just fine; highvoltage is mistaken here, it doesn't have to specifically know about hardy
[12:34] <cjwatson> I'm using the version of debmirror in Debian stable to mirror hardy, with no problems
[12:35] <highvoltage> hmm, yes I am. I confused debmirror with debootstrap a bit, I think.
[12:36] <cjwatson> debootstrap does have to be up to date, yes
[12:36] <cjwatson> (when one is using it)
[12:36] <canburak> hmm thanks i'll give a try
[13:54] <pitti> soren: what's up?
[13:54] <pitti> soren: ah, virtinst; I'll deal with it
[13:58] <soren> pitti: Thanks very much.
[14:34] <evand> cody-somerville: I'll tkae care of it.  Sorry, your email got buried in my inbox.
[14:38] <evand> oh, that was Scott Ritchie
[14:38] <highvoltage> who was?
[14:39] <evand> Scott emailed me saying that his emails were being put in the moderation queue.
[14:41] <cody-somerville> evand, mine too :P
[14:41] <evand> cody-somerville: hrm, better check with the sysadmins, I'm not sure how that bit is set up as no one on the list has the moderation bit set.
[14:42] <cody-somerville> It says I'm not a developer
[14:46] <evand> cody-somerville: I've asked in #canonical-sysadmin as this is beyond my control.
[14:46] <pecisk> hi guys, in our locale, in last minute changes, some update in xlrunner b0rked security exception dialog, therefore lots of people can't log in on internet via wifi. It is known issue for other languages/more general issue?
[14:46] <sistpoty|work> can an archive admin please shove libitpp through hardy-proposed binary new? thanks.
[14:46] <pitti> sistpoty|work: ugh, NEW for stables?
[14:46] <sistpoty|work> pitti: yeah, it's a leftover from the gfortran transition... but fortunately it doesn't have rdepends
[14:49] <ScottK> Unfortunately I didn't find the missed spot in the transition until about 6 hours after the archive closed
[14:49]  * sistpoty|work didn't find it at all
[14:50] <ScottK> pitti: Would you be up for some archive work on clamav updates/backports?
[14:52] <pitti> sistpoty|work: ok, I'll have a look later; I'd like to finish my current hack job first, to not loose so much brain state
[14:52] <sistpoty|work> pitti: sure, no hurries ;)
[15:01] <evand> YokoZar: Can you join #canonical-sysadmin?
[15:04] <cjwatson> evand: it's done in the sender filters, but people in ubuntu-dev ought to be processed automatically without needing to be in there by hand
[15:04] <cjwatson> so adding them by hand will make it work short-term but isn't really the right answer long-term
[15:05] <cjwatson> evand: (privacy options -> sender filters -> accept_these_nonmembers is what we use for non-developer autoaccepts)
[15:05] <evand> cjwatson: ah
[15:06] <evand> cjwatson: I've since taken it to #canonical-sysadmin to see if there's a problem with the processing code.  Spads is looking into it.
[15:35] <edugonch_> Hello, I have install code:blocks but I have problems when I try to compile a project, an error with a file wx/setup.h that not is found
[15:35] <edugonch_> any help please
[15:37] <Kim^J> edugonch_: Install wxwidgets.
[15:38] <Kim^J> Aslo, this is not the channel for such a question I think.
[15:39] <sistpoty|work> (-dev to be more precise... headers will always be in the -dev packages)
[15:39] <Kim^J> sistpoty|work: Ah correct... Too used to Arch now, which installs it all.
[16:17] <cjwatson> mvo: how are the Translation-* files updated? I notice that their mtimes are 22 Feb
[16:17] <cjwatson> which seems a bit old for hardy, really
[16:18] <mvo> cjwatson: there were issues with rosetta that made updates difficult - I can manually work around those, its more effort. I will prepare a update now (its a special upload handler)
[16:19] <mvo> cjwatson: I discussed a spec how the translations can/should be improved within rosetta/LP but I is not implemented yet (or even assigned)
[16:19] <cjwatson> I'm wondering whether we can/should push that into hardy directly, rather than updates
[16:20] <mvo> hm, not sure if that is possible
[16:20] <cjwatson> I don't approve of updating Packages post-release, obviously, but this doesn't seem like it would cause practical problems
[16:20] <cjwatson> if you upload to hardy directly, it ought to get held for approval
[16:21] <mvo> we may push them into -proposed first, I don't think there will be problems, its probably best to test them there first
[16:21] <mvo> but for .1 we should have them in main hardy IMHO
[16:22] <cjwatson> I very much doubt that the usual copy-out-of-proposed trick will work, so they'll need to be reuploaded
[16:22] <mvo> that should be ok I think
[16:22] <cjwatson> my only concern is that the Rosetta export should happen ASAP so that it's close to what real hardy ought to have been
[16:23] <mvo> ok, I request the export now
[16:23] <cjwatson> mvo: what's the translation component in Launchpad called, just for my curiosity?
[16:24] <mvo> cjwatson: translated-package-descriptions is the name of Source: and Binary: in the changes file
[16:25] <cjwatson> https://translations.edge.launchpad.net/ubuntu/hardy/+source/translated-package-descriptions doesn't exist though
[16:26] <cjwatson> just wondering where to direct people who want to translate it
[16:26] <mvo> cjwatson: oh, sorry - I misunderstood the question. its called ddtp-ubuntu
[16:26] <mvo> https://edge.launchpad.net/ddtp-ubuntu
[16:27] <mvo> I keep syncing those with debian
[16:27] <cjwatson> ah, right, thanks
[16:27] <cjwatson> syncing those must be a horrible job
[16:28] <mvo> yeah, the scripts do most of the work, but its still time consuming
[16:28] <Hobbsee> mvo: you screwed up!
[16:34] <tseliot> mvo: any ideas as to why this happens? bug #221977
[16:34] <ubotu> Launchpad bug 221977 in gnome-control-center "Cannot enable compiz with the fglrx driver (ATI X1600)" [Undecided,New] https://launchpad.net/bugs/221977
[16:35] <mvo> tseliot: that is a mobile ati one? that might be a (unintended) side-effect of our mobile ati test :/
[16:35] <tseliot> no, a standalone card
[16:36] <tseliot> ﻿mvo: Compiz works well if launched from the command line
[16:36] <tseliot> ﻿mvo: my card is an ATI X1600 Pro 256MB
[16:36] <mvo> tseliot: could you please check what .xsession-errors says when you enable it?
[16:37] <mvo> tseliot: strange that it works fine if you run it on the commandline
[16:37] <mvo> it should be just the same
[16:37] <tseliot> mvo: sure, let me check
[16:39] <ramvi> ﻿I have an openPGP key on Launchpad. I'm trying to build a deb, but I'm not allowed because I don't have an gpg key (fresh ubuntu install). How do I activate my old gpg key in Ubuntu?
[16:40] <mjg59> ramvi: If you no longer have the secret key, you don't
[16:40] <ramvi> I have the secret key
[16:40] <mjg59> Then use it
[16:40] <ramvi> Right, how? I can only find how to make a new pgp thing in the wiki
[16:40] <mjg59> Just copy the keyring into your new install
[16:41] <mjg59> It's in the .gnupg directory
[16:41] <Amaranth> ramvi: You're supposed to keep the key
[16:41] <cjwatson> ramvi: are you just building the .deb for personal use?
[16:41] <Amaranth> Guard it with your life and never delete it
[16:41] <ramvi> cjwatson: no
[16:41] <cjwatson> ah
[16:41] <ramvi> Amaranth: Oh, I have deleted it..
[16:41] <tseliot> mvo: here it is: http://pastebin.com/m1e3a1f55
[16:42] <Amaranth> ramvi: Then you don't have a key
[16:42] <ramvi> Amaranth: Didn't know I was suppost to keep it. So I have to create a new one? Can't download from the keyserver?
[16:42] <ramvi> Amaranth: right, stupid me
[16:42] <Amaranth> The keyserver only has the public key
[16:42] <mjg59> ramvi: The secret half of the key never goes anywhere near the keyserver
[16:42] <mjg59> That's why it's secret
[16:43] <ramvi> :)
[16:43] <ramvi> thanks
[16:43] <mjg59> (please don't upload the secret part of your key anywhere at any point!)
[16:44] <mvo> tseliot: right, that is a bug in "jockey" (There is no available graphics driver for your system which supports the composite extension.) - that message comes from jockey
[16:44] <mvo> ^--- pitti could you please have a look?
[16:47] <tseliot> mvo: ok, I have edited the bugreport so that it affects jockey
[16:47] <pitti> mvo: right, I already talked with tseliot about it; I'm still not sure which exit codes the applet expects
[16:48] <pitti> mvo: ATM it exits with 0 iff it installed a compositing-enabling driver
[16:48] <d1nker_> has anyone fixed the smb:// and network:// bug yet?
[16:48] <pitti> mvo: i. e. it exits 1 if installation failed or there is no driver, or you don't need one
[16:48] <pitti> mvo: I guess we need more exit codes then
[16:48] <pitti> mvo: jockey currently does not know about all possible X drivers and whether they already support composite
[16:49] <pitti> mvo: it only knows composite support for the drivers it can handle (nvidia, fglrx)
[16:49] <pitti> mvo: I guess we need to redefine the exit code protocol
[16:49] <slangasek> d1nker_: that's a bit imprecise; do you have a bug number for the issue you're referring to?
[16:49] <slangasek> d1nker_: or can you explain what the bug is that you're talking about?
[16:50] <d1nker_> https://bugs.launchpad.net/ubuntu/+source/gvfs/+bug/185756
[16:50] <ubotu> Launchpad bug 185756 in gvfs "Couldn't display "network:///" - Nautilus cannot handle network:locations" [Medium,Fix released]
[16:50] <mvo> pitti: can/should it not just do the same as it does when nvidia is already installed/enabled on the system?
[16:50] <Amaranth> d1nker_: They work fine here
[16:50] <slangasek> d1nker_: well, yes, that bug was fixed some time ago
[16:50] <seb128> d1nker_: that bug has been fixed some months ago as indicated
[16:50] <pitti> mvo: so far I thought the applet would only call jockey if the current driver would not support composite
[16:50] <d1nker_> how do I apply the patch?
[16:50] <pitti> mvo: I didn't know it would *always* call jockey before enabling
[16:51] <Amaranth> d1nker_: The bug did not affect 7.10 and is fixed for 8.04
[16:51] <pitti> mvo: I think it does
[16:51] <slangasek> d1nker_: all that's required is to log out and log back in after upgrading from the older version of pre-release hardy to the official Ubuntu 8.04
[16:51] <d1nker_> I still have the exact same problem as described with 8.04 (fresh install this morning) should I reopen the bug
[16:51] <slangasek> d1nker_: if network:/// still doesn't work for you, then you have some other bug and we'd need more information about your symptoms
[16:52] <slangasek> d1nker_: if you freshly installed 8.04, then no, you don't have that bug
[16:52] <Amaranth> In a bug report, preferably
[16:52] <mvo> pitti: I check the code out, we should milestone it for .1
[16:52] <pitti> mvo: right; compiz itself knows whether the current driver is composite-capable, right? can the applet ask compiz, and only ask jockey to install a driver if the current driver doesn't work?
[16:53] <pitti> mvo: (the option should really be --install-composite, not --check)
[16:54] <mvo> pitti: yeah, the option is pretty unfortunately named - well, its a bit more complicated, compiz does not really know until it is started, this is why we have the wrapper script
[16:54] <pitti> oh
[16:54] <mvo> pitti: we would have to add a --check-only option to the wrapper to get this information
[16:54] <d1nker_> when I go to places - connect to server and select windows share, fill the info out I get an error that says "Can't display location smb://SERVERNAME" No application is registered as handling the file
[16:54] <mvo> pitti: maybe the best solution is to just remove the cal to jockey entirely
[16:54] <pitti> mvo: or, just try to enable it, and if that fails, call jockey?
[16:55] <mvo> pitti: the user knows about the misisng nvidia driver already from the notification bublle
[16:55] <mvo> pitti: hm, yeah, we could do that too (start, if it fails call jockey)
[16:55] <pitti> mvo: so, don't check if compiz would work, just start it and check if it fails?
[16:55] <pitti> if the magic test script would return "thumbs up", you'd just start compiz anyway, right?
[16:55] <seb128> d1nker_: that's a different issue, does places, networks give an error?
[16:56] <slangasek> d1nker_: to clarify, that's the Places -> Network menu in the desktop
[16:56] <pitti> mvo: btw, bug 208026 is also related to that
[16:57] <mvo> pitti: yeah, start it, if that fails, run jockey
[16:57] <ubotu> Launchpad bug 208026 in gnome-control-center "gnome-appearance-properties asks for reboot if you cancel ATI driver install" [Low,Confirmed] https://launchpad.net/bugs/208026
[16:58] <d1nker_> Places -> Network works fine it's when you enter information into the Places -> Connect to Server -> Service Type: Windows Share
[16:58] <pitti> mvo: I think that part does need a separate exit code
[16:58] <mvo> pitti: yeah
[16:58] <pitti> mvo: although, hmm
[16:59] <pitti> mvo: it only asks for reboot if the driver was installed, right?
[16:59] <pitti> mvo: i. .e if jockey returns 0
[16:59] <pitti> mvo: in that case, the bug might just be that jockey returns with 0 if the install was cancelled
[16:59] <mvo> hm, right
[17:00] <slangasek> d1nker_: for shares that I know to work, I'm able to connect without error on 8.04.  Is this a share that requires authentication?
[17:00] <d1nker_> yes
[17:00] <pitti> mvo: right, I just checked the code
[17:00] <pitti> mvo: I'll get that fixed in jockey
[17:01] <d1nker_> which if I access it using the control+L option I can get to
[17:01] <seb128> ok, so it seems to be an issue in the connect to server dialog, could you open a bug on nautilus and describe the steps you are using?
[17:01] <d1nker_> yah can you send me a link, I've never submitted a bug before
[17:02] <pitti> mvo: thanks; I'll summarize in the bug
[17:03] <mvo> pitti: thanks
[17:03] <pitti> mvo: done; I didn't assign the bug yet
[17:03] <slangasek> d1nker_: in this case, https://bugs.launchpad.net/ubuntu/+source/gvfs/+filebug
[17:04] <strider_> hi , where can I find why there is no more multimedia tab in gnome-volume-properties ?
[17:05] <pitti> tseliot: did you already file a bug about the 'envy superseded by lrm' problem?
[17:06] <tseliot> pitti: the one about "-envy" lrms?
[17:06] <d1nker_> thanks I'll submit the bug this afternoon
[17:07] <pitti> tseliot: problem (1) in our chat
[17:07] <tseliot> ﻿pitti: bug ﻿221968
[17:08] <pitti> ah, an upstream bug
[17:09] <tseliot> ﻿pitti: d'oh!
[17:09] <seb128> d1nker_: replaces gvfs by nautilus in the url, that's a nautilus issue
[17:09] <pitti> tseliot: that's fine, don't worry
[17:10] <slangasek> seb128: ah, oops :)
[17:14] <MrDakoki> hi all
[17:14] <MrDakoki> im trying to code an app
[17:14] <MrDakoki> but i have a linker error  for XTest.h
[17:14] <MrDakoki> i cant find the library
[17:14] <slangasek> MrDakoki: see the /topic, please
[17:15] <gnomefreak> mvo: doesnt update-manager have the do-upgrade flag? to upgrade gutsy > hardy?
[17:15] <MrDakoki> ok
[17:16] <mvo> gnomefreak: it will/should pick that information up automatically (i.e. from the internet)
[17:16] <gnomefreak> ah ok thanks
[17:23] <jburd> I think the upgrade tools should resolve configuration conflicts using a GUI diff/merge tool instead of asking for a file level "keep/replace"
[17:23] <jburd> After upgrading I have to make changes in so many configuration files.  Using meld as part of the dist upgrade tool would have saved me that.
[18:02] <duaneb> hello?
[18:02] <duaneb> where can I find gtk-doc.m4?
[18:02] <duaneb> it's missing from /usr/share/aclocal, and I can't find it in any package
[18:04] <duaneb> ahh, gtk-doc-tools
[18:04] <duaneb> ubuntu never ceases to piss me off in its separation of packages... like manpags-dev
[18:04] <duaneb> took me about five months to find that
[18:07] <ion_> Duane Barry? Is that you? :-) Please read the topic.
[18:11] <cjwatson> 18:11 [Freenode] -!-  ircname  : Duane Bailey
[18:11]  * cjwatson hands /whois to ion_
[18:12] <cjwatson> duaneb: manpages-dev has been separated out in Debian for well over a decade, FWIW; not an Ubuntu-specific thing
[18:12] <ion_> cjwatson: That was a joke, an X-Files reference actually.
[18:12] <ion_> Thus the smiley.
[18:12] <duaneb> cjwatson, it's still stupid >.<
[18:12] <jdong> duaneb: it shouldn't take you 5 months to find a file in a package though :)
[18:12] <cjwatson> duaneb: I disagree, and I'm sure we can agree to disagree
[18:12] <duaneb> I mean, it should still be a dependency of build-essential
[18:12] <cjwatson> no, it *really* shouldn't
[18:13] <duaneb> cjwatson, why would you EVER want it separate
[18:13] <cjwatson> build-essential is for a very specific purpose; it is not a general "help me develop stuff" package
[18:13] <cjwatson> I think it's perfectly reasonable to have only manpages installed and not manpages-dev on a system that isn't used for development
[18:13] <duaneb> cjwatson, then there SHOULD be a "help me develop stuff" package
[18:13] <cjwatson> it's a suggestion of gcc
[18:13] <duaneb> cjwatson, but if build-essential is installed...
[18:13] <cjwatson> Suggests: autoconf, automake1.9, bison, flex, gcc-doc, gcc-multilib, gdb, libtool, make, manpages-dev
[18:13] <slangasek> Package: gcc Depends: diveintopython ;)
[18:14] <duaneb> nobody actually looks at that, though :P
[18:14] <cjwatson> build-essential is there to tell you what the build daemons install by default when building packages
[18:14] <cjwatson> that is its purpose
[18:14] <cjwatson> and it's a very useful purpose for package maintainers
[18:14] <duaneb> cjwatson, still, there should be a development bpackage
[18:14] <duaneb> i.e. 'build-tools'
[18:14] <cjwatson> if you want to find packages, I recommend using packages.ubuntu.com
[18:14] <cjwatson> development for whom?
[18:15] <duaneb> that would include the basics for development, like gcc and the manpages
[18:15] <cjwatson> it would end up pulling in half the archive
[18:15] <jdong> build-tools will turn into .*-dev
[18:15] <duaneb> cjwatson, console development
[18:15] <cjwatson> many developers never touch gcc
[18:15] <cjwatson> why is your development more special than others/
[18:15] <cjwatson> ?
[18:15] <duaneb> cjwatson, find, gcc developers
[18:15] <duaneb> cjwatson, because manpages should come with libc-dev (or whatever the package name is).
[18:15] <cjwatson> I think it's a fair question. Why is your development more special than others? We have to consider a pretty wide base of people.
[18:15] <persia> jdong: *-dev self-conflicts :p
[18:15] <duaneb> documentation, especially those manpages, is an integral part of development
[18:16] <cjwatson> it's an integral part of C development
[18:16] <duaneb> cjwatson, because their development is based on c development :P
[18:16] <cjwatson> I do think we need to do better at serving developers
[18:16] <laga> i demand that a standard ML interpreter comes with the proposed build-tools package
[18:16] <cjwatson> but lumping it all together doesn't really help anyone
[18:16] <cjwatson> in the meantime, I recommend packages.ubuntu.com for finding packages
[18:16] <jdong> IMO we need an apt-file like mechanism more readily accessible
[18:16] <jdong> i.e. within Synaptic
[18:16] <cjwatson> or a good package manager that shows suggestions
[18:17] <duaneb> cjwatson, installed, manpages-dev is ~3 megabytes.
[18:17] <duaneb> that's hardly 'lumping it all together'
[18:17] <cjwatson> I see I am not getting anywhere
[18:17] <jdong> duaneb: so's libxmms-dev and python-setuptools.
[18:17] <jdong> let's put those in too
[18:18] <duaneb> jdong, there's no reason to think that c developers would EVER use python
[18:18] <cjwatson> in the meantime, we provide excellent search tools, and I recommend that you use them
[18:18] <jdong> and g++ is too big too, and everything that g++ does can be done by gcc anyway
[18:18] <jdong> let's take out g++.
[18:18] <duaneb> but every c developer uses manpages-dev
[18:18] <cjwatson> there is absolutely no reason it should take you five months to find anything
[18:18] <duaneb> jdong, I wouldn't protest :)
[18:18] <jdong> duaneb: *every* C developer uses manpages-dev?
[18:18] <duaneb> jdong, I would say yes
[18:18] <jdong> interesting.
[18:19] <slangasek> jdong: everything that g++ does except for building C++ code, I guess you mean?
[18:19] <cjwatson> gcc doesn't depend on manpages-dev because gcc is on our CD images and those three megabytes would put us over our limit.
[18:19] <jdong> slangasek: you can rewrite C++ code anyway ;-)
[18:19] <jdong> slangasek: that's a luxury
[18:19] <slangasek> anyway, g++ is part of build-essential because of the definition of build-essential
[18:19] <cjwatson> similarly libc6-dev
[18:20] <cjwatson> (well, that's not the original reason why it doesn't have that dependency, but it's a very good reason not to add it now.)
[18:22] <User633> hello
[18:22] <User633> anyone in here
[18:22] <User633> is this ubutnu irc channel correct
[18:22] <pitti> !ask | User633
[18:22] <ubotu> User633: Please don't ask to ask a question, ask the question (all on ONE line, so others can read and follow it easily). If anyone knows the answer they will most likely answer. :-)
[18:22] <pitti> User633: you probably want #ubuntu
[18:22] <User633> yes
[18:22] <User633> can i have to server for that to join ubunth help room
[18:23] <pitti> User633: just /join #ubuntu, it's the same server
[18:23] <slangasek> same server
[18:23] <User633> oh really thank you
[19:03] <genii> Quick question... is default user created now at install time allowed sudo? A user is telling me he can't issue sudo visudo for instance
[19:11] <jdong> genii: yes, the default user is allowed sudo
[19:11] <jdong> genii: add users to the "admin" group to give them sudo access
[19:12]  * ScottK looks up at /topic
[19:13] <jdong> ScottK: does ubotu need =~ /question/i && forward #ubuntu? :D
[19:13] <ScottK> Sounds reasonable to me, but I just volunteer here.
[19:13] <cjwatson> I think that's a bit much. Ubuntu developers ask each other questions too.
[19:14] <genii> jdong: Were there some changes made perhaps to the sudo system ? He's telling me vi opens but he can't change or alter anything as before
[19:15] <genii> I only ask here since there may be recent changes the main support channels have no knowledge of, don't mean to get support here :)
[19:17] <cjwatson> genii: not since before hoary
[19:17] <cjwatson> (not at that kind of level anyway, of course modulo bugs)
[19:17] <genii> cjwatson: OK, thanks
[19:25] <juliank> cjwatson: my local debimg tree can now read the Germinate output files and use them instead of calculating the dependencies itself.
[20:06] <emgent> heya
[20:36] <smallfoot-> Blender runs in fullscreen (even when I tell it to run in window mode) if I use Compiz. Without Compiz, Blender behaves properly though.
[20:39] <slangasek> smallfoot-: please see https://help.ubuntu.com/community/ReportingBugs for information about reporting bugs; reporting them on IRC is not a reliable way to get them seen by the relevant developers
[20:42] <Caesar> slangasek: so the timeframe for the first Hardy point release is July?
[20:43] <slangasek> Caesar: yes
[20:45] <Caesar> tjaalton: can you be more precise? I can't see any xorg source packages in dapper-proposed on archive.ubuntu.com, and there's no obvious links from https://bugs.launchpad.net/ubuntu/dapper/+source/xorg-server/+bug/113679
[20:45] <ubotu> Launchpad bug 113679 in xorg-server "xorg freezes when running openoffice" [High,Fix committed]
[21:31] <lucent> soren: thanks for the fix to virt-manager w/re OS settings that have no info for vt
[21:32] <lucent> at the moment I'm not 100% sure, but I think there is a bug or undocumented problem with 8.04 LTS release of virt-manager for the networking aspects
[21:32] <lucent> I am a member of libvirtd and kvm
[21:33] <lucent> However, I am not able to configure network settings?
[21:41] <sistpoty> hey LaserJock... motu meeting going on right now
[21:41] <LaserJock> oh really?
[21:41] <sistpoty> yes
[21:42] <laga> ludo_: you need to connect to the system-wide session i think
[21:54] <laga> mario_limonciell: i'd like to do an SRU for https://bugs.launchpad.net/bugs/222009 - is that OK with you or do you want to wait for other issues to come up?
[21:54] <ubotu> Launchpad bug 222009 in mythplugins "package mythvideo 0.21.0+fixes16838-0ubuntu2 failed to install/upgrade: " [Undecided,New]
[21:55] <mario_limonciell> laga, no i think that's a good idea to do an SRU, but i don't know that is the proper solution for it
[21:56] <laga> mario_limonciell: well, what do you suggest? trying to fix permissions if we can and just ignoring it if we can't sounds like a pragmatic idea ;)
[21:57] <mario_limonciell> laga, well i guess the only alternative that comes to mind is checking whether the file system is mounted on nfs
[21:57] <laga> which will also result in doing nothing. and that's probably error-prone and requires a lot more shell code
[21:57] <mario_limonciell> this solution seems like it can be a bad idea in one circumstance that i can think (and the original lines have  this same problem), overriding uid/gid on a remote system that doesn't necessarily have the same uid/gid
[21:58] <mario_limonciell> so that's why it almost makes more sense to me to check if /var/ /var/lib /var/lib/mythtv or /var/lib/mythtv/videos are on nfs
[21:58] <mario_limonciell> and if they are do nothing
[22:00] <laga> is overriding permissions a good idea anyways? there is probably a reason why the user changed them
[22:00] <laga> mario_limonciell: don't forget about cifs
[22:00] <mario_limonciell> laga, yeah maybe it's better to only override if the directory ends up getting made
[22:00] <mario_limonciell> or its a first install
[22:00] <mario_limonciell> etc
[22:02] <laga> mario_limonciell: let's assume the user is running NFS. the postinst runs with UID0. shouldn't any UID0 calls be mapped to "nobody" on the server?
[22:02] <laga> unless the user uses "no_root_squash"
[22:03] <laga> that'll also explain the failures people are seeing. UID/GID isn't changed
[22:03] <mario_limonciell> yeah
[22:03] <mario_limonciell> is there a mkdir -p for that directory i nthe postinst?
[22:03] <laga> so || true ought to be enough
[22:03] <mario_limonciell> or is it explicitly made?
[22:03] <mario_limonciell> or shipped with the package
[22:03] <laga> i think it's in the package
[22:03] <mario_limonciell> ah.
[22:04] <mario_limonciell> well i guess in that case || true should do the trick, but "laga> is overriding permissions a good idea anyways? there is probably a reason why the user changed them"
[22:04] <kees> slangasek, bdmurray: can you check on bug 222108 ?
[22:04] <ubotu> Launchpad bug 222108 in ubuntu "md5sum check fails on wubi.exe in Hardy" [Undecided,Incomplete] https://launchpad.net/bugs/222108
[22:05] <laga> mario_limonciell: now remember all the times users have thought it was a good idea to override the mysql password
[22:05]  * mario_limonciell shrugs
[22:05] <mario_limonciell> okay yeah.
[22:05] <mario_limonciell> this solution is good enough for now then
[22:05] <mario_limonciell> apply it across the board to the other plugins
[22:05] <laga> yeah
[22:07] <slangasek> mario_limonciell, laga: looking at that bug, I'm not sure why NFS has anything to do with this code failing?
[22:07] <slangasek> kees: duplicate of 222018, oddly
[22:08] <mario_limonciell> slangasek, there have been several reports of  the exact same thing popping up and it turns out that on NFS you can't chown or chmod the directories
[22:08] <mario_limonciell> bryce, and kees both ran into it too afaik
[22:08] <laga> slangasek: .. because root access is mapped to nobody using default NFS export options
[22:09] <kees> slangasek: ah-ha, thanks.  I have at least one other dup of that, I'll mark it.  thanks.
[22:09] <slangasek> oh, a concern about root squash, meh
[22:09] <slangasek> mario_limonciell: well, you can't if you're running with root_squash, correct; but then, er, don't do that :)
[22:09] <laga> no_root_squash is bad
[22:10] <slangasek> if all of /var is mounted on NFS, you are generally screwed if it's root_squash
[22:10] <slangasek> ditto /var/lib
[22:10] <slangasek> so there's no reason to solve this problem on a per-package basis
[22:11] <mario_limonciell> well that goes back to the question should permissions really be overridden on a per package basis like that
[22:11] <laga> slangasek: there are a few very good reasons to have /var/lib/mythvideo mounted via NFS. it contains your video files
[22:11] <slangasek> mario_limonciell: sure they should
[22:11] <mario_limonciell> slangasek, well i mean every time the package is upgraded though
[22:11] <mario_limonciell> the first install sure, but on every upgrade i dont think
[22:12] <laga> slangasek: and mythbuntu actually allows you to export them, so users might just mount them on another box in /var/lib/mythvideo/
[22:12] <slangasek> mario_limonciell: /var/lib == package state; if you think you know better than the package what the perms on its own state should be, you need to use dpkg-statoverride
[22:12] <slangasek> mario_limonciell: which the cited postinst code already checks for
[22:12] <laga> i meant to say /var/lib/mythtv/videos of course
[22:14] <laga> (have we reached a consensus yet? ;))
[22:16] <mario_limonciell> to at least stop the bug traffic on it, i think the || true would be fine
[22:16] <laga> good.
[22:16] <slangasek> and if the user did mount it over NFS, it has the wrong perms, and the mythtv user won't be able to access it?
[22:17] <slangasek> package installs successfully and then fails mysteriously at runtime? :)
[22:17] <laga> slangasek: an angel gets its wings? ;)
[22:17] <laga> slangasek: we can't do anything about that. i blame NFS for not having a sane UID remapping mechanism. (dunno about nfsv4)
[22:18] <slangasek> I'm asking whether you really think it's better for the package to claim to have successfully installed in this case, as opposed to bailing at install time
[22:18] <laga> nfs-user-server has something usable, but it doesn't do files over 2GB
[22:19] <laga> slangasek: it's still possible that the user made it work without invoking dpkg-statoverride.
[22:19] <slangasek> I know mvo disagrees on this point because he has to deal with all the upgrade fallout when packages do it badly :), but I'm a firm believer that packages should *not* be lenient in their maintainer scripts, because it's better to find the cause of failures up front rather than let the package install continue and leave behind an inconsistent system
[22:21] <laga> slangasek: yeah, you're talking about *installing*, i'm talking about upgrading :) i guess we can safely assume that it worked before the upgrade.
[22:21] <laga> hum
[22:21] <slangasek> well, yes, instead of the || true, you could put in a check for [ -z "$2" ]; I wouldn't object as loudly to that
[22:22] <laga> ah, $2 only shows up when upgrading. i guess it's the version number of the previously installed package
[22:23] <Caesar> laga: http://women.debian.org/wiki/English/MaintainerScripts
[22:23] <laga> okay, that makes sense. is there any way to show a nice message to the user if it fails?
[22:23] <slangasek> using debconf error templates you could
[22:24] <slangasek> but, er, that's a change that definitely shouldn't pass muster in an SRU without first being uploaded to intrepid (when it opens)
[22:25] <laga> no, we need the fix now and i don't want to do so much fiddling nowe
[22:25] <laga> now*
[22:25] <laga> thanks, you've been a great help
[22:25] <slangasek> one hopes so, at least :)
[23:16] <soren> lucent: Sorry for breaking it to begin with :)
[23:19] <lucent> soren: I have filed a new bug, with a different python error
[23:19] <lucent> soren: you will have to catch any further info from me soon, because this laptop system is going to be sent in tonight for repairs :)
[23:20] <lucent> bug #222191
[23:20] <ubotu> Launchpad bug 222191 in virt-manager "Virtual Machine Manager start-up python error as user belonging to libvirtd group" [Undecided,New] https://launchpad.net/bugs/222191
[23:23] <soren> lucent: No, I see what the problem is there. It's completely unrelated to that fix, though.
[23:34] <hwilde> why doesn't network manager put auto [iface] into /etc/network/interfaces :(
[23:34] <lucent> hwilde: network manager supercedes the use of interfaces file to define interfaces, I think
[23:35] <slangasek> because n-m can't be trusted to write the file correctly ;)
[23:35] <lucent> as a user, I remember having to disable or remove most of my /etc/network/interfaces file in order for early versions of nm-applet to function appropriately
[23:35] <hwilde> fair enough, but lots of people asking why their wireless doesn't work automaticall on boot,  they pastebin the interfaces file and it's missing auto eth1 or auto wlan0
[23:35] <crimsun> they also could simply remove the $wifi_iface stanza altogether from interfaces(5)
[23:35] <slangasek> that doesn't sound right; n-m isn't supposed to require anything in interfaces to work
[23:36] <slangasek> so "missing" auto eth1 should have no effect on n-m
[23:36] <lucent> hwilde: this is a big change from the way people are used to doing it in debian
[23:36] <lucent> or older versions of wifi access and wpa-supplicant script hacks
[23:36] <hwilde> hey I wrote one of those supplicant hacks :)  and it works great
[23:36] <slangasek> there are no uncommented lines in /etc/network/interfaces related to the wifi interface I'm currently using n-m on
[23:37] <hwilde> but what generates the interfaces file
[23:37] <lucent> nothing in your case
[23:37] <lucent> it's not generated... there are other subsystems controlling the networking functions now
[23:37] <lucent> I'm guessing HAL and/or DBUS
[23:37] <hwilde> there is an /etc/network/interfaces file
[23:38] <hwilde> and people don't type it in themselves
[23:38] <hwilde> so where does it come from
[23:38] <lucent> ohhh
[23:38] <hwilde> and why is it missing auto [interface]
[23:38] <lucent> the network configuration applet in gnome
[23:38] <lucent> I understand now
[23:38] <hwilde> which causes it to not come up on boot
[23:38] <lucent> There is a conflict between the wifi configuration of the Gnome network configuration applet, and network-manager
[23:38] <lucent> when a person enters the Gnome network configuration applet to configure their wifi device, it writes out the /etc/network/interfaces file
[23:39] <lucent> this in turn screws up network-manager from working
[23:39] <lucent> is that still a problem in Hardy?
[23:39] <slangasek> hwilde: no, see what I wrote above; lack of "auto" has nothing to do with network-manager using the interface or not
[23:39] <soren> slangasek: Er... Sure it does. n-m needs to know that it's not going to fight with ifupdown about managing an interface. It does so by checking if there's a config stanza and (I believe) an "auto $iface" as well.
[23:39] <hwilde> i've answered the same question in #ubuntu about a dozen times.  they pastebin the interfaces file and there is no auto.  and they say their wirelss doesn't come up on boot
[23:40] <hwilde> personally, I do not have these problems, and I do not use the guis, so I cannot say....
[23:40] <soren> slangasek: (Not sure about the auto part, though)
[23:40] <slangasek> soren: no, what n-m does is refuse to expose the interface if there's anything *other* than the 'auto' line
[23:40] <soren> slangasek: Ah.
[23:40] <slangasek> soren: or rather, an 'auto' line and/or an 'iface $foo inet dhcp' line, IIRC
[23:40] <soren> slangasek: Ah... Ok, reading what you said again, I see what you mean now.
[23:42] <lucent> what is "Enable roaming mode", does this toggle the use of /etc/network/interfaces ?
[23:42] <lucent> from Network settings applet
[23:42] <slangasek> hwilde: yes, so what *do* they have in their interfaces file?  they probably have /other/ stuff set, which was not set by network-manager and is a hint to network-manager to not manage that interface
[23:42] <hwilde> they have everything looking good, except the "auto" line
[23:42] <slangasek> er, no
[23:42] <soren> slangasek: I thought you were saying that n-m didn't care at all what was in /e/n/i.
[23:42] <slangasek> what is "looking good"?
[23:43] <slangasek> soren: it does care, but it doesn't care about adding/removing an 'auto' line.. :)
[23:43] <hwilde> they have everything else that is required, static ip or dhcp, wireless, whatever, but there is   no    auto
[23:43] <soren> slangasek: Right.
[23:43] <hwilde> let me look through the url grabber and see if any pastebins are still available
[23:43] <lucent> hwilde: this should not be a problem in a fresh install of Hardy
[23:43] <slangasek> hwilde: those are not things that are required, those are things that directly interfere with network-manager agreeing to manage the interface
[23:43] <slangasek> hwilde: because network-manager doesn't know how to manage /etc/network/interfaces directly, so if you configure your interface there, n-m will not touch it
[23:44] <hwilde> telling them to add the auto line fixes it, but what should they be doing, commenting out everything and letting n-m handle it?
[23:44] <slangasek> adding the auto line causes the interface to be managed with ifupdown
[23:44] <slangasek> it will not cause the interface to be configurable through network-manager
[23:45] <slangasek> if you want to use n-m, you have to comment out everything, yes
[23:45] <hwilde> like I said, I don't personally have this issue
[23:46] <hwilde> so the problem is they are going into gnome network config, and that writes the interfaces file, so n-m won't touch it, but it's missing the auto line so ifupdown is not up on boot
[23:46] <hwilde> what a mess :/
[23:47] <slangasek> seems so
[23:47] <lucent> soren: w/re the new virt-manager bug I filed, do you need any information from me?  I would like to see if it is related to bug #202432
[23:47] <ubotu> Launchpad bug 202432 in virt-manager "Unable to use anything but usermode networking" [Undecided,Invalid] https://launchpad.net/bugs/202432
[23:48] <soren> lucent: No, they're completely unrelated. I see what the problem is and I have a patch, but I need some sleep before I do anything useful about it.
[23:48] <lucent> okay, thank you
[23:48] <lucent> the user symptom is that we cannot configure networking
[23:48] <lucent> because qemu:///system does not show in the virt-manager
[23:49] <hwilde> https://bugs.launchpad.net/ubuntu/+source/ifupdown/+bug/50099  "network has to be started each time system is booted"
[23:49] <ubotu> Launchpad bug 50099 in ifupdown "network has to be started each time system is booted (dup-of: 44194)" [Undecided,Confirmed]
[23:49] <ubotu> Launchpad bug 44194 in wpasupplicant "wpasupplicant doesn't start when the network start" [Medium,Confirmed]
[23:49] <soren> lucent: Yeah, I know.
[23:49] <hwilde> in that bug post two people pasted the interfaces file missing the auto line
[23:50] <lucent> soren: when that bug gets squashed, would I then be able to configure networking?
[23:50] <lucent> or am I confused
[23:50] <slangasek> hwilde: right, definitely a bug in whatever wrote the interfaces file then...
[23:50] <soren> lucent: You can already configure networking if you just follow the guides (start virt-manager as: virt-manager -c qemu:///system).
[23:51] <soren> lucent: But if you're seeing that bug, you probably can't start virt-manager right now?
[23:51] <slangasek> hwilde: actually, in that bug I see config snippets that *do* list 'auto'?
[23:51] <lucent> soren: I can start virt-manager, and I can run and do all things except configure Networking
[23:52] <soren> lucent: Start it as "virt-manager -c qemu:///system" just once, and qemu:///system will stay there.
[23:52] <hwilde> slangasek, yeah I think that would be the thing he is talking about n-m won't touch it
[23:52]  * lucent does this
[23:52] <lucent> soren: Yes, this is working
[23:52] <hwilde> slangasek, I have specifically answered a dozen people who added the "auto" line and that fixed their problem... that's about all I know.  I use command line for everything and i've never had the problem myself
[23:52] <slangasek> hwilde: right, fair enough
[23:53] <soren> lucent: Yeah. It's on the wiki about this: http://wiki.ubuntu.com/KvmVirtManagerEtc
[23:53] <soren> Anyhow, as I was saying: It's bed time. :)
[23:53] <lucent> good night and thank you again
[23:53] <slangasek> I would love to know why avahi-autoip thinks it's clever to set a default route
[23:53] <hwilde> slangasek, I actually go as far as uninstalling n-m :)  so I don't have this interfernece
[23:53] <lucent> I read that guide and somehow missed the need to run a terminal command
[23:54] <slangasek> hwilde: I normally don't use n-m and prefer to have it uninstalled; I've been using it for the past two months in the run-up to the hardy release because there were a number of major wireless bugs that needed attention
[23:55] <hwilde> lol good excuse
[23:55] <hwilde> it doesn't like my wpa so I just run the supplicant from rc.local
[23:55]  * hwilde ducks
[23:55] <slangasek> hwilde: it's actually doing much better now than it did in the past; if someone would just fix the bug whereby n-m doesn't bother connecting to anything until after I log in, preventing me from using kerberos authentication correctly, we'd be all set
[23:56] <slangasek> hwilde: ah, there were a number of wpa-related bugs fixed in the past months, in fact
[23:56] <laga> n-m doesn't always set an IP address when connecting to my wlan :/
[23:57] <hwilde> it works for awhile, then it jsut stops... Ican't have that
[23:57] <hwilde> it also latches on to free public wifi whenever it has the chance and I can't have that either
[23:58] <hwilde> even with a strict .conf file, it roams to open networks :/