[12:16] <dholbach> good night guys
[12:44] <doko> ogra: building LTSP chroot: the progress bar jumps to 50% and stays there ...
[12:55] <Robi-> does anyone run soft-raid?
[01:36] <robertj> theoretically to get rid of grub, one could just copy over a stock mbr and munge the partition table so that the windows partition containing that file was active, right?
[03:46] <zul> thats weird none of the irc logs are being displayed with firefox
[03:57] <mdz> zul: the server which runs people.ubuntu.com has shat itself
[03:57] <mdz> assuming you're talking about those irc logs
[03:58] <zul> yeah i am
[04:01] <zul> damn it..
[06:06] <fabbione> morning
[06:28] <bytee_> good morning, fabbione 
[06:46] <pitti> Good mor*yawn*ing
[06:47] <tritium> fabbione, is your irc logger working properly?  I was trying to find the motu meeting from the 18th, but -meeting and -motu logs from that date are empty
[06:50] <ajmitch> tritium: probably because of the last item in /topic
[06:50] <tritium> ajmitch, didn't see that.  Thakns.
[06:50] <tritium> thanks even
[06:50] <fabbione> tritium: mostlikely
[06:51] <fabbione> i am pretty sure that -current are all borked
[06:51] <tritium> fabbione, okay, thanks.  Sorry, didn't see the topic
[06:51] <fabbione> i think that daily are ok
[06:51] <fabbione> but i will need to check
[06:53] <tritium> okay, thanks again
[07:27] <poningru> whats the eta on NM becoming default?
[07:27] <poningru> is it on Dapper or 6.04+1?
[07:27] <poningru> Network manager btw
[07:31] <pitti> is it just me, or do other people also have troubles with reading from people.u.c.?
[07:31] <pitti> I tried it from two completely different servers
[07:33] <ajmitch> pitti: see topic :)
[07:34] <ajmitch> apparantly the server has 'issues' :)
[07:35] <pitti> oops, sorry
[07:36] <pitti> I tend to ignore the topic since it is 6 lines long
[07:36] <mdz> it overflowed and required additional editing
[07:37] <ajmitch> yes, I saw :)
[07:37] <ajmitch> a common problem 
[07:37] <infinity> Do we really need pointers to DeveloperResources and HelpingWithBugs in the topic, when both of those shold be in the topic of #ubuntu-motd, which we already point people to? :)
[07:38] <infinity> s/motd/motu/
[07:39] <infinity> Also, when I see jdub, I'm going to strangle him for cursing me.
[07:39] <infinity> My laptop had been up for days when he showed up last night complaining about recent hardlocks, and blaming the ipw2200 driver...
[07:39] <infinity> Sure enough, I wake up this morning, and my laptop is locked hard with the wireless light solid.
[07:40] <Seveas> hehe
[07:40] <Seveas> jdubs laptop misbehaved during his presentation a few days ago
[07:40] <Seveas> beagle fubared his user_xattr so /home got mounted read only
[07:41] <infinity> Beagle?  FUBAR?  What are the odds?
[07:41] <Lathiat> heh
[07:41] <Lathiat> infinity: hahaha
[07:41] <Lathiat> i have a laptop using ipw2200 on like, 24/7
[07:41] <infinity> It still boggles my mind that people think it's both stable enough and generally useful enough to have it in the default session.
[07:41] <Lathiat> and it never skips a beat
[07:42] <infinity> Lathiat : Yes, I could say the same thing until this morning.  Just wait, your time will come.
[07:42] <Lathiat> infinity: heh
[07:42] <Lathiat> well its been on 80% of the last year
[07:42] <Lathiat> poor laptop
[07:42] <Lathiat> its probably not supposed to be on that much :)
... Mine is.
[07:42] <infinity> I just have to remember to be nice to the battery.
[07:43] <infinity> But my laptop is my only computer with a monitor, so I don't have much choice.  It gets used a lot.
[07:43] <Lathiat> heh
[07:43] <Lathiat> the same
[07:43] <Lathiat> cept i have 2
[07:43] <infinity> Well, okay, I have two computers with monitors, the other one being my OLD laptop that this one replaced.
[07:43] <Lathiat> but between the both of them i use them all the time
[07:43] <Lathiat> only other machine is just a server box with no monitor
[07:44] <infinity> And when I say OLD, I mean OLD.  I'm pretty sure people were laughing at me behing my back at UDU as I walked around with it.
[07:44] <Lathiat> haha
[07:44] <Lathiat> at lca2004
[07:44] <Lathiat> i had a p133 toshiba
[07:44] <ajmitch> infinity: surely not
[07:45] <Lathiat> altho i managed to steal a 1ghz machine towards the end of the week
[07:45] <ajmitch> I thought mine was bad enough, and it was just a p2
[07:45] <Lathiat> that toshiba was quite nice tho
[07:45] <Lathiat> it wasnt overly chunky
[07:45] <Lathiat> 160CDT or something
[07:45] <Lathiat> 14" i think
[07:45] <Lathiat> run xfce4 quite nice
[07:46] <infinity> Oh well, nothing beats dilinger's keyboard.
[07:46] <Lathiat> dilingers keyboard?
[07:46] <infinity> I was complaning because I was missing one key, then he showed me his.
[07:46] <infinity> I'll have to find a picure for you to really understand, but his laptop is missing, like, 10 keycaps.  Maybe more.
[07:46] <Lathiat> haha
[07:46] <Lathiat> nice
[07:46] <jdub> infinity: i have cursed you?
[07:46] <ajmitch> I saw a photo on tseng's site, if he still has them up
[07:46] <infinity> jdub : Yes.  Jerk.
[07:46] <Lathiat> jdub: yes! you evil man!
[07:47] <infinity> jdub : My laptop was locked solid this morning, with the wireless light solid green.  And it's ALL YOUR FAULT.
[07:47] <jdub> heh
[07:47] <jdub> fix it!
[07:47] <fabbione> infinity, jdub: buy better hw
[07:47] <fabbione> why my machines never lock solid? because i use sane hw
[07:48] <infinity> Mine's sane.  The kernel just needs to catch up.
[07:48] <fabbione> infinity: no no no.. the kernel is sane :)
[07:48] <fabbione> your hw sucks
[07:48] <infinity> I out-blinged the speed of kernel and X devel.
[07:49] <jdub> fabbione: i'm shipping off to italy today :-)
[07:49] <doko> good morning
[07:50] <fabbione> jdub: yeah but you are going to milan.. that's not really italy.. it's padania :P
[07:52] <fabbione> jdub: well they will explain that to you ;)
[07:53] <jdub> :-)
[07:53] <pitti> Hi JaneW
[07:53] <pitti> ouch, even ssh onto rookery fails
[07:53] <jdub> "Hi, my Italian friend who no longer lives in Italy says that you are not real Italians."
[07:54] <JaneW> hi pitti
[07:54] <jdub> oh man, someone put the ubuntu poster on flickr
[07:54] <jdub> so if i sync the fridge's feed, there will be boobies on the shuffle
[07:54] <pitti> jdub: oh, there is an Ubuntu poster`
[07:54] <Lathiat> hahaha
[07:54] <pitti> jdub: some guys from the "Linux Info Tag Dresden" asked me for one
[07:55] <jdub> pitti: the one with the big boobies, "because naked people are totally appropriate for business dektops" or whatever
[08:01] <highvoltage> what codec do i need to play shuttleworth_zimmerman_band.avi?
[08:02] <highvoltage> ah, gxine plays it :)
[08:30] <pef> hello
[08:32] <pef> ogra: Hello Olivier, I've made the changes needed to my kcheckgmail debdiff (malone #2018)
[09:05] <Treenaks> hi bubulle 
[09:05] <bubulle> very short concern: I try syncing with mvo's repository for APT and I currently can't
[09:05] <bubulle> cannot connect to archive (michael.vogt@ubuntu.com--2005)
[09:05] <ajmitch> people.ubuntu.com is down, see topic
[09:05] <pitti> bubulle: people.u.c. is screwed ATM
[09:05] <bubulle> ah, thanks
[09:06] <Treenaks> pitti: fixably so?
[09:06] <pitti> Treenaks: let's hope so
[09:07] <bubulle> and /me waves bye..:)
[09:07] <daniels> http://www.p0z3r.org/images/kde_hackers.png
[09:07] <daniels> i wonder if that password still works
[09:07] <Treenaks> daniels: all of them?
[09:07] <daniels> Treenaks: hm?
[09:08] <daniels> \sh_away: my god man, what the hell sort of blog software are you running?
[09:08] <Treenaks> daniels: kde_hackers.png.. I see 7 people..
[09:08] <daniels> Treenaks: note the whiteboard in the background
[09:08] <Treenaks> daniels: yeah, saw that :)
[09:08] <Treenaks> daniels: \sh blames planet for being broken
[09:09] <infinity> For the curious, not the password no longer works. :)
[09:09] <Burgundavia> daniels, it happens about once a week it seems
[09:09] <mdke> it's a pain
[09:09] <Burgundavia> infinity, at OSCON 2004, orielly gave everybody full access to safari but forgot to turn it off for 6 months
[10:27] <fabbione> people is back
[10:27] <Treenaks> \o/
[10:29] <ogra> bah, but i still get no support for dapper :)
[10:32] <Kamion> whiprush: I don't think you can say "it's a <foo> of ducks" given that I had just turned up earlier with half a dozen possible alternative collective nouns for ducks and was trying to pick the best one. :)
[10:33] <Kamion> paddling is one of the alternatives, but imho not the best one
[10:35] <Treenaks> gaggle!
[10:35] <Treenaks> or waddling
[10:35] <ogra> mob, troop, shoal, drove, swarm ?
[10:36] <Treenaks> ogra: a drove of dapper drakes?
[10:36] <ogra> why not ?
[10:36] <Kamion> it's a gaggle of geese normally, not ducks
[10:37] <Treenaks> Kamion: it's both, according to quite a few places
[10:37] <Treenaks> ogra: (troop? not troup? :P)
[10:37] <ogra> hehe
[10:38] <ogra> would quiver fit ? 
[10:39] <HiddenWolf> \sh_away, is your blog messed up? It's re-posting everything you posted the last week(s) to planet
[10:41] <seb128> Kamion: does useradd behaves the same way as adduser?
[10:44] <ogra> seb128, i think one of them creates a homedir by default iirc (dont ask which though)
[10:45] <Kamion> seb128: useradd's a low-level tool and is not obliged to
[10:45] <tsume> mm, one question. packages are based on stability and changed towards stability, yes?
[10:45] <Kamion> Treenaks: I think geese is the association most native speakers have for "gaggle", but shrug
[10:45] <tsume> once released that is. What if there is a bugfix update for a package?
[10:46] <tsume> would it be able to get pushed out in a tree, or would it only be maintained for current bugs in the release of the software?
[10:47] <seb128> Kamion: users-admin users useradd, should it use adduser rather?
[10:50] <Kamion> oh, yeah, I'd've thought so, if it can ... adduser has better default policy, and uses useradd under the hood anyway
[10:50] <Kamion> and it automatically creates groups for you and stuff
[10:51] <infinity> Does useradd support different backends, or is it just one of the supported adduser backends?
[10:51] <Kamion> the arguments it takes are quite different though
[10:51] <infinity> The nice thing (in theory) with adduser is that it doesn't have to be dealing with /etc/passwd at all.
[10:51] <Kamion> infinity: adduser always uses useradd
[10:51] <Kamion> wouldn't you do different backends with PAM?
[10:52] <Kamion> dunno, I haven't looked at the useradd code
[10:52] <infinity> Oh, I suppose you would these days.
[10:52] <infinity> I must be living in 1995.
[10:52] <infinity> Anyhow, the arguments aren't so drastically different that one couldn't patch an application to use one instead of the other with minimal effort.
[10:53] <infinity> They're just incomaptible as drop-in replacements.
[10:53] <Kamion> yeah
[10:53] <Kamion> infinity: it's more that the default behaviour's different, so it needs testing
[10:54] <Kamion> I wouldn't quit or anything if we flipped the adduser/homedir-permission default :-) I just think many of the arguments being advanced are bogus and paranoid and miss the UI point ...
[10:55] <infinity> I'd walk. ;)
[10:55] <infinity> (Well, I'd just dpkg-reconfigure adduser and go about my day, but still..)
[10:56] <infinity> My guess is that 99% of the people who argue that 700 or 711 is a better default than 755 have never actually worked on large multi-user systems.
[10:56] <infinity> It becomes very obvious, very quickly, that convenience is more important than paranoid security in most setups.
[10:56] <infinity> (Odds are that most of them are home users, bent on shoving their paranoid permission scheme down the throats of others who actually DO use large multi-user systems and tihnk the current default is fine)
[10:57] <infinity> But, that's just a guess. :)
[10:57] <Kamion> yeah, that's been my experience on large multi-user systems too
[10:57] <Kamion> but what do I know, I've only tried it

[10:59] <infinity> To be fair, I also run one system where the users aren't all buddy-buddy and I did reconfigure adduser to use 711.  But is that really so hard to do, if you feel the need?
[10:59] <infinity> That's one system out of many, so I know which default makes more sense for me.
[11:08] <zyga> can anyone tell me the rough estimate for the number of files in a full mirror of ubuntu?
[11:09] <Yagisan> My take on it (as a "security guy" ) is that the tools should be consistent in setting permissions, and info on how to reconfigure
[11:09] <Yagisan> them should be stuck in a securing ubuntu manual
[11:09] <Znarl> zyga : Archive or releases mirror?
[11:09] <zyga> Znarl: archive
[11:10] <Znarl> 163903 files and directories.
[11:11] <zyga> thanks
[11:11] <zyga> Znarl: could you also tell me the total size?
[11:11] <Znarl> 79G presently
[11:11] <zyga> cool, thanks alot
[11:12] <Keybuk> Znarl: gah, you beat me
[11:12] <Keybuk> 79G     /srv/archive.ubuntu.com/ubuntu
[11:12] <Keybuk> :p
[11:13] <Znarl> Sorry
[11:21] <zyga> does the installer ask the user if s?he would like to participate in the popularity contest?
[11:29] <HiddenWolf> zyga, no
[11:42] <jdub> http://lists.ubuntu.com/archives/ubuntu-announce/2005-October/000043.html
[11:42] <jdub> UBUNTU LOVE DAY!
[11:42] <ajmitch> hi jdub 
[11:43] <ajmitch> mm, nice
[11:44] <fabbione> jdub: the wiki page is empty?
[11:44] <fabbione> and 2 emails to announce :) cool
[11:44] <fabbione> at least i got 2
[11:48] <highvoltage> that's scary.
[11:48] <infinity> Yagisan : Yeah, no one's arguing that the tools should lack consistency.  gnome-system-tools should obviously be using adduser.
[11:48] <infinity> Yagisan : We're all just arguing about defaults. :)
[11:54] <slomo> pitti: i have two debdiffs for warty and hoary to fix CVE-2005-2718
[11:55] <slomo> pitti: http://slomosnail.de/~slomo/mplayer_1.0-pre5-0.6ubuntu1.1.debdiff and http://slomosnail.de/~slomo/mplayer_1.0-pre6-0.3ubuntu6.1.debdiff
[11:55] <slomo> pitti: it's the same patch we already have for breezy... applies cleanly on the older versions
[11:59] <pitti> slomo: yes, sounds fine; thanks
[12:00] <slomo> pitti: and your tool reports it is still unfixed for breezy... i uploaded a fix for it ages ago but didn't mention the CVE in the changelog
[12:02] <pitti> slomo: I can mark it manually
[12:02] <slomo> thanks :)
[12:02] <pitti> slomo: so mplayer breezy is not vuln to CVE-2005-2718?
[12:03] <slomo> pitti: yes... it has exactly the same patch
[12:03] <ajmitch> pitti: there's yet another phpmyadmin hole, not sure if you saw it
[12:03] <pitti> ajmitch: I saw it, but universe :-/
[12:04] <ajmitch> yeah
[12:04] <ajmitch> want me to put a debdiff together?
[12:04] <infinity> ajmitch : And update prepared would be nice.
[12:04] <ajmitch> ok
[12:04] <Kamion> at this rate cdimage is going to gain the ability to read mail
[12:04] <infinity> ajmitch : phpmyadmin is reasonably popular.
[12:04] <Kamion> I just added a counting semaphore implementation to it ...
[12:04] <ajmitch> certainly
[12:04] <infinity> Kamion : It can't already?
[12:04] <pitti> ajmitch: sure, and send it to security-review
[12:04] <ajmitch> pitti: will do
[12:04] <infinity> Kamion : Dude, even buildd can read mail.  You're so behind.
[12:05] <Kamion> infinity: not in any code I've written. It's possible it's started to write code for itself.
[12:08] <Kamion> sorry, I should've said. A counting semaphore implementation IN SHELL.
[12:09] <Kamion> it's possible shell has stopped being the best language for cdimage, but ...
[12:09] <Diziet> Woah.
[12:11] <hunger> Kamion: When I was in uni I was told that GNU products were only complete if they were able to read mail and news...
[12:14] <pitti> hmm, why do we enable DMA on CD-ROMs by default now?
[12:14] <pitti> this hasn't been the case in hoary
[12:15] <pitti> and it breaks on so many machines out there (and I get all the hal bugs)
[12:15] <ajmitch> great, phpmyadmin fix is only 3 lines
[12:15] <Treenaks> pitti: I thought we didn't, but HiddenWolf's machine disagrees
[12:15] <Treenaks> pitti: so maybe the disabling is not working for some hardware?
[12:15] <pitti> Treenaks: I thought that, too, and I always replied with "then don't enable DMA, dude"
[12:16] <pitti> but I just noticed that DMA is active for my drive, too
[12:16] <HiddenWolf> Treenaks, /dev/hda is my windows drive, which is not in fstab
[12:16] <pitti> seems to WFM, but it is wrong
[12:16] <HiddenWolf> Treenaks, dma is on for cd-roms too tho.
[12:16] <pitti> yes, for hard disks its fine
[12:16] <Treenaks> HiddenWolf: on hard disks DMA gets enabled
[12:16] <pitti> Hi carstenh 
[12:16] <carstenh> hi pitti :)
[12:16] <Treenaks> pitti: we will need to figure out a way to determine if the CD-ROM/Controller combination is "DMA-OK"
[12:17] <Treenaks> pitti: because no-dma sucks hard
[12:17] <HiddenWolf> Treenaks, no, it's off for both my cdrom drives, in fact.
[12:17] <bob2> lincvs is fux0red in breezy, if anyone cares
[12:17] <pitti> Treenaks: DMA sucks even harder on many hardware
[12:18] <Treenaks> pitti: So we need a black/whitelist
[12:18] <Treenaks> pitti: because on _most
[12:18] <Treenaks> _ systems it seems to work fine
[12:18] <Treenaks> (most systems I've seen)
[12:30] <Yagisan> infinity: you are a desktop distro, targeted at single users. 755 works for them. People (like me) that do run
[12:31] <Yagisan> multi-user systems will run dpkg-reconfigure adduser (but that should be in a manual for new sysadmins)
[12:31] <Kamion> mdz: FYI I've implemented parallel CD builds now at least for different projects (so we can run Ubuntu, Kubuntu, and Edubuntu builds at the same time, though not parallel install/live builds for the same project)
[12:31] <Kamion> Yagisan: s/will/may/
[12:33] <Yagisan> Kamion: yeah - it depends on the system, I know.
[12:33] <infinity> Kamion : How much of a win is that?  Won't you just be hampered by disk speed anyway?
[12:36] <crimsun> infinity: is this applicable for your ipw2200 freeze? It's a simple diff. jdub might want to take a look too if it's relevant. http://bughost.org/bugzilla/show_bug.cgi?id=740
[12:36] <crimsun> (patch being http://bughost.org/bugzilla/attachment.cgi?id=479)
[12:37] <mvo> hey doko! are you at berlinux?
[12:37] <infinity> crimsun : Dude, I've only hung once.  It's a bit hard to debug that sort of (in)frequency.
[12:38] <Kamion> infinity: there's a fair amount of waiting around for jigdo-file in there
[12:38] <Kamion> which is mostly md5summing the world
[12:38] <Kamion> I think there's enough of a mix between CPU-intensive work and I/O-intensive work that they should parallelise somewhat, but we'll see
[12:39] <infinity> I mean, I could write one that basically had livefs.sh embedded in it, but I'm not sure we want to let that hideous code see the light of peer review^W^Wday.
[12:41] <doko> mvo: yes
[12:41] <lucas> hi
[12:45] <mdke> jdub, i'm still not on planet afaics :/
[12:50] <fabbione> later fellas
[12:51] <pitti> cu fabbione 
[12:51] <vuntz_> hey lucas :-)
[12:52] <lucas> hi vuntz 
[12:56] <pitti> Kamion: does the installer put USB CD-ROM entries into fstab under some circumstances?
[12:56] <zyga> pitti: hi
[12:56] <pitti> Hi zyga
[12:57] <zyga> pitti: about the installer, I hate bug reports and all but I've ran across a strange behaviour
[12:58] <zyga> pitti: I had 1 ata and 2 sata disks, with sata disks the installer borked and died saying it cannot resize ext2 filesystems (even though I didn't even ask it to mount either of the sata disks) on the first sata disk it found
[12:58] <pitti> zyga: bugzilla is exactly the right place for this report :-/
[12:58] <Kamion> pitti: if mounted, if you're booting from them ... would have to dig into partman to say any more
[12:58] <Kamion> mounted> as in explicitly in the partman UI
[12:59] <pitti> Kamion: ok, thanks
[12:59] <pitti> Kamion: I'm asking because of #6106
[01:02] <pitti> anyone with an USB cdrom drive?
[01:03] <Lathiat> pitti: yeh
[01:04] <Lathiat> pitti: need somethign tested?
[01:04] <mvo> pitti: not here, but I had one last weekend and it worked fine with breezy
[01:04] <pitti> Lathiat: if you plug it in, do you get /dev/sr0, /dev/scd0, or both?
[01:04] <Lathiat> both if i recall
[01:04] <Lathiat> but i'll try
[01:04] <pitti> Lathiat, mvo: maybe you can say something to #6106
[01:04] <Kamion> pitti: oh, no, I think it always puts removable media into fstab
[01:05] <Kamion> partman-target/finish.d/fstab_removable_media_entries
[01:05] <pitti> ah, bad
[01:05] <pitti> interesting, I never installed Ubuntu with my USB stick attached
[01:06] <Kamion> we can probably just turn that off in dapper and see what happens, although new installs will no longer be able to use the traditional 'sudo mount /cdrom'
[01:06] <pitti> Kamion: well, "removable" media in fstab are fine, just not "hotpluggable" ones
[01:06] <pitti> Kamion: i. e. entries for an ATAPI card reader are fine, but it will break for an USB CD-ROM or USB stick
[01:06] <Kamion> uh
[01:07] <Kamion> I don't think partman knows the distinction
[01:07] <pitti> (card reader, ATA CD-ROM: removable; USB stick: hotpluggable)
[01:07] <Kamion> would take some hw-detect hacking to arrange tht
[01:07] <Kamion> that
[01:07] <pitti> Kamion: can we teach it to not add USB and Firewire devices?
[01:07] <Kamion> like the existing hw-detect hacking that tells netcfg about hotpluggable network cards
[01:08] <Kamion> I don't want to hardcode device types if I can avoid it; a hw-detect trick that remembers when disk devices have been hotplugged would be better if possible
[01:08] <Lathiat> at one point the installer was putting /dev/sda instead of /dev/sda1 for removable usb drives in /etc/fstab
[01:08] <Lathiat> dunno if that was fixed
[01:08] <Kamion> but yeah, I could just delete the USB storage bit; I don't think it ever adds Firewire devices
[01:09] <Lathiat> i think it adds firewire too
[01:09] <Lathiat> i can test if you like
[01:09] <pitti> seb128: thanks for adding the watch to malone #3418 - I guess I still need to get used to work with Malone
[01:09] <Lathiat> ive got a dual firewire/usb enclosure
[01:09] <Kamion> Lathiat: #10900
[01:09] <Kamion> Lathiat: I've got the relevant code open in front of me and I can't see anything that adds Firewire
[01:10] <Lathiat> interesting
[01:10] <Lathiat> i reported that ages ago
[01:10] <Kamion> it iterates through /proc/scsi/usb-storage-* /proc/scsi/usb-storage
[01:10] <Lathiat> Kamion: ok it might not then
[01:10] <Kamion> Lathiat: yeah, and see my explanation at the end of why the trivial fix wouldn't work and it needs something more complicated
[01:10] <Lathiat> ah yeh my comment is in the bug haha
[01:11] <Lathiat> Kamion: i take it from the bug you dont need an example now
[01:11] <Lathiat> hm i need to hunt for a usb cable
[01:12] <pitti> Keybuk: do you want to merge udev, or shall I?
[01:12] <Kamion> I still haven't found my USB sticks though :(
[01:13] <Keybuk> pitti: I will, it's not a simple matter of just taking the Debian version
[01:13] <Keybuk> I have some pre-prepared packages here
[01:13] <pitti> Keybuk: okay, great; I'm just eager to try it to play with libgphoto
[01:13] <Keybuk> we need to figure out what to do with those
[01:13] <pitti> Keybuk: I currently stack my uploads in chinstrap:~pitti/uploads until dapper opens :-)
[01:14] <Keybuk> the "new udev" doesn't give us a move away from /proc/bus/usb, we need the new kernel for that too
[01:14] <pitti> ah, right
[01:14] <Keybuk> as the kernel needs to drop usbfs
[01:15] <pitti> Keybuk: since the new hal's udev rule approach (instead of hotplug.d) works fine here, would the "new" way already work with our udev (0.063)?
[01:15] <Keybuk> yes
[01:15] <pitti> ah, cool
[01:15] <pitti> so we need the kernel's "sysfs devices" patch
[01:16] <Kamion> Lathiat: no, I don't, thanks
[01:17] <Lathiat> pitti: 
[01:17] <Lathiat> [4299920.567000]  sr1: scsi3-mmc drive: 2x/48x writer cd/rw xa/form2 cdda tray
[01:18] <Lathiat> i get scd0 too
[01:18] <Lathiat> and a sr1
[01:18] <Keybuk> I'm still trying to understand how the kernel guys use git
[01:18] <Lathiat> and my internal (sata) cdrom
[01:18] <pitti> Lathiat: ok, thanks
[01:18] <Lathiat> is sr0 and scd0 too
[01:18] <pitti> Lathiat: the guy in #6106 apparently does not have scdN
[01:18] <Keybuk> it seems that gregkh actually puts patches in git, rather than has separate branches
[01:19] <Lathiat> anyone else want any debugging
[01:19] <Lathiat> usb/firewire drive during install?
[01:19] <pitti> Lathiat: not from my side; thanks!
[01:21] <Keybuk> pitti: http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=fbf82fd2e1f4e679c60516d772d1862c941ca845
[01:21] <Keybuk> is the patch we want
[01:21] <Keybuk> I'm trying to work out when that's due/or got folded into a release
[01:22] <pitti> Keybuk: ah, cool, so it already cares for the major/minor (in the "dev" attribute)
[01:23] <Keybuk> yeah
[01:23] <Keybuk> that's the nice thing about all the "proper" device stuff now
[01:23] <Keybuk> the kernel can assign random numbers to them, it doesn't matter
[01:23] <pitti> and it doesn't even look particularly scary
[01:23] <Keybuk> aye, and totally replaces usbfs
[01:24] <Keybuk> we just need to change /proc/bus/usb to /dev/bus/usb (or wherever we put them)
[01:24] <pitti> Keybuk: the patch is small enough to be included even in our current kernel (AFAICS), if it doesn't get upstream any soon
[01:27] <Keybuk> assuming that linus's linux-2.6.git tree is what gets released, it's in the current 2.6.14-rc series
[01:32] <tepsipakki> is there a mechanism for you to force mirror-admins to update their mirrors? for instance fi.archive.ubuntu.com is currently broken on ppc
[01:33] <Kamion> depends on the mirror
[01:33] <tepsipakki> a friend of mine complained about it
[01:33] <Kamion> fi == heanet
[01:33] <Kamion> Znarl: ^--
[01:33] <tepsipakki> yes
[01:33] <Kamion> I thought we'd taken heanet out of rotation for (?)nl
[01:33] <tepsipakki> kamion: the installer defaults to it (in Finland) so it is quite critical in that sense =)
[01:33] <Kamion> tepsipakki: yes I know :P
[01:34] <Kamion> tepsipakki: Znarl runs the mirror network, which is why I drew his attention above; explaining to me how critical it is won't help ;)
[01:35] <tepsipakki> ok
[01:35] <tepsipakki> just wanted to inform you
[01:36] <Kamion> yeah, thanks, as you can see I've passed it on
[01:37] <tepsipakki> yes, thanks ;)
[01:39] <Znarl> Kamion : Ok, I'll chase heanet.
[01:44] <Kamion> ta
[01:48] <Kamion> jesus, you post to ubuntu-devel and get a huge load of replies in your private inbox
[01:49] <Kamion> mailers so badly need to respect m-f-t
[01:49] <Treenaks> set a bounty :)
[01:51] <infinity> I'd be very happy if Tbird respected M-F-T.
[01:51] <Kamion> my concern about it is slightly less than my desire to own a house
[01:51] <infinity> But M-F-T is something only nerds think the world should respect, and the reality is that pretty much no "friendly" MUAs even look at it.
[01:52] <lucas> how do I send a mail to all members of a launchpad team ?
[01:52] <infinity> It actually irritates me a little bit every time I see a user get smacked down by someone with "jesus, can't you read M-F-T?!"... Seriously, how many "average users" read the full headers on all their mail?
[01:52] <Kamion> infinity: that's only because nerds get far more mail than everyone else
[01:53] <infinity> Kamion : Yes, and nerds should implement M-F-T in all the common mailers if they care, then.
[01:53] <infinity> Kamion : Or, at least, all the ones we have the power to change.
[01:53] <infinity> Get it into Evo and TBird, and I'll bet gmail will fix their webmailer for you.  That should kill 99% of Ubuntu list mailers.
[01:53] <Kamion> I was under the impression evolution already had reply-to-list
[01:54] <infinity> Does it use M-F-T, or list headers, or use an educated guess?
[01:54] <Kamion> no idea
[01:54] <infinity> I tend to just use reply-to-all in tbird, then remove the poster's email address iff I recall them being subscribed.
[01:55] <infinity> (So, I remove developers, but will reply-to-all with most users I don't recognise)
[01:55] <infinity> Nothing more irritating than replying-to-list with a user who comes back a week later asking the same question, cause they didn't actually think to check the list for replies.
[02:11] <doko> photos from berlinux.de - http://people.ubuntu.com/~doko/berlinux/
[02:12] <azeem> doko: nice thumbnails :P
[02:12] <ogra> hehe
[02:13] <tepsipakki> hmm, shouldn't the live/install-dvd load all the language-packs for the chosen language? I chose "finnish" but it didn't load them, locale was correct though
[02:15] <azeem> doko: who are the guys in 42 and 46?
[02:15] <doko> ogra: edubuntu even works ;-P
[02:15] <ogra> tepsipakki, it does load them at the end of the install... if you put in the dvd after reboot at any time it will pull them from there ... (some space for improvement)
[02:15] <ogra> doko, yes, but its easier to install if you read the fine manual :P
[02:16] <ogra> *g*
[02:16] <doko> azeem: herzi and a nice guy from openoffice
[02:16] <azeem> ah, looked a bit like herzi
[02:17] <ogra> i like DSCN9044 :) a lot of edubuntu there :)
[02:17] <tepsipakki> ogra: oh, this is the live-session I'm talking about..
[02:17] <ogra> tepsipakki, oh, ok...
[02:18] <Znarl> tepsipakki : Can you help me?
[02:18] <Znarl> tepsipakki : Do you have any more details about brokenness with heanet so I can bully the admin?
[02:19] <tepsipakki> znarl: for instance mozilla-firefox-locale-fi-fi_1.0.4lang20050515-1ubuntu3_all.deb complains about size mismatch
[02:20] <Znarl> OK, thanks.
[02:21] <tepsipakki> or, actually dpkg/apt is the one complaining, but anyway
[02:22] <dilinger> infinity: my keyboard will be making an appearance at UBZ
[02:23] <ogra> dilinger, you have still the same one you had at UDU ?
[02:30] <tepsipakki> ogra: is there a component I can file a bug on?
[02:31] <ogra> casper probably ...
[02:41] <Kamion> tepsipakki: we don't have a separate live fs build for the DVD yet; it's a fairly well-known problem
[02:41] <Kamion> it's not a casper bug
[02:42] <Kamion> tepsipakki: #17548
[02:49] <tepsipakki> kamion: ok, good to know
[02:57] <Lathiat> can someone dpkg -S gnome-settings-daemon for me
[02:59] <tseng> gnome-control-center
[02:59] <Lathiat> thanks
[02:59] <Lathiat> eh that pulls in lots of shit
[02:59] <Lathiat> i just want my themes to work damnit :(
[03:01] <seb128> pitti: np for the watch :)
[03:01] <seb128> Lathiat: what is the issue?
[03:01] <Lathiat> seb128: using kde
[03:01] <seb128> bah
[03:01] <seb128> you can stop here :p
[03:02] <Lathiat> its not my fault gnome doesnt have enough bling
[03:02] <Amaranth> naughty
[03:03] <Lathiat> so kubuntu whisked me away
[03:03] <seb128> I like my desktop beeing clean and not a jukebox :)
[03:03] <Amaranth> clean is nice
[03:03] <Lathiat> i feel dirty but its kinda like a good dirty
[03:03] <Lathiat> ;p
[03:03] <Lathiat> YAY clearlooks
[03:03] <Lathiat> also can we enable animation in clearlooks for dapper :(
[03:03] <Lathiat> it entirely depresses me not havign animated progress bars
[03:03] <Lathiat> i mean, kde has them
[03:04] <Amaranth> Lathiat: they spin backwards, it makes you think the progress bar is counting down
[03:04] <Lathiat> it doesnt actually
[03:04] <Amaranth> and it needs to be a gtk thing so apps can choose whether or not they want to use them
[03:04] <Lathiat> its a result of it aligning itself to the right of the progressbar
[03:04] <Lathiat> or just a visual screw up or something
[03:04] <Lathiat> at least i think
[03:04] <seb128> some people bugged the other way
[03:04] <Lathiat> what it looks like
[03:04] <Amaranth> they spin backwards
[03:04] <seb128> than animation are annoying
[03:04] <Amaranth> remenic said so
[03:04] <Lathiat> seb128: whats the bug number? i'll flame them to death :)
[03:05] <Lathiat> PLZ TO TURN BACK ON OR I WILL KEEP USING KDE. KTHXBAI.
[03:05] <seb128> sure, that's the way to fix an issue
[03:05] <Lathiat> seb128: obviously
[03:05] <seb128> happy KDE using
[03:05] <Lathiat> thanks. :)
[03:05] <Amaranth> Lathiat: build your own gnome-themes package
[03:05] <Lathiat> i did ;p
[03:06] <seb128> no need of that to install a theme
[03:06] <seb128> you can do to your user folder
[03:06] <Lathiat> yeh but that was easier
[03:06] <Amaranth> you need to change a gtkrc file or something too to enable the animation
[03:06] <Lathiat> you do?
[03:06] <Lathiat> i just compiled it and it worked
[03:06] <Amaranth> oh :P
[03:07] <Amaranth> i could have sworn he was going to make it configurable from gtkrc
[03:07] <Lathiat> i wish it was
[03:07] <Lathiat> :)
[03:07] <seb128> "was going"
[03:07] <seb128> not made yet
[03:09] <seb128> pitti: you use an amd64 install on your day to day work box?
[03:09] <seb128> mvo: you too?
[03:10] <mvo> seb128: yes
[03:10] <mvo> it's my main development box
[03:10] <mvo> (because it's fast!)
[03:10] <seb128> mvo: does "gthumb /" crash?
[03:11] <mvo> seb128: yes
[03:11] <seb128> k
[03:11] <seb128> amd64 specific issue
[03:11] <seb128> let's put dholbach on it :)
[03:12] <seb128> thanks
[03:12] <dereks__> mvo: you are using amd64?
[03:12] <dereks__> smp or 1 cpu?
[03:14] <mvo> 1cpu
[03:14] <dereks__> ahh, nm then
[03:14] <mvo> nm?
[03:15] <dereks__> never mind
[03:15] <dereks__> i was going to ask you if you had trouble wiht an smp setup
[03:15] <dereks__> i couldn't get the smp kernel to load
[03:15] <mvo> seb128: crash happens in resolve_all_symlinks
[03:15] <dereks__> so i am using 686 smp instead :(
[03:15] <mvo> do you want me to have a look?
[03:16] <mvo> derek__: you may try #ubuntu-kernel
[03:16] <seb128> mvo: http://bugzilla.ubuntu.com/show_bug.cgi?id=6523 if you have a good bt
[03:16] <dereks__> mvo: didn't know about that one :)
[03:16] <seb128> mvo: we lag 2 upstream versions behind, dholbach packaged the new one, I would let him give it a try before bothering
[03:18] <Amaranth> does xgettext no longer exist in breezy or something?
[03:19] <mvo> seb128: do you know if his package is available somewhere? 
[03:19] <seb128> gettext: /usr/bin/xgettext
[03:20] <seb128> mvo: on his disk I guess :) Don't worry, the bug is here for months, it can wait a few day
[03:22] <mvo> seb128: ok
[03:24] <dconlon> Anyone know anything about the disappearance of libXcursor.la from the libxcursor-dev package in Breezy?
[03:25] <seb128> .la are evil, daniels has decided to destroy them :)
[03:26] <dconlon> Fair enough, breaks my Enlightenment builds though
[03:27] <seb128> you have some other .la mentionning it probably
[03:27] <seb128> just trash all the .la doing that
[03:27] <seb128> or rebuild stuff with the right order so they stop doing that
[03:28] <dconlon> cheers I'll get on with that, what's the particular problem with .la's though?
[03:28] <seb128> they create such issues :p
[03:29] <doko> http://gnome-look.org/content/preview.php?preview=1&id=29911&file1=29911-1.jpg&file2=&file3=&name=ubuntu+badger
[03:30] <seb128> doko: cool
[03:30] <Keybuk> dconlon: they show you up when you fuck up something to do with the builds
[03:31] <Keybuk> so rather than fix the problems, people just remove the .la files and hide them
[03:31] <Keybuk> "the .la file has the wrong information" just means you gave it the wrong information, and have a problem
[03:31] <Keybuk> but some people are stupid and/or lazy
[03:31] <seb128> I agree with the lazy
[03:31] <seb128> but rebuilding 100 packages to update the .la files when something change is rather annoying
[03:32] <seb128> ie: when cairo stopped to use glitz
[03:32] <Keybuk> that's going to break people using static libraries too
[03:32] <seb128> you have to rebuild, with the right order, all the stack
[03:32] <Keybuk> so you need to delete all the .a files as well
[03:33] <seb128> I doubt that his e17 build will need any static lib
[03:33] <Keybuk> bet you they still get shipped though
[03:33] <Keybuk> tbh, though, I'm somewhat inclined to write something that looks and smells like libtool. but actually isn't
[03:33] <Keybuk> and just very thinly wraps gcc
[03:34] <Keybuk> so when we build our packages, we end up with valid libtool .la files -- but don't actually use libtool to do it
[03:34] <Keybuk> it takes away our build problems, without breaking the users (who can still use libtool itself)
[03:36] <dconlon> There seem to be a lot of .la's created when I do my e17 build so it seems there seems to be some static libs being generated that reference each other, hence the problem with libXcursor.la
[03:36] <seb128> no
[03:36] <seb128> have the .la created doesn't mean that you use them
[03:36] <seb128> it just give you the option to do so
[03:38] <dconlon> is there a way then of doing a dynamic only build?
[03:40] <Lathiat> arent .la files fairly useless on linux anyway
[03:41] <seb128> you don't want to start this discussion with Keybuk :)
[03:41] <Lathiat> haha
[03:41] <seb128> dconlon: --enable-static=no ?
[03:43] <Keybuk> Lathiat: no :p
[03:44] <Keybuk> they aren't needed on Linux for building shared libraries
[03:44] <Keybuk> but they are still needed for building and linking to static libraries
[03:44] <Keybuk> if you ship a .a file in your -dev package, you should also ship the .la
[03:44] <Lathiat> Keybuk: oh ok
[03:44] <Keybuk> but if you don't ship the static, on Linux, there's no reason to ship the .la
[03:44] <Lathiat> whats it needed for?
[03:44] <Lathiat> for statics?
[03:44] <Keybuk> static libraries are just an archive of object files
[03:44] <Lathiat> (i know its needed on some other platforms where you cant get the right info or something)
[03:44] <Keybuk> they don't contain dependency information
[03:44] <jbailey> Lathiat: defining relationships
[03:44] <Keybuk> or version information, etc.
[03:45] <Lathiat> ah ok
[03:45] <jbailey> Shared ELF objects have that information
[03:45] <Keybuk> so if I link to libxml2.a, it will fail
[03:45] <Lathiat> but linux shared elfs do
[03:45] <Keybuk> because I _also_ need to link to libz.a and libdl.a and libm.a
[03:45] <Keybuk> and the .la file contains that information
[03:45] <Keybuk> the .la file also contains the version information for the .a file, to make sure you don't link incompatible ones together
[03:45] <Keybuk> all this is in the shared ELF on Linux, yes
[03:45] <Keybuk> but not in the .a
[03:48] <Lathiat> kernel stores TTL __s16
[03:48] <Lathiat> oh
[03:48] <Lathiat> actually
[03:48] <Lathiat> mc_tlt
[03:48] <Lathiat> __u8
[03:48] <Lathiat> mc_ttl
[03:49] <Lathiat> struct inet_sock {
[03:49] <Lathiat> ...
[03:49] <Lathiat>     __u8            mc_ttl;     /* Multicasting TTL */
[03:49] <Lathiat> mc_loop is just 1 bit
[03:49] <Lathiat> so let me chck what it decodes that as
[03:50] <Lathiat> = !!val
[03:50] <Lathiat> val being an int
[03:50] <Lathiat> setsockopt takes an int
[03:50] <Lathiat> so its probably casted in the kernel
[03:50] <Lathiat> oh actually
[03:50] <Lathiat> i lied
[03:50] <Lathiat> its taken as  _u8
[03:51] <Lathiat> and then casted into an int in the kernel
[03:51] <Lathiat> its a pointer tho
[03:51] <Lathiat> so it can really read whatever it wants out of it
[03:51] <Lathiat> so mc_loop is !!val;
[03:51] <Lathiat> for val = int
[03:51] <Lathiat> and val = (int) optval
[03:51] <Lathiat> and optval is a _u8
[03:52] <Lathiat> for ttl its then stored without explicit cast
[03:52] <Lathiat> in a _u8
[03:52] <Lathiat> so
[03:52] <Lathiat> go figure all of that out
[03:52] <Lathiat> oh wtf
[03:52] <Lathiat> irssi just typed that all into th wrong window
[03:52] <Lathiat> but it said #avahi in the left
[03:54] <ogra> EINDENTATION
[03:54] <Lathiat> sorry guys
[03:54] <Lathiat> must have got my terminal corrupted or something
[03:54] <sivang> 'afternno all
[03:54] <ogra> oh, its not code :)
[03:55] <Lathiat> no
[03:55] <Lathiat> i was just discussing the 3000 type changes
[03:55] <Lathiat> something passed to setsockopt goes through
[03:55] <Lathiat> in an attempt to dtermine the type we shoudl be passing to IP_MULTICAST_LOOP and IP_MULTICAST_TTL
[04:00] <HiddenWolf> crimsun, ping
[04:05] <sivang> pitti: Hey Martin
[04:05] <pitti> yay, network back
[04:12] <HiddenWolf> seb128, you took rhythmbox versions from cvs, right?
[04:12] <infinity> dilinger : Will you be there with your keyboard?
[04:12] <dilinger> infinity: yes
[04:12] <seb128> HiddenWolf: no
[04:13] <Lathiat> HiddenWolf: package claims to be 0.9.0
[04:13] <HiddenWolf> seb128, i remember a lot of -cvs-etc version numbers, but must be some other package then.
[04:13] <HiddenWolf> Lathiat, yes, that's why I was confused. I thought it was -cvs
[04:15] <infinity> dilinger : \o/
[04:15] <infinity> dilinger : For how much of it?
[04:17] <seb128> HiddenWolf: I did package the CVS before 0.9.0
[04:17] <seb128> HiddenWolf: but that was before 5.10, not for a stable
[04:18] <dilinger> infinity: not sure yet.  i would like to stick around for the kernel stuff (but since the schedule's not known in advance, i have no idea whether i'll be able to), and i wanted to do some brainstorming/chatting w/ jeff about cdbs related things
[04:19] <HiddenWolf> seb128, right, good. :) 0.9.1 changelog looks like it'll solve all my gripes with th current package. I was just checking it it'd all apply to the version now in breezy.
[04:28] <freeflying> why don't we make grub perform more beautifuly with patch from susel 
[04:30] <Lathiat> hrm, if you try compile a .cpp with gcc
[04:30] <Lathiat> it spits out some undefined symbol error
[04:30] <Lathiat> (compiles fine as a .c, g++ compiesl it fine)
[04:31] <infinity> freeflying : Define more beautifully?... Ideally, it should be completely silent, except when displaying menus.
[04:31] <infinity> freeflying : I'm indifferent about making the menus "prettier" at the expens of them perhaps not working anymore for some people.
[04:32] <freeflying> but many newie prefer this 
[04:32] <Kamion> freeflying: we tried using a graphical grub menu and it made it look worse in the common case due to VGA mode switches, plus caused problems on some weird graphics cards
[04:32] <Lathiat> /tmp/ccAz06lX.o:(.eh_frame+0x11): undefined reference to `__gxx_personality_v0'
[04:32] <Lathiat> collect2: ld returned 1 exit status
[04:32] <freeflying> so they may think thak ubuntu perform so bad compare to suse at first glance
[04:33] <Amaranth> slomo: wanna be a guinea pig for my translation stuff?
[04:33] <infinity> Lathiat : Not linking libstdc++
[04:33] <slomo> Amaranth: depends... what shall i do?
[04:34] <Lathiat> infinity: does having the '.cpp' extension do some magic in gcc?
[04:34] <Lathiat> infinity: i mean, it compiles fine if its renamed to .c 
[04:34] <Amaranth> slomo: in a minute i'll put up a new beta of alacarte with your old german translation in it, you download it, update the translation, install it, and see if it all works
[04:34] <slomo> Amaranth: ok, fine :)
[04:35] <infinity> Lathiat : Err, in both cases, it's being compiled with gcc, not g++?
[04:35] <Amaranth> hrm
[04:35] <infinity> Lathiat : I would guess so.
[04:35] <Lathiat> infinity: yes
[04:35] <Amaranth> i don't know where to install the translations
[04:35] <Lathiat> infinity: and if i compile the .cpp with g++ its fine too
[04:35] <Lathiat> its just compiling .cpp with gcc throws that error
[04:35] <Lathiat> indeed -lstdc++ fixes it
[04:35] <infinity> Lathiat : That's new behaviour of gcc to automagically use C++-mode for .cpp files.  It would also explain another bug I've seen recently.
[04:35] <Lathiat> that seems kinda broken
[04:35] <infinity> Hrm.
[04:36] <Lathiat> what bug is that?
[04:36] <infinity> Well, basically the same bug.
[04:36] <Lathiat> infinity: looking at man gcc
[04:36] <Lathiat> infinity: it says this behavior is correct
[04:36] <Lathiat> line 480ish
[04:36] <infinity> That compiling MySQL with 'gcc -fno-exceptions' used to work, and now it doesn't, cause it wants libstdc++ (though it shouldn't)
[04:36] <Lathiat> infinity: ah
[04:37] <Lathiat> at a guess its throwing it through the c++ preprocessor
[04:37] <Lathiat> rather than the C preprocessor
[04:37] <Lathiat> whats the name of the c++ preprocessor?
[04:38] <infinity> cpp? :)
[04:38] <Lathiat> ah
[04:38] <infinity> But it's then shoving you off to cc1plus instead of cc1, probably.
[04:39] <Lathiat> that doesnt exist?
[04:39] <Lathiat> ah its in /usrlib
[04:47] <freeflying> kaimon: how to make grub display graphically after use this patch
[04:49] <Kamion> freeflying: splashimage directive, if I remember correctly; but #ubuntu for support questions, please
[04:58] <Amaranth> slomo: http://dev.realistanew.com/alacarte-0.8beta4.tar.gz
[04:59] <Amaranth> slomo: extract it, update de.po, run sudo ./setup.py install, see if you actually get everything translated
[05:00] <Amaranth> slomo: sorry it took so long, this installer stuff is a PITA
[05:01] <slomo> yes, it's even worse than autofoo ;)
[05:03] <Amaranth> slomo: also, if you're running alacarte while it's translated into german in Help->About the credits button should pull up a dialog with your name in it
[05:03] <slomo> yes
[05:03] <Amaranth> slomo: that's the translator-credits string if you want to change it
[05:04] <Amaranth> it works?
[05:04] <slomo> no idea, i'm currently translating the missing strings ;)
[05:04] <slomo> but i already noticed it
[05:04] <Amaranth> ah
[05:06] <slomo> Amaranth: "A menu can't be named \"Other\"." <--- is Other to be translated?
[05:06] <Amaranth> oh, no
[05:07] <Amaranth> ooh, WINE 0.9 coming tuesday
[05:10] <slomo> Amaranth: you'll get a patch for some sources ;)
[05:10] <Amaranth> eh?
[05:11] <slomo> Amaranth: translations must be enabled for glade
[05:11] <Amaranth> *facepalm*
[05:13] <pvh> I keep getting "kernel-image not in control info" errors when trying to build a custom kernel using make-kpkg.
[05:13] <pvh> I believe this is because Debian calls the custom kernel package "kernel-image", and Ubuntu calls it "linux-image", but I can't see how to specify that I want the resulting package to match the debian/control specs.
[05:14] <pvh> Can anyone advise me on what I'm doing wrong?
[05:15] <Amaranth> whoa, glade won't open my file
[05:15] <Kamion> pvh: --stem linux
[05:15] <pvh> Kamion: Thanks, I'll try that.
[05:15] <Kamion> Debian's moving to linux-image too, so this will probably get cleared up eventually
[05:16] <pvh> Kamion: I won't file a bug report then.
[05:16] <Amaranth> slomo: i have them enabled for glade
[05:16] <slomo> Amaranth: look at the query ;)
[05:16] <pvh> Kamion: That did it! I've been trying to figure that out for two days now.
[05:21] <pvh> Kamion: If I want to build a kernel that can coexist peacefully alongside another existing kernel, will I need to manually edit the debian/control file? I haven't read anything about having to do that before.
[05:22] <Kamion> pvh: I'm not an expert on this, was just randomly answering a question I knew :)
[05:23] <Kamion> I believe you can use --append-to-version to change the package name; I'd rather strongly recommend avoid editing debian/control by hand when using make-kpkg, as it all gets rather confusing
[05:23] <pvh> Kamion: tell me about it. Thanks for your help anyway.
[05:23] <Kamion> np
[05:58] <carlos> doko, hi, around?
[06:00] <jdub> jbailey: ping
[06:01] <jbailey> jdub: pong
[06:02] <mdz> morning
[06:02] <doko> carlos: yes, but not long, we have to leave the booth now
[06:02] <pitti> Hi mdz
[06:03] <carlos> doko, it's fast
[06:03] <carlos> doko, now that breezy has been released, could you take sometime to sync the spec at https://launchpad.net/products/launchpad/+spec/rosetta-openoffice-support with current implementation so we can approve and set as implemented it?
[06:04] <carlos> doko, no need to do it today, just remember to do that
[06:04] <doko> carlos: ok
[06:04] <carlos> doko, thank you
[06:04] <sivang> ogra: wow, edubuntu wiki is so visually appealing :)
[06:05] <ogra> :)
[06:06] <jdub> jbailey: is montreal canada/atlantic timezone?
[06:06] <jbailey> jdub: No, canada/eastern
[06:06] <jdub> thanks
[06:26] <moyogo> seb128: could the patch actually be tested and included in pango on breezy?
[06:26] <moyogo> seb128: this is a major bugfix for some african languages and for IPA
[06:26] <moyogo> seb128: ... except there's no font to actually use the feature :(
[06:26] <seb128> no
[06:27] <seb128> that's quite a big change
[06:27] <moyogo> ok
[06:27] <seb128> not something you want to push to a stable version
[06:27] <moyogo> cool
[06:27] <moyogo> when is dapper in developpement then?
[06:28] <Mithrandir> "soon"
[06:28] <Kamion> next week sometime, we hope
[06:30] <moyogo> sweet
[06:43] <HiddenWolf> Kamion, decided on the name of the test releases already?
[06:44] <Lathiat> "broken stuff"
[06:45] <mvo> Keybuk: I was wondering if dpkg stores md5sum of the original conffiles of a pkg on the fs somewhere. or would that be a stupid idea?
[06:46] <Lathiat> its not in the standard apckage md5sum file?
[06:46] <Lathiat>  /var/lib/dpkg/info/*.md5sums ?
[06:47] <Lathiat> they seem to be
[06:47] <Lathiat> at least for some packages
[06:47] <Keybuk> mvo: do you mean of the original package file?
[06:47] <Keybuk> or the original contents of the package file?
[06:47] <Keybuk> dpkg stores md5sums of configuration files in status
[06:47] <Keybuk> some packages ship an md5sums control file which ends up as *.md5sums -- but that's not actually used or understood by dpkg
[06:47] <Lathiat> ah, so it does
[06:48] <mvo> Keybuk: great, that's what I was looking for (but didn't rememberd)
[06:48] <Lathiat> Keybuk: is that md5sum taken postinst?
[06:48] <Keybuk> Lathiat: "taken" ?
[06:48] <Lathiat> calculated
[06:48] <mvo> Keybuk: thanks!
[06:48] <infinity> mvo : It would have to, how else do you think it handles conffile updates? :)
[06:48] <Lathiat> as in, if the packages postinst messes with config files, does it md5 before or after that?
[06:48] <Keybuk> it's calculated during unpack
[06:48] <Lathiat> ok
[06:48] <Keybuk> package postinst's are forbidden from messing with conffiles
[06:49] <Keybuk> anyone caught doing it has their hands chopped off
[06:49] <Keybuk> anyone caught doing it again has their package chopped off :p
[06:49] <infinity> If only...
[06:49] <sivang> Keybuk: cool, that's pretty good
[06:49] <Lathiat> whats the proper definition of a conffile?
[06:49] <infinity> Lathiat : A config file that's handled by dpkg.
[06:49] <Lathiat> ah ok
[06:50] <Keybuk> Lathiat: conffile = a file registered with dpkg as a conffile; appears in the conffiles control file and is managed by dpkg
[06:50] <Lathiat> as opposed to say /etc/X11/xorg.conf, gotcha
[06:50] <Keybuk> not all configuration files are conffiles
[06:50] <Keybuk> in particular, if a package wants to ditz about with it in postinst, it is not allowed to register it as a conffile
[06:51] <sivang> Keybuk: what is X11/xorg.conf then ?
[06:51] <Keybuk> (because dpkg would issue a conffile prompt [the Y/N/D/Z thing]  every single upgrade)
[06:51] <Keybuk> sivang: it's a file on the disk
[06:51] <infinity> Keybuk : Any plans to eventually pull ucf's functionality into dpkg, so we can have a decent C implementation (and something blessed by policy)?
[06:51] <Keybuk> infinity: no, not ucf
[06:51] <sivang> Keybuk: but some package puts is there, no? :)
[06:51] <Keybuk> sivang: it's not owned by any package, no
[06:51] <infinity> Keybuk : Note I said the functionality, not the code.
[06:51] <Keybuk> it's most likely written by one of the X postinsts
[06:51] <sivang> Keybuk: ah, created on runtime by dexconf?
[06:51] <Keybuk> I can't recall which one, maybe -common
[06:51] <infinity> Keybuk : ie: Tha ability to magle conffiles, but still have them registered in their mangled states.
[06:52] <Keybuk> infinity: http://www.dpkg.org/Classes
[06:52] <infinity> xserver-xorg's postinst does xorg.conf
[06:52] <HiddenWolf> infinity, shush ;)
[06:52] <sivang> hmm, /me finds yet another expection for backing up conffiles....
[06:52] <sivang> thoe "not owned by any package" files
[06:53] <sivang> maybe get a hardcoded list of those..
[06:54] <infinity> Keybuk : Cool.  Got a reference implementation yet? :)
[06:54] <infinity> sivang : You should back up all of /etc anyway.
[06:55] <Keybuk> infinity: yes :p
[06:55] <Keybuk> shockingly
[06:55] <Lathiat> a lot of /etc is crap tho
[06:55] <infinity> Sweet.  Is this dpkg 2.0 stuff, or feasible in the near (next year or so) future?
[06:56] <infinity> Lathiat : Only due to gconf living there, which should be moving RSN.
[06:56] <Lathiat> i tend to see a whole bunch of other crap that im not particularly interested in backing up
[06:56] <Lathiat> alternatives
[06:56] <infinity> Lathiat : Without the evil that is gconf, /etc is tiny, and "crap" or not, you're better off with a "better safe than sorry" attitude to grabbing it with your backups.
[06:56] <Lathiat> lots of hotplug
[06:56] <Lathiat> infinity: oh, sure
[06:56] <sivang> infinity: I want to be able to backup only things that  user changed ....
[06:57] <infinity> alternatives is a mess of symlinks.  It's it hurts to copy those. :)
[06:57] <Lathiat> udev rules, a whole whack of X11 crap
[06:57] <Keybuk> infinity: probably 2; on the basis it's from-scratch code
[06:57] <infinity> s/It's/Like/
[06:57] <Keybuk> sivang: beware of the "files in /etc that are obsolete conffiles" trap too
[06:58] <infinity> sivang : Then you need to start with a system snapshot of what you believe is "clean", and work from there.  The "full+incremental" backup model.
[06:58] <infinity> sivang : There's no other way you'll maigcally know what they did and didn't change.
[06:58] <sivang> infinity: ah right :)
[06:58] <sivang> infinity: me and mvo discussed the MD5 way :)
[06:58] <infinity> Right, but you need "unchanged" MD5s of everything first.
[06:58] <Kamion> sivang: "dpkg-registerfile" (let postinsts register files with dpkg so that at least it knows who owns them) was proposed about a zillion years ago but hasn't made it into implementation yet
[06:58] <infinity> Which takes you back to "full+incremental"
[06:59] <infinity> No initial snapshot, no dice.
[06:59] <sivang> e.g., comparing right, and as Keybuk noted, they are not held for any package
[06:59] <sivang> Kamion: I see
[07:00] <infinity> sivang : I suppose you could just do incremental for everything you have md5sums of, then "everything else"  (ie: all unregistered files would be considered changed)
[07:00] <infinity> sivang : That would work well enough.
[07:02] <sivang> infinity: yes, that seem as a sane defulat
[07:03] <infinity> I assume you're only using md5sums for conffiles, you're not summing everything in /usr.... Right? :)
[07:03] <infinity> Cause that'd be evil.
[07:04] <infinity> I'm all about dpkg --get-selections as a backup tool, and people wh omodify packaged binaries can cope with the loss. :)
[07:05] <Lathiat> debsums does the whole md5sum everything
[07:06] <infinity> Lathiat : Yes, but a backup tool that CHECKS md5sums of the whole system would be painfully slow and useless.
[07:06] <Lathiat> ah
[07:06] <Lathiat> right
[07:07] <sivang> infinity: that's the Kamion argument :) my vision is to have classes of snapshots, so for each membet of MarketingTeam, install snapshot:marketing-workstation :)
[07:07] <sivang> infinity: not just "plain" bakcup..
[07:07] <infinity> I mean, okay, a backup tool with a checkbox that says "I'm a really bad UNIX admin, and I put files ALL OVER MY DISK, so check it all" would be fine.
[07:07] <Lathiat> heh
[07:07] <infinity> But the default should be "backup /etc, home, /usr/local, the good bits from /var, and a package list"
[07:08] <Lathiat> or /var/lib/mysql
[07:08] <infinity> Take all of /var, if you're paranoid.
[07:08] <infinity> "The good bits" required actuall remember what's there, afterall.
[07:08] <infinity> s/required/requires/
[07:08] <Keybuk> sivang: one other thing that's worth making clear
[07:08] <Keybuk> dpkg is a bit crap and litters /etc with obsolete conffiles
[07:08] <Keybuk> if a package drops a conffile, dpkg doesn't remove it
[07:08] <infinity> Keybuk : Which are often wanted.
[07:08] <Keybuk> so watch out for those too
[07:08] <Keybuk> yeah
[07:09] <Keybuk> some of them are conffiles that have been turned into "edited by postinst" files
[07:09] <Keybuk> and some of them really are supposed to have been deleted, but the maintainer didn't in postinst
[07:09] <infinity> More fun, though, is backing up the "user deleted the conffile" state.
[07:09] <Keybuk> ooh, didn't think of that one
[07:09] <Keybuk> yes, you need to back that up to <g>
[07:09] <Keybuk> deleting a conffile is a user change
[07:09] <infinity> I assume you'd check for the file, if it doesn't exist, reocrd that, then when restoring, install the package and remove the conffile.
[07:09] <Keybuk> ie. you might delete something from /etc/cron.daily
[07:10] <sivang> Keybuk: right 
[07:10] <sivang> damn :) that's a pandora box 
[07:10] <Keybuk> (dpkg preserves that, if you upgrade it won't put the conffile back -- a source of many people's miseries
[07:10] <Kamion> you could reconstruct that from /var/lib/dpkg/status plus your backup of /etc
[07:10] <Keybuk> ie. I wiped /etc, reinstalled all the packages, and it didn't put any conffiles back, what gives?)
[07:10] <sivang> yes, I mysself have been bitten by this
[07:10] <Lathiat> yes thats caught me out a few time
[07:10] <Lathiat> --force-confmiss or something i think 
[07:10] <Keybuk> indeed
[07:10] <Lathiat> only learnt that recently
[07:10] <sivang> Lathiat: right :)
[07:10] <Lathiat> i had many pains about that in previous years
[07:11] <sivang> Lathiat: me too, from mvo :-)
[07:11] <Lathiat> yeh i read it in here
[07:11] <infinity> Kamion : If you backed up all of /etc, but sivang wants to only backup "the bits that changed", which is where the fun comes in.
[07:11] <sivang> infinity: indeed :)
[07:11] <Keybuk> these days I go out of my way to never change /etc
[07:11] <infinity> (It's really much less hassle to juct backup all of /etc)
[07:11] <Kamion> infinity: if it's full+incremental you can still work it out, as long as you can represent deletions in the incrementals
[07:11] <Kamion> (which, surely, you can ...)
[07:12] <infinity> Indeed.
[07:12] <infinity> I argued this already. :)
[07:12] <Keybuk> of course, being a developer on the distro you use, so you can change the package defaults to match your taste _helps :p
[07:12] <infinity> He seems to be talking something fancier and/or crazier.
[07:13] <sivang> Keybuk: I'm not sure I'm following you ... :)
[07:13] <Keybuk> sivang: which bit?
[07:13] <sivang> Keybuk: the "taste _help" one
[07:13] <Keybuk> oh
[07:13] <Keybuk> just muttering about me never changing /etc
[07:13] <infinity> Keybuk : I just have apt configured to deny all new conffiles by default, then a post-invoke hook to show me find /etc -name \*.dpkg\*
[07:14] <infinity> Keybuk : So I can peruse the diffs at my leisure, rather than ACT ON IT RIGHT NOW, DAMNIT.
[07:14] <Keybuk> the only file I had to change when installing my desktop was /etc/hdparm.conf -- and I'm sure we can automate that one in dapper or dapper+1 :p
[07:15] <sivang> infinity: hmm, so maybe work the other way around , track in a way similar to what you've mentioned, all the changed stuff. that prepares stuff in advance, instead of borking when trying to create a snapshot
[07:18] <sivang> Keybuk: so anyway, every packages should have that .md5sums file in /var/lib/dpkg/info/ and if its not there, I can always check status ?
[07:19] <Keybuk> sivang: no, there's no requirement for an md5sums file in info
[07:19] <Keybuk> and for all you know, the contents could be the recipie for the author's meatloaf
[07:19] <sivang> lol
[07:20] <infinity> Yeah, the md5sums aren't always right (well, these days, if they're there, they seem to be right, but in the past there were oddities with that file not matching reality)
[07:20] <infinity> But you can't count on them in any way, really.
[07:22] <sivang> I wonder if given all this, starting with something that stores package selections, together with possibly gconf settings of a user/system can be a good starting point towards the snapshots/usage classes eventually..
[07:23] <sivang> e.g., starting small which I intend to do anyways, just was nice to get some feedback re: the grand scheme
[07:23] <sivang> (thanks all)
[07:35] <Kinnison> ciao all
[08:21] <dilinger> http://www.zwane.com/blog/?p=59
[08:21] <dilinger> how many people are taking one of those to ubz? :)
[08:23] <sivang> dilinger: ROTFL
[08:23] <sivang> :)
[08:39] <crimsun> HiddenWolf: pong
[08:39] <HiddenWolf> crimsun, It seems /all/ audio on my system is delayed by like 2 seconds.
[08:39] <HiddenWolf> crimsun, any clue how I can figure out why?
[08:40] <crimsun> HiddenWolf: ->#ubuntu
[09:05] <\sh> jdub: pign
[09:05] <\sh> ping even
[09:05] <jdub> \sh: pong
[09:06] <\sh> jdub: did u change something today on the planet?
[09:07] <\sh> jdub: reloading or something? ,-)
[09:07] <jdub> nup
[09:07] <\sh> jdub: or should I change to atom 0.3 / atom 1.0?
[09:07] <jdub> i don't have access to it
[09:18] <zul> did the archive go caput?
[09:19] <crimsun> Fetched 22.7kB in 2s (11.2kB/s), seems fine to me.
[09:20] <zul> seems to be just me then
[10:35] <mdz> fabbione: https://launchpad.net/distros/ubuntu/+spec/installer-volume-management
[10:35] <mdz> fabbione: you don't really mean "by default", do you?
[10:44] <ajmitch> yay, I have a passport so I can get to UBZ now ;)
[10:47] <Treenaks> ajmitch: wooohoo :)
[10:47] <ajmitch> Treenaks: I put in the application a bit late, my old passport was a bit water-damaged
[10:48] <Treenaks> ajmitch: mine looks used, but isn't
[10:48] <Treenaks> ajmitch: but then, it's 4 years old
[10:48] <ajmitch> I'll be travelling through the US, I've heard they don't like passports that look worn, have bad photos, etc :)
[10:49] <ajmitch> which mine did
[10:49] <Treenaks> ajmitch: I'm going from Amsterdam to Heathrow, and then straight to Trudeau (Montreal)
[10:49] <crimsun> definitely don't want a warn passport
[10:49] <crimsun> worn
[10:49] <crimsun> geez, mind is befuddled today
[10:49] <Treenaks> crimsun: not worn, just.. well, used