[00:03] <crimsun> classvoid: other channel, please.
[03:53] <psusi> woohoo... sped up e2defrag by 25-50%!
[03:58] <hyperair> there's an e2defrag?
[04:04] <psusi> yea... the old defrag package.. was written by linus, ted tso, and remmy card in the 90s... debian and ubuntu dropped the package 2 years ago because it had not been maintained for like a decade... I rescued it and have been updating it to work on ext4 and combining it with ureadahead to speed up boots
[04:05] <psusi> it has a super sexy ncurses gui ;)
[04:05] <psusi> was cool as hell when I used it the first time on slackware in 95 ;)
[04:06] <hyperair> ooh that's cool.
[04:06] <hyperair> but isn't there an e4defrag around already?
[04:07] <psusi> they are working on it.... who knows when it will be usable... but it's for doing online defrag... so far they are just trying to get it to defragment files, does not defrag free spcae or try to relocate files to better places
[04:07] <psusi> basically it just copies a file and hopes it is less fragmented than the original, then replaces the original with the copy
[04:08] <psusi> e2defrag is an offline defragmenter... works on the raw block device when it is not mounted, and totally reorganizes the fs, packing all blocks as far left as possible, and 100% in order... and really, really, really tries to only move any given block once
[04:09] <psusi> it also can take a list of files to prioritize and pack first... like those read by ureadahead
[04:09] <hyperair> ah i see. that's interesting
[04:09] <hyperair> cool =p
[04:09] <hyperair> now if that would work with lvm + cryptsetup, even cooler
[04:10] <psusi> I'm using lvm and my test volume on the old first gen 36 gig 10krm wd raptor raid0 array boots in 13 seconds with my changes to ureadahead and a good defragging
[04:10] <TheMuso> ooo nice.
[04:10] <hyperair> ooh
[04:11] <hyperair> but it needs unmounting, making it only feasible during initrd stage..
[04:11] <psusi> took the boot time on my new laptop from 35 to 30 seconds... it's only a 1.3 ghz celeron, so it spends most of its boot time cpu bound.. the ureadahad part now only takes 4-5 seconds
[04:11] <psusi> aye... which is why I'd like to get initramfstools to build an alternative initrd and add it to the grub boot menu
[04:11] <psusi> so you can choose from the grub menu to do a defrag boot
[04:14] <psusi> when I defragged the laptop I just had grub pass the break parameter to make the initrd stop so I could manually mount the root fs, copy the defragger to the root tmpfs, umount it, then defrag it, and reboot
[04:15] <psusi> the ncurses interface didn't work but I've figured out what needs to be included for that... just need to automate it now with initramfs-tools
[04:20] <hyperair> cool
[04:20] <hyperair> i look forward to testing it =p
[04:21] <hyperair> i wonder if btrfs would also have a defragging option..
[04:21] <arand_> psusi: Cheers for the help with looking at the fsck error before, it seems like Ts'o fixed up e2fsprogs within 30mins or so :) , And released a new version which is already in debian :D
[04:22] <psusi> ohh, wow.. you mean he looked at my comments and jumped right on it?  was my guess basically right?
[04:22] <xnox> arand_, =)))))) will it appear on lwn.net as a follow-up?
[04:23] <psusi> now if only he would implement my idea to fix fsck's handling of finding inodes from previous incarnations of the fs by checking if the ctime is < fs creation time and ignoring them if it is ;)
[04:28] <arand_> psusi: Again, I have no idea :) but this is the commit: http://pastebin.ubuntu.com/435904/
[04:32] <psusi> arand_, well, glad it's fixed ;)
[04:42] <psusi> ok... pushed my local e2defrag changes up to lp and I think it's about ready for another release, and some sponsorship to upload to the universe archive for maverick... well, there might be one more thing I can do to improve performance first...
[06:03] <superm1> cody-somerville, not really sure, what interface does it use?
[06:04] <superm1> i'm guessing it doesn't
[06:05] <cody-somerville> it sorta works but sometimes causes X to crash or my music applications to freeze up. sometimes it'll give me -1 value instead of correct temp as well for example. Sometimes it also takes two attempts to modprobe it.
[06:56] <imbrandon> superm1: gotta few minutes this evening ( or early morning depending on your sleep schedule ) , lol
[06:57] <imbrandon> for a /query
[07:45] <\sh> moins
[07:45] <\sh> does anybody know where keybuk hide his slides of his talk "the plumbing layer" from Tursday Plenary last week ?
[07:51] <dholbach> good morning
[08:19] <pitti> Good morning
[08:20] <ajmitch> morning pitti
[08:20] <pitti> hey ajmitch, how are you?
[08:22] <ajmitch> good, how are you?
[08:22] <ajmitch> recovered from UDS? :)
[08:40] <sabdfl> morning all
[08:40] <pitti> ajmitch: yes, wasn't that much to recover from; no jetlag this time
[08:40] <pitti> hey sabdfl, good morning
[08:55] <Usama> pitti, may I ask about the mlterm issue
[08:56] <pitti> Usama: you can ask in the channel, sure (probably not me personally since I don't know about mlterm)
[08:56] <Usama> https://bugs.launchpad.net/bugs/562130
[08:58] <Usama> it needs main inclusion request but many people (not RTL users) don
[08:58] <Usama> ... don't seem to like the idea
[09:00] <pitti> Usama: you can file one if you have a good reason for it to be in main
[09:02] <Usama> I don't want it to be in main but people are saying that it's necessary to have it in main if it will be installed when selecting RTL language
[09:04] <pitti> Usama: right, it could be hooked into language-selector
[09:04] <pitti> Usama: then the main question is if there is someone who will maintain it
[09:05] <Usama> some one already have built version 3 packages
[09:08] <Usama> would it be enough if some one build new versions and update it?
[09:09] <pitti> Usama: well, the point is there needs to be someone who will continuously test and look after it
[09:09] <pitti> and upload fixes if it breaks, or if there's a new version
[09:10] <pitti> this is not an one-off task
[09:10] <pitti> we can probably sync from Debian most of the time, so it's not a lot of work
[09:10] <pitti> Usama: debian has 3.0.0, FYI
[09:10] <pitti> so this will be autosynced into maverick anyway
[09:10]  * pitti does an autosync run, he needs the new postgresql
[09:11] <Usama> some one said that debian do install it by default for the Hebrew language but I couldn't check if it's true
[09:12] <Usama> could you help with that. it might support my case
[09:14] <pitti> I don't know, but I'm not sure that Debian even has magic like this
[10:05] <ct529> hi everybody. I have upgraded to 10.04. I have nvidia quadro 1600 m. I tested the graphic drivers thoroughly, both os and proprietary. There are problems and conflicts on libgl and lbglx, between mesa and the driver. I believe they need to be fixed.
[10:06] <dim> Hola!
[10:08] <RAOF> ct529: How did you install the drivers?  Sadly, those problems are essentially unfixable, although we have systems in place to make it as little trouble as possible.
[10:09] <dim> Hey I'm interested in finding some resources for doing some more basic linux dev stuff
[10:09] <dim> like right now im thinking I should make a program that manually interprets /dev/input/mouse0 or something
[10:09] <dim> but im not really sure where to start, should I just make a python program read from that device?
[10:11] <RAOF> You could certainly do that.
[10:11] <dim> RAOF, alrighty then
[10:11] <dim> is this the right place for those kinds of questions?
[10:11] <dim> or an acceptable place at least
[10:12] <RAOF> Not really, no :)
[10:12] <RAOF> “Development of Ubuntu (not support, not app development)”
[10:12] <dim> hm, alright
[10:12] <ct529> RAOF: I made a thorough test, by trying all the possible combinations in all possible orders.
[10:13] <dim> well I'm also in #ubuntu-app-devel
[10:13] <RAOF> That's probably a better place.
[10:13] <ct529> RAOF: I do realise you are right, it is a problem difficult to fix.
[10:13] <ct529> RAOF: I think there is actually a chain of problems.
[10:14] <RAOF> Well, actually it's a problem *impossible* for us to fix; we don't have access to the source.
[10:15] <RAOF> It should work well as long as you use the the jockey tool - System→Administration→Hardware Drivers under Ubuntu.  Somewhere else for Kubuntu.
[10:15] <ct529> RAOF: well, you have access to the mesa and the os driver source I believe.
[10:15] <ct529> RAOF: and the proprietary driver installs a clear set of files, so it is potentially possible to fix it.
[10:16] <ct529> RAOF: but I am not saying it is practically possible.
[10:16] <ct529> RAOF: but I would like to look into it, if you do not mind.
[10:16] <RAOF> The proprietary driver replaces parts of X with its own code, that doesn't work with the non-proprietary drivers.
[10:16] <RAOF> You're welcome to look into it if you like.  I don't think it will be productive.
[10:16] <RAOF> Anyway… dinner.
[10:17] <ct529> RAOF: it does not work well, even using the jockey tool.
[10:30] <chrisccoulson> could someone please reject the gnome-screensaver SRU that i uploaded to lucid-proposed?
[10:30] <chrisccoulson> i'm just talking to a user and it doesn't completely fix one of the bugs it's meant to fix
[10:30] <pitti> chrisccoulson: *flush*
[10:31] <chrisccoulson> pitti - thanks :)
[10:32] <vandenoever>  /usr/lib/pkgconfig/libsoup-2.4.pc has the wrong version number
[10:33] <vandenoever> (in karmic)
[10:34] <lool> james_w: Heya, it seems the package import issue I mentionned yesterday affects a bunch of packages, rendering merges quite unpractical
[10:34] <james_w> lool: which issue was that?
[10:35] <lool> james_w: Some packages were not imported on LP
[10:35] <lool> james_w: Sorry, that was two days ago, not yesterday
[10:35] <james_w> lool: the LP mirror of Debian isn't up to date?
[10:35] <lool> james_w: Yes, I think that's what you mentionned
[10:36] <ajmitch> it's been that way for a month or more
[10:36] <james_w> right, there's not a lot I can do about that, the issue is apparently on the Debian side
[10:36] <lool> That's a problem for bzr merges
[10:36] <lool> james_w: Ouch, I think I know what you mean
[10:36] <lool> Is it mirroring experimental?
[10:36] <james_w> yeah
[10:36] <lool> My debmirrors have been failing since a month+ due to a bug in apt
[10:36] <ajmitch> I think it's mirroring in general now
[10:37] <lool> james_w: How can I check whether it's mirroring?
[10:37] <ajmitch> there's a new apt-ftparchive for lenny, ftpmasters just need to install it afaik
[10:37] <StevenK> Or DSA
[10:38] <james_w> lool: you can check individual packages to see if they are up to date
[10:38] <james_w> I don't know if there is a status page
[10:38] <james_w> ...or you could ask StevenK :-)
[10:38] <StevenK> The Debian mirror for LP is also failing to update due to the same error
[10:39] <lool> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=577759
[10:39] <ajmitch> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=577759 is tracking the progress on that side
[10:39] <lool> stable has 0.7.20.2+lenny1 and proposed-updates lenny2
[10:39] <ajmitch> StevenK: poke a few people please :)
[10:40] <StevenK> ajmitch: I did, at UDS.
[10:41] <ajmitch> maybe not enough beer was used
[10:42] <StevenK> I'm not about to fly to Germany and sit on Ganeff's lap until he fixes i
[10:42] <ajmitch> heh
[10:44] <lool> ajmitch, james_w, StevenK: I've mailed DSA and Ftpmasters to point at the update, requesting to install it
[10:46] <StevenK> lool: With an RT ticket?
[10:46] <Riddell> vandenoever: Version: 2.28.1  that's wrong?
[10:46] <lool> StevenK: No
[10:46] <vandenoever> Riddell: actually, my mistake, it's correct, i was confused with 2.2
[10:52] <Omahn> I would like to fix a simple bug in vsftpd that currently has a version of 2.2.2-3ubuntu6, should my debdiff with the fix be versioned 2.2.2-3ubuntu7 or 2.2.2-3ubuntu6-1 or something else?
[10:56] <raphink> Omahn:  2.2.2-3ubuntu6-1 is not a valid version number
[10:56] <raphink> ubuntu7 is good
[11:06] <Omahn> raphink: Great, thanks.
[11:43] <gartral> who's bloody idea was it to release empathy without a window title that says empathy? i see 25 people a day in #ubuntu asking "what's the name of this chat application in ubuntu?"
[11:45] <Usama> this is the first time I notice that :)
[11:45] <Usama> there is the about button I guess
[11:45] <pitti> Help->About says it quite clearly, though, as well as the app menu
[11:45] <gartral> i did it twich today cause it has such an obscure frigging name
[11:55] <chrisccoulson> pitti - i reuploaded gnome-screensaver if you have time for some SRU review later ;)
[12:07] <apw> pitti, i am about to get expired out of work-items-hackers, seems its not something i can fix myself ... could you
[12:07] <pitti> apw: done 10 mins ago, and fixed the option for self-renewal
[12:08] <apw> pitti, thanks :)
[12:08] <pitti> yw :)
[12:08] <pitti> I don't want to loose my best contributor!
[12:11] <apw> pitti, we have a bunch of blueprints over on another project, which have some WIs for people on my team.  am wondering if that is something we can cope with
[12:11] <pitti> apw: ubuntnu-arm?
[12:11] <apw> yeah
[12:11] <pitti> apw: yesterday I implemented support for project WIs
[12:12] <pitti> apw: you mean like this? http://people.canonical.com/~pitti/workitems/ubuntu-arm/
[12:12] <pitti> :)
[12:12] <apw> yeah like that :)
[12:12] <apw> will the ones for people on my team blead over or is it separate
[12:13] <pitti> apw: if you have WIs assigned there, they should appear
[12:13] <apw> awsome
[12:13] <pitti> we don't do the per-team thing for projects
[12:13] <pitti> well, at least the configuration for it doesn't do
[12:13] <pitti> (we potentially could)
[12:13] <pitti> it's just "all"
[12:14] <apw> yeah, for now that likely is plenty for over there
[12:14] <apw> and if the items bleed to my charts, rocking is all i can say
[12:15] <pitti> apw: btw, you don't happen to have a recent lucid.db laying around somewhere?
[12:15] <pitti> (cf my platform mail about destroying it by accident)
[12:16]  * apw looks around, i think i renamed it to current.db and it got zapped by maverick
[12:17] <pitti> darn
[12:17] <pitti> the current workitems/lucid/ stuff has the #WIs all correct, but the history (and charts) are gone
[12:18] <pitti> I now know how it happened at least
[12:18] <apw> pitti, i have this
[12:18] <apw> -rw-r--r-- 1 apw apw 6907904 2010-01-27 10:42 lucid.db
[12:18] <apw> which isn't very new
[12:18] <pitti> hm, very old
[12:18] <pitti> apw: ok, thanks for checkign
[12:18] <apw> i'll have a little look around, my cron job was collecting it for the longest time
[12:19] <fabbione> Keybuk: the more I read 94065, the less I udnerstand why they are duplicated
[12:19] <fabbione> Keybuk: maybe there is implementation overlap
[12:19] <apw> pitti, i assume your one got broken: http://people.canonical.com/~pitti/workitems/lucid/lucid.db
[12:19] <fabbione> Keybuk: but what I understad in 94065 is that they want the equivalent of removing rcX.d/ symlinks
[12:20] <pitti> apw: right
[12:20] <fabbione> Keybuk: mind to explain why are they the same?
[12:20] <pitti> hey fabbione
[12:20] <fabbione> hi pitti
[12:20] <Keybuk> fabbione: 94065 is a wishlist for "disabling a job", which is exactly what your bug was
[12:20]  * pitti lunch &
[12:20] <fabbione> Keybuk: we are asking 2 different things... even if it can be achieved by similar code :)
[12:21] <fabbione> otherwise why would I bother to file a bug? :)
[12:21] <Keybuk> maybe I didn't read your bug hard enough, what were you asking for?
[12:21] <fabbione> Keybuk: from boot options.. they are asking as normal maintainance thingy.. similar to symlink handling for old rcX.d/
[12:22] <fabbione> Keybuk: the concept might be similar in terms of internal implementation (AFAICT anyway)
[12:22] <fabbione> Keybuk: but the user interface looks different to me
[12:23] <fabbione> Keybuk: anyway.. as long as you can give it a thought.. then if it's referenced by bug A or B... matters not
[12:24] <Keybuk> yeah, yours is somewhat of an overlap of 94065 and 389113,
[12:24] <Keybuk> I understand what you're asking for - "don't start this service in single-user mode" kind of thing
[12:24] <Keybuk> it's certainly on the roadmap ;)  just doesn't quite map onto the existing bug#s
[12:24] <fabbione> hmm i wasn't presented 389113... need to read that
[12:24] <fabbione> no, even if i am booting into 2 or 3 or 5
[12:25] <fabbione> don't start this service at boot. <- full stop :)
[12:25] <Keybuk> right, 389113 covers being able to put things like "noapache" on the kernel command-line
[12:25] <fabbione> then optionally, you can say at this runlevel
[12:25] <Keybuk> and having upstart disable that job
[12:25] <fabbione> yeah something like that
[12:25] <fabbione> i didn't see it in the list
[12:25] <fabbione> noapache sounds scary :)
[12:26] <fabbione> ok well.. good to know I am not the only one with crazy idea
[12:26] <fabbione> +s
[12:27] <Keybuk> :-)
[12:28] <fabbione> tho I admit, in a cluster enviroment it is extremely useful
[12:28]  * fabbione will go for the workaround in the meantime
[12:30] <cjwatson> pitti: I have: -rw-r--r-- 1 cjwatson cjwatson 34606080 2010-04-10 01:18 lucid.db
[12:40] <apw> i have gotten a new machine which was installed from a usb stick, installed and rebooted normally.  did an update and grub updated and it popped a gconf popup for grub which asked me to select grub install location, i ignored it and hit forward, then realised what it was
[12:40] <cjwatson> cody-somerville: I've merged base-installer now
[12:41] <apw> 1) back was not possible, and 2) having gone forward i ended up unbootable. booting from the USB i oddly get booted from the disk now, but grub-install works but does not make me bootable
[12:42] <cjwatson> apw: a number of people have reported that; what I really need is a recipe for reproducing the previous state, since by the time people report it they've typically lost the information I need
[12:42] <apw> that i can believe
[12:43] <apw> i am suspicious that the trigger here is that the disk has moved from sda to sdb when i have the usb stick in
[12:44] <apw> hrm.  i simply cannot get grub-install to install a boot block
[12:44] <jussi> join #ubuntu-artwork
[12:44] <cjwatson> pitti: I've put that lucid.db in my home directory on lillypilly, in case it's useful
[12:44] <jussi> argh
[12:44] <apw> it says it has, but ... it does not boot afterwards
[12:44] <cjwatson> apw: it's supposed to use /dev/disk/by-id/ paths
[12:45] <cjwatson> but 'sudo grub-install /dev/sdwhatever' should work ...
[12:45] <apw> now that i've broken it its not happy about fixing itself for sure
[12:45] <apw> that says it works, but the machine does not become bootable as a result
[12:45] <apw> gah
[12:45] <tjaalton> cjwatson: are you planning to migrate console-setup to keyboard-configuration for maverick?
[12:45] <cjwatson> then I suppose it's actualy booting from wqsomewhere else
[12:45] <cjwatson> tjaalton: probably
[12:45] <cjwatson> sigh, lag
[12:46] <cjwatson> tjaalton: it's on the "really nasty merges" list
[12:46] <tjaalton> cjwatson: heh, ok
[12:46] <apw> cjwatson, when i did the original install, it booted just fine so it used to be able to do it ... that was an official release CD image on stick
[12:46] <tjaalton> that one has to go right ;)
[12:50] <apw> cjwatson, ok this is perverse.  if i have the stick inserted it boots from the hard drive ... as if grub seconadary is coming from there
[12:52] <apw> cjwatson, can i ask grub where it loaded itself from?
[12:52] <cjwatson> apw: do you have a /boot/grub/device.map?
[12:52] <cjwatson> is it wrong?
[12:53] <cjwatson> "boots from the hard drive" - there are a bunch of things that might mean, you mean it loaded the kernel from the hard drive with a root= pointing there?
[12:53] <ogra_cmpc> cjwatson, https://wiki.ubuntu.com/Specs/M/ARMPreinstalledSDCardImages mind to take a glance if i miss anything obvious ?
[12:53] <apw> cjwatson, the machine does not boot from the USB without manual intervention
[12:53] <ogra_cmpc> (not urgent)
[12:53] <apw> but if i don't ask it to then the mahcine loads with the OS from the real drive when the USB stick is present, not when it is not
[12:54] <apw> when the stick is present the harddrive moved to (hd1)
[12:54] <cjwatson> ogra_cmpc: I don't have it all in my head, but it seems to match roughly what we discussed
[12:54] <apw> so i suspect the /boot/grub/device.map will be wrong
[12:54] <cjwatson> apw: if /boot/grub/device.map is there at all, I suggest removing it
[12:54] <ogra_cmpc> cjwatson, great, thanks
[12:54] <cjwatson> modern grub can work without it and frankly is better off without it
[12:55] <cjwatson> we're deprecating the (hd1)-style names
[12:55] <apw> cjwatson, not there anyhow ... hrm
[12:56] <cjwatson> silly question, are you sure that bios boot order is correct?
[12:57] <cjwatson> try running grub-install with 'sh -x' and send me the output
[12:57] <apw> heh new machine, seemed to boot windows before it was zapped, will check
[12:59] <apw> ok it does fall back to usb it seems after hard drive, so if its not on the hard drive at all it would do something, why the usb stick then confusingly boots the real disk ... i don't know ... though doesn't the CD have syslinx?  and i end up with grub
[12:59] <apw> so perhaps the usb stick got wacked with grub
[13:01] <cjwatson> I'm not sure I'm following
[13:01] <cjwatson> running grub-install on the USB stick should render it self-contained, assuming that the /boot/grub/grub.cfg on the USB stick is vaguely sane
[13:01] <apw> if we assume that the reason it boots with the stick installed is that the fallback order is disk/something/something/USB disk
[13:02] <cjwatson> and of course assuming that the BIOS is actually booting it
[13:02] <apw> then i'd expect to see the syslinux, as i get grub i assume one of the two spindles have grub
[13:02] <cjwatson> I don't follow where the musing about isolinux is coming from
[13:02] <pitti> cjwatson: thank you!
[13:02] <apw> the USB stick should be untouched, and thus have syslinux, but i am now suspicuus the install has wacked it with grub
[13:02] <cjwatson> you've been repeatedly running grub-install on things
[13:03] <cjwatson> oh, I see
[13:03] <cjwatson> I thought you *wanted* to boot from the USB stick
[13:03] <apw> yep, to be sure it happened then would require a scratch install i cannot now be sure it was me
[13:03] <apw> no i want my real disk to boot :)
[13:03] <cjwatson> Marjo reported a pretty similar bug during the release sprint
[13:03] <cjwatson> so I know that USB installs sometimes get the grub target wrong
[13:04] <apw> yeah its mentioning hd1, which the spinning disk is right nwo
[13:04] <cjwatson> forget about hd1
[13:04] <apw> (in the sh -x output)
[13:04] <cjwatson> that naming is (a) unhelpful (b) unused
[13:04] <cjwatson> can I see the output?
[13:04] <apw> working on it
[13:06] <apw> http://pastebin.com/YQpPrTdT
[13:08] <Keybuk> cjwatson: are we doing a team meeting today?
[13:08] <cjwatson> ok, so the hd1 stuff there is definitely a red herring - with a single-disk biosdisk configuration, it doesn't even end up in any configuration file
[13:08] <cjwatson> Keybuk: a quick one, no need for activity reports though since we've all been at UDS :)
[13:08] <cjwatson> I'll send a mail in a moment
[13:10] <apw> cjwatson, can i tell if grub is in the boot block trivially ?
[13:13] <apw> cjwatson, ok i think i have it sorted ... i had to make the linux root disk a boot partition in fdisk
[13:14] <cjwatson> apw: you can compare it with /boot/grub/boot.img; there'll be some differences for variable bytes and the partition table and such
[13:14] <cjwatson> grub doesn't care about active partitions, but maybe your BIOS does
[13:14] <apw> cjwatson, yeah i suspect the latter
[13:14] <apw> i am going to double check and confirm
[13:14] <cjwatson> I know of some BIOSes that refuse to boot from a disk unless it has an active partition
[13:14] <cjwatson> the installer is supposed to deal with that already though
[13:15] <apw> that is exactly the symptom i am suspecting, and the installer did not do that
[13:15] <cjwatson> well, if the installer was confused about which the boot disk was supposed to be (which is borne out by your comments above), then it wouldn't have known that it needed to take care to give it an active partition
[13:16] <cjwatson> so this is probably just all part of the same bug, which is the one Marjo saw
[13:16] <cjwatson> if I can manage to put together a test case in KVM, I want to try to fix that one for 10.04.1
[13:17] <apw> ok confirmed, removing the active flag has broken boot
[13:17] <Laney> :q
[13:17] <apw> so i think this is as you say a combination of known issues
[13:18] <cjwatson> right.  useful to have confirmed that there are still BIOSes with that property, though :-)
[13:18] <cjwatson> the last definite sighting I had of that was seb128's laptop a few years back
[13:18] <apw> cjwatson, this looks to be an efi bios underneath too
[13:18] <apw> so it may become more common not less
[13:18] <cjwatson> but of course once it was fixed in the installer we stopped caring much
[13:18] <cjwatson> lovely
[13:19] <apw> i has a ticky to switch modes, i am somewhat scared of trying it
[13:19] <cjwatson> guess I'd better get that patch sorted out and pushed to Debian at some point
[13:20] <apw> cjwatson, this machine also has the lovel M-P from the screen switch button
[13:21] <cjwatson> M-P?
[13:21] <apw> meta/windows-p
[13:21] <apw> Super_L down, P down, P up, Super_L up ... from pressing the output screen select button
[13:24] <cjwatson> ah
[13:25] <apw> cjwatson, i recon we should just give up and map that key to the monitors applet
[13:27] <txwikinger> Riddell: what needs updating in debootstrap?
[13:29] <Riddell> txwikinger: turns out it already supports maverick so nothing needs updating
[13:29] <Riddell> unless you want to do the merge with the current Debian version
[13:31] <cjwatson> please don't
[13:32] <cjwatson> a merge there is a waste of time - I'll do a new upstream release and then we can just sync
[13:56] <G> sorry asked a few minutes ago in -motu but I think here is more the right place, anyone able to take a look at https://bugs.launchpad.net/ubuntu/lucid/+source/libvirt/+bug/571093 to me it looks like a udev bug and I think some udev trained eyes might be a good idea :)
[13:58] <G> it's getting triggered by udev calls and udev is all over the valgrind output
[14:06] <seb128> cjwatson, slangasek, jdong: do you plan to do some SRU review soon? since pitti is on rotation SRU seems to be stucked a bit...
[14:06] <seb128> would be nice to get those moving at least to get fixes in the LTS
[14:07] <cody-somerville> cjwatson, thanks! :)
[14:32] <cjwatson> seb128: yikes, suppose I'd better
[14:32] <seb128> cjwatson, thanks!
[14:39] <apw> pitti, this dbg image reaping issue, can i confim that this filename is in the right form: linux-image-2.6.34-2-generic-dbgsym_2.6.34-2.9_amd64.ddeb
[14:44] <apw> pitti, we are getting pounded about them going missing
[14:45] <jcastro> Keybuk: are you planning on proposing to give a talk at debconf?
[14:46] <pitti> apw: that seems right; I suppose it's due to the old -image-debug hack that I placed on the ddeb machine
[14:46] <pitti> apw: I reverted it now
[14:47] <pitti> oh, hang on, we still need it for older releases
[14:47] <apw> yeah old releases still use the old form
[14:47] <pitti> apw: ok, I extended the hack, should cover both forms now
[14:48] <apw> pitti, thanks ... you are the man
[14:48] <pitti> let's hope that it works now
[14:48] <pitti> http://ddebs.ubuntu.com/pool/main/l/linux/ -> seems that's missing -22 :(
[14:48] <apw> yep, thats how we noticed that it wasn't working
[14:48] <pitti> apw: can you watch this for a bit with the maverick ones?
[14:48] <apw> yeah will be watching for a while
[14:48] <pitti> the hack should have kept the lucid -22 ones
[14:48] <pitti> unless -22 used -dbgsym already?
[14:49] <apw> lucid also uses the new form
[14:49] <pitti> i. e. did you change the name in updates?
[14:49] <pitti> aah
[14:49] <pitti> ok, that explains it
[14:49] <apw> good thanks, will write it up and close the bug out
[14:58] <amikrop> Hello, how can I run JVM source?
[14:58] <amikrop> What do I need to install?
[14:59] <joaopinto> amikrop, the support channel is #ubuntu
[14:59] <amikrop> ok thanks
[15:00] <cjwatson> seb128: there's no bug in the changelog for librsvg - how do we track verification in that case?
[15:03] <pitti> cjwatson: I usually reject those and ask to reupload with a bug #
[15:04] <cjwatson> oh, wait, I just found a bug number buried in there
[15:04] <cjwatson> hard to see amid the list of upstream bugs
[15:05] <pitti> cjwatson: you don't use queuediff? that opens the bugs automatically
[15:05] <cjwatson> no, my net connection is slow
[15:05] <pitti> ah, it's a new upstream version
[15:05] <cjwatson> I don't want to have to download all the files
[15:05] <pitti> if it's the same upstream version, that'll only download the diff.gzs and use interdiff
[15:06] <cjwatson> because it can take ages for me, my habit is not to use it
[15:07] <nigelb> pitti: is there a problem with the nondefault gconf value reporting via apport?
[15:07] <pitti> nigelb: define "a problem"?
[15:08] <nigelb> someone was writing a hook and he noticed that its reporting a lot more values that its supposed to be and most of them were normal
[15:08] <pitti> nigelb: not known to me
[15:08] <nigelb> pitti: I'll ask him to log a bug with exact details against apport
[15:10] <apw> cjwatson, ok confirmed that doing an install on this machine breaks the USB stick, grub ends up on it instead
[15:17] <juliank> mvo: Just want to inform you that software-center is 2.0.3debian1 in maverick (has been synced automatically). This means a bit of Ubuntu branding is currently missing in debian/control package description.
[15:17] <juliank> Apart from the description change, you can merge anything else though, so this is not a large issue.
[15:17] <pitti> so much for stealing Debian's version numbers for ubuntu..
[15:17] <mvo> juliank: thanks, I know about this, but maverick is so fresh that it does not matter that much yet
[15:18] <mvo> juliank: I did merge most (all?) in trunk/ - thanks for this
[15:18] <ogra> just make sure that never happens with initramfs-tools ... that will be painful :)
[15:25] <Keybuk> jcastro: I've proposed a BoF
[15:25] <Keybuk> jcastro: I think
[15:25] <Keybuk> jcastro: the Debconf registration/papers/etc. system makes canonicaladmin look like the work of a web genius
[15:25] <jcastro> Keybuk: ok, you might want to doublecheck, the deadline is past but they told me it's a "soft" deadline
[15:25] <jcastro> Keybuk: it's like a more obsfucated version of summit!
[15:26] <Keybuk> if I click "Own Events" on the left, I see
[15:26] <Keybuk> "Upstart in Debian", event state: undecided
[15:27] <ScottK> jcastro should volunteer to help since he's so experienced at this stuff.
[15:27] <jcastro> Keybuk: ok then you're all set
[15:27] <jcastro> (just doublechecking)
[15:39] <apw> pitti, gah didn't realise you had moved the db ... doh ... no wonder my pages are stale
[15:40] <pitti> apw: only after lucid
[15:40] <pitti> from workitems/ to workitems/lucid/
[15:40] <apw> well and * into maverick
[15:40] <pitti> right
[15:40] <apw> which they weren't last week :)
[15:40] <pitti> I announced that today
[15:40] <pitti> and they were moved today
[15:41] <apw> yeah i'd been waiting on a dx-blueprint appearing ... and it didn't :)
[16:01] <seb128> cjwatson, sorry about the changelog not being clear, you don't have the .changes available for review to get those listed easily?
[16:02] <cjwatson> seb128: yeah, I do, was just being a bit dense, sorry
[16:02] <cjwatson> it's been a while since I did serious SRU processing
[16:03] <seb128> cjwatson, np, thanks for reviewing those sru!
[16:13] <juliank> mvo: Could you merge lp:~juliank/software-center/jak
[16:13] <mvo> juliank: sure, merging now
[16:15] <jibel> asac_, hey, could you please have a look at bug 429841
[16:15] <jibel> asac_,  it was introduced with fp-installer 10.0.22.87ubuntu2 in jaunty and breaks the upgrade from lts to lts
[16:17] <mvo> juliank: hm, actually using deb822 is kind of nice as it iirc works with gzip files too. or can TagFile() do that nowdays?
[16:19] <juliank> mvo: Then ignore this revision, but add a dependency on python-debian.
[16:19] <mvo> juliank: good point! thanks
[16:20] <chrisccoulson> jibel - it should probably be me who looks at that ;)
[16:21] <asac_> chrisccoulson: take care. that package is a mess ;) ... lots of upgrade paths need to be tested before uploading something
[16:21] <asac_> also partly its owned by security team etc.
[16:21] <asac_> but yeah. you are the right person ;)
[16:21] <chrisccoulson> asac_ - ok, no worries
[16:22] <chrisccoulson> yeah, i'll take a look at that in a bit
[16:24] <jibel> chrisccoulson, as long as someone can look at it, that's perfect. Many thanks. asac_ sorry for pinging.
[16:27]  * mvo waves to jibel
[16:33]  * jibel waves back
[17:35] <didrocks> cjwatson: thanks for netbook-launcher approval
[17:41] <chrisccoulson> thanks for gnome-screensaver approval too :)
[18:40] <pwnguin> pitti: is LP #582946 a dupe?
[18:41] <seb128> pwnguin, it's not-gvfs-bug in any case
[18:41] <seb128> pwnguin, seems a file corruption, either locally or wrong download
[18:41] <rickspencer3> seb128, I
[18:41] <rickspencer3> m not sure who should own the touch one
[18:42] <seb128> rickspencer3, did you mean to reply on this channel?
[18:42] <rickspencer3> nope
[18:43] <qense> robbiew: MaverickReleaseSchedule says our current focus should be A-2, but shouldn't that be A-1, as the first alpha hasn't even been released yet?
[18:45] <pwnguin> seb128: crazy
[18:46] <pwnguin> downloading from lp via wget worked
[18:51] <robbiew> qense: no, as the iterations are more focused at how we plan the work for the release. So we scope out enough work for the 1st iteration, which ends at Alpha 2.  Then we look at what's remaining and what others may have discovered they need and scope out additional work for the next iteration, which ends on Alpha 3
[18:52] <robbiew> the iterations allow us to be a bit more agile in our approach to the release, instead of setting in stone everything we will do upfront
[18:52] <qense> robbiew: ah. The first Alpha release is just to get people to install it? ;)
[18:52] <qense> It's got a label, lets install it!
[18:52] <robbiew> and see if the damn iso builds :P
[18:52] <qense> Those things are useful to know. :)
[18:52] <robbiew> indeed
[18:52] <qense> robbiew: OK, thank you for your answer. Sounds like a good way to manage the cycle.
[18:53] <robbiew> np...and I admith it's confusing though...maybe I should just call them "Iteration 1", "Iteration 2", etc
[18:54] <qense> yeah
[18:54] <qense> Although the current names to tell when the iteration ends.
[18:56] <kees> pitti: hi! can you reset the canonical-security workitem history?  I'd like the burndown line to start with today's view of the blueprints (as of 5 minutes ago) since now we're finally done filling in the initial work items.
[18:57] <robbiew> qense: yeah, true
[19:03] <ccheney> bug 582980, I would appreciate a check to see if the patch I made looks sane for an update for Lucid SRU, its needed for OOo to save to cifs mounts properly
[19:15] <jdstrand> pitti: fyi, I've got the psql updates now. I will build in ubuntu-security-proposed as mentioned in the bug
[20:23] <psusi> anybody know of a C library that implements a binary tree data structure?  like std::map<> or something?
[20:24] <soren> psusi: glib?
[20:25] <psusi> soren: gimme something I can man? ;)
[20:25] <soren> psusi: hm?
[20:25] <Pici> grep the manpages for that? then select one?
[20:25] <psusi> the name of a function or something I can look up in man
[20:26] <psusi> for what is the question?
[20:26] <Pici> Ignore me if this doesn't make any sense.
[20:26] <juliank> psusi: http://library.gnome.org/devel/glib/unstable/glib-Balanced-Binary-Trees.html
[20:26] <soren> psusi: http://library.gnome.org/devel/glib/2.24/glib-Balanced-Binary-Trees.html [+]
[20:27] <psusi> oh, I don't want to bring in all of gtk ;)
[20:27] <soren> psusi: glib is separate.
[20:27] <psusi> hrm..
[20:28] <juliank> glib is small (smaller than libstdc++), about 800k on my system (although the package is larger as it includes gio-2.0 and others as well)
[20:33] <psusi> this might do the trick then... once I figure out how to use it... right now e2defrag keeps a forward and reverse translation map for disk blocks, but they are just arrays... so one entry per disk block... so you need 671 mb of ram just for the translation maps on a 320 gig drive... I need to rework it to use a more efficient data structure ;)
[20:45] <psusi> I'm confused... what is g_tree_insert() going to do with key and value pointers?  when it doesn't know anything about what they point to?
[20:46] <psusi> how is it going to copy a new value into the tree without at least knowing what its length is?
[20:48] <psusi> and how is g_tree_replace() any different than g_tree_insert()?  insert says it sets the value if the key already exists
[20:54] <psusi> hrm... ok, I think I see... internally the tree only stores a pair of opaque pointers, the key and the value... so I just have to allocate my own object and pass the gtree functions pointers to the key and value members of that object...
[20:55] <psusi> that's a bit disappointing... would be faster to have the object itself stored in the tree
[20:55] <SEJeff> As if richard didn't hate ubuntu enough already: https://bugzilla.gnome.org/show_bug.cgi?id=618952
[21:25] <xnox> slomo, are you building webm gstreamer packages in ppa:gstreamer-team/ppa ?
[21:26] <slomo> xnox: yes
[21:26] <xnox> slomo, awesome =) can't wait to test them ;-)
[21:26] <slomo> xnox: already there for lucid, i don't plan to upload them for karmic
[21:26] <xnox> slomo, i've upgraded to maverick already.... =) should be fine though ;-)
[21:27] <slomo> xnox: then you'll get the packages when they're in debian :P
[21:27] <xnox> well I'll just play my odds and install lucid packages without rebuilding ;-) and see what happens
[21:28] <slomo> xnox: should work just fine :) you need gst-plugins-{base,good,bad} and libvpx
[21:28] <xnox> slomo, kk
[21:29] <xnox> slomo, and I need to wait for i386 build to finish =(
[21:39] <ion> Woot, WebM sounds promising.
[21:39] <ion> Considering all the major browser vendors are behind it.
[21:40] <ion> Crap. I forgot about IE. :-P
[21:59] <xnox> ion, Microsoft issued a statement IE9 will support h.264 out of the box and WebM if plugin is installed... The better news is that loads of hardware manufacturers are behind it & Adobe as well. Apple is missing though so far?!
[21:59] <ion> http://x264dev.multimedia.cx/?p=377 though
[22:01] <xnox> ion, yeah I read that. But i don't care. Google gives the patent liceses for VP8 for free to everyone. So if anyone sues, they will have to deal with google and 30+ leading software and hardware companies.
[22:02] <xnox> ion, i don't care that x264 is better. I need to buy a patent license to use it =)
[22:02] <xnox> the question is will blueray switch to VP8 or not......
[23:10] <psusi> so I'm trying out this glib btree and it seems glib.h is trying to include glibconfig.h, which does not exist... why idea why?
[23:13]  * psusi wonders what an include file is doing in /usr/lib
[23:20] <slangasek> psusi: you need to use 'pkg-config glib-2.0 --cflags'
[23:22] <psusi> what's pkg-config?  I just had to add the -I flag, but header files should be in /usr/include not /usr/lib shouldn't they?
[23:23] <slangasek> 1) being in /usr/include doesn't guarantee you can properly include all the dependencies without having to add a bunch of things to cflags; 2) glibconfig.h is split out because it's auto-generated and machine-specific.
[23:24] <slangasek> (I don't think that's sufficient reason to put it in /usr/lib, really, but given 1) you're already doing it wrong wrt portability if you're hard-coding -I options)
[23:28] <psusi> hrm.... I'll have to look into this pkg-config then... but for now I'm fiddling with GTree to create a double binary tree indexed extent list for the block relocation map ;)
[23:29] <slangasek> pkg-config is a dependency of libglib2.0-dev.  You shouldn't have to look far.
[23:41] <temugen> k
[23:42] <temugen> hey, this isn't my vim window...