[00:08] Anyone in here have the jaunty kernel git repo? [00:10] kees: I would be happy if we could document all distinct jaunty issues though; it's one thing not to manage to have them working, but another to be aware of it ;-) [00:10] kees: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/364345 is what I filed so far [00:10] Launchpad bug 364345 in linux "Misses many important kernel configs" [Undecided,New] [00:11] Apart of the cups bug [00:11] kees: Should I file more bugs for the apparmor side of the issues? [00:11] lool: okay, cool. I just filed 364358 and flipped 364290 to "apparmor" instead of kernel. [00:12] Thanks! [00:12] lool: np, thanks for helping me debug it :) [00:12] kees: I'm happy to give you remote access to an installed system if that can help further testing if you like [00:13] lool: if possible, that'd be great. I'd like to debug why the one of the qrt tests fails too (unrelated to all this stuff). [00:13] kees: Do you have IPv6? [00:14] lool: I don't have an ipv4/v6 gateway set up, and neither of my ISPs support it :( [00:14] That's ok [00:14] jameswf: git://kernel.ubuntu.com/ubuntu/ubuntu-jaunty.git [00:21] statik: ping [00:21] statik: you going to be at the MySQL O'Reilly conf this week? [00:22] mneptok: do you mean the Oracle-SQL O'Reilly conf ? :-) [00:28] spm: i mean the MariaDB .5 Conference. 1.0 is next year. ;) [00:28] heh === ScottK2 is now known as ScottK [00:42] soren: fyi, ubuntu-server ISOs are going to be respun; linux-server needs to be moved from main to restricted now that it depends on linux-restricted-mumble-tweet [00:42] kirkland: ^^ (not sure who all is doing ISO testing this round) [01:39] hey folks, can anyone help me to report a bug in jaunty? [01:40] i don't know the cause of it, in wich application, i just know the symptom [01:41] Pretto, whats the symptom? [01:42] the problem is that all video player closes when i have "Virtual 2560 800" in Xorg's display section [01:43] sorry, in the screen section cody-somerville [01:43] think we should mess with the apt-get -moo easter egg on ubuntu? [01:44] No, it deserves not to be messed with :-) [01:45] well apt-rpm messed with it [01:49] It doesn't look like a cow :( [01:49] apt-rpm's -moo is more cow-like [01:51] and it says "Use the moo, Luke" [01:51] http://www.linuxforums.org/forum/426298-post14.html [01:52] aptitude's moo support is better :) [01:53] There's an open bug about the realism of the cow. [01:54] any help cody-somerville ? [01:54] Pretto, Filing it against the apps that crash is probably a good start [01:55] cody-somerville, so i must filing every video player? [01:55] You can't possibly have tried *every* video player, have you? [01:55] Pretto, Maybe xorg-server would be better than [01:56] cody-somerville, what about mplayer, vlc, xine, gxine, gnome-mplayer [01:56] Pretto, just file it against xserver-xorg [01:56] cody-somerville, ok [01:56] thank you [02:31] mneptok: no, unfortunately i won't be there [02:34] slangasek: have updated the release notes, in case there is a respin, please remove it [03:01] statik: grim. :( [03:58] Rafik_: net problems? [04:00] I've been wondering about that myself. Based on the timing I would guess it's because of long WHO replies or something. [04:01] slangasek: i can do some iso testing tomorrow [04:02] slangasek: mathiaz is at the mysql conference in santa clara [04:02] slangasek: i can cover for him [04:02] slangasek: are the iso's ready now? [04:02] slangasek: my cronjob should pick them up at 6am tomorrow [04:03] kirkland: I think it's been some time since he respun, so I'd guess they are. [04:03] kirkland: they're ready now, yes [04:03] serial 20090421.1 [04:03] slangasek: oh, great, then i have them now [04:03] slangasek: i've been informally testing them already, then, and haven't hit anything yet. i'll make sure to record my results tomorrow [04:04] yay [04:06] slangasek: how's the blood pressure this release? [04:10] kirkland: a little bit higher after finishing this coffee; but reasonable overall [04:11] slangasek: good to hear then, overall ;-) [04:26] Full test coverage for ubuntustudio disks, all tests have had at least one result. [06:31] LaserJock: ffmpeg is not allowed to be on any ISOs (bug #364440); we need to either cut kino out of edubuntu, or ffmpeg out of kino [06:31] slangasek: yeah, I just got the bug [06:31] ok [06:32] I'm reviewing Debian bug #457988 now, to see which approach is actually sensible [06:32] hmm, seems to be fairly core functionality for kino, so that doesn't help [06:34] we also have kino seeded on the DVD and on ubunutstudio [06:34] ubuntustudio, too [06:34] Debian bug 457988 in kino "kino: Please add Recommends: ffmpeg" [Normal,Closed] http://bugs.debian.org/457988 [06:39] slangasek: how about we drop it? [06:39] drop which? [06:39] slangasek: would that cause too much mess? [06:39] slangasek: kino [06:40] it's been part of the DVD seed in Ubuntu for quite a while, so I'm inclined to think that it would be better to drop the recommends on ffmpeg [06:40] slangasek: I mean, we can bump ffmpeg to Suggests but LP already has bugs from people requesting ffmpeg as they can't import videos without it [06:40] slangasek: ok [06:40] but leaving things as-is, and having ffmpeg just not pulled in unless you have the network available, might also be acceptable [06:41] I'm going to wait for cjwatson to wake up so I can get his opinion [06:41] LaserJock: do I have your permission to proceed with whatever edubuntu changes are appropriate, to match what we decide for the other images? [06:41] slangasek: yes [06:41] okie [06:41] slangasek: I don't have any special affinity for it. It's a bit dead and buggy. [06:42] slangasek: but if we can keep it that's good to as it is our only video editing software [06:42] ah, heh [06:42] seems to be a common problem in educational software [06:43] "dead upstream, kinda buggy, but darn it, it's all we got" [06:45] sbeattie: in the case of an edubuntu install w/o network, does the install complete, or does it bail because ffmpeg's deps can't be satisfied? [06:48] slangasek: using the add/remove installer, it asks if you'd like to attempt install anyway, even though some packages couldn't be found; if you say okay (vs cancel) it does successfully install. [06:48] (just libavcodec52 doesn't get installed) [06:52] good morning [06:55] sbeattie: hrm, it installs ffmpeg without libavcodec52? [06:56] that would be a deeper bug, the deps shouldn't allow that [06:57] slangasek: sorry, I take that back, it did not install ffmpeg [06:57] ok; that's what I expected [07:02] Good morning [07:03] asac, kees: yes, pam-foreground has gone for good for ages; our CK has a compatibility patch which creates the /var/run/console/* stamps [07:25] pitti: morning [07:49] TheMuso: can you please apply http://paste.ubuntu.com/155171/ to the ubuntustudio ship seed? [08:02] hi [08:07] asac: hey, you already had the patched libxcb for 220628 on your ppa.. mind pushing that to jaunty-proposed? [08:21] persia: can iso files be used to install from a card reader too? === mkorn is now known as thekoern_ === thekoern_ is now known as thekorn_ === thekorn_ is now known as thekorn [08:26] mdke: if you can boot from the card reader, sure? [08:28] slangasek: ok, thanks. [08:35] pitti: we don't have a DVD seed [08:36] TheMuso: well, the alt seed, which only fits on a DVD. [08:36] TheMuso: yes, it needs to be applied to the ship seed [08:37] TheMuso: right, as I said, ship seed [08:37] oops, what the others said [08:39] woops, we had that in there, but it got lost somehow. [08:39] * TheMuso sighs, a respin means more testing tomorrow. [08:41] yes, sorry [08:41] np [08:41] pushed [08:41] this should have been caught a long time ago; we apparently had this in intrepid too [08:42] As I said, I caught it for studio earlier in jaunty, and addressed it, however some fiddling with the seeds recently to put the RT kernel back in again somehow got it lost. [08:42] TheMuso: depending on what we decide (when cjwatson wakes up), we will probably need one more change to ubuntustudio seeds plus an ubuntustudio-meta upload; will you be around to help with that? [08:43] slangasek: I can be. [08:43] ok, thanks [08:44] Oh well, at least it will mean the others don't have to re-test, and testing again for me is no big deal. [09:03] how can i set the owner:group and the permissions for a directory (/var/lib/appname) when i build a debian package? [09:04] i build a deb-package for a django-web-application [09:09] is there no debian team for such things, with documentation? [09:20] c [09:23] d [09:30] evand: hi, are you there? [09:30] ogra: hm, the UNR stick still fails to be recognized on my wife's PC [09:30] (from the BIOS) [09:31] pitti, but the machine can boot from usb ? [09:31] ogra: yes [09:31] hmm [09:31] ogra: but it looks like the partition table is totally screwed [09:31] oh, it's mounted as a raw device [09:31] aww [09:31] so that isn't a normal usb HD for the bios [09:31] it should only have one vfat partition [09:31] but rather like an USB CR-ROM? [09:31] evand: it's about a bug (310804) on usb-creator. You did commited a fix, but the bug is still witht the issues and the fixes are there since 3 weeks. could you or someone commit the fix, please? [09:32] ogra: well, it doesn't have any kind of partition table _at all_ [09:32] * pitti tries with USB-CDROM or USB-FDD [09:32] thats really strange [09:32] if you plug it into your ubuntu machine it should be automounted [09:35] ogra: yes, automounting works, but as I said, raw device (/dev/sdb) [09:35] ogra: I got it to boot with "USB-CDROM" in the BIOS [09:35] TheMuso: please unseed kino from ubuntustudio; we're keeping the recommends as-is [09:35] ogra: but now I get a "boot:" prompt and nothing happens [09:35] TheMuso: which means it can't go on the ISOs [09:35] ogra: and if I press "enter", it says image "linux" not found [09:35] what does file say about the image ? [09:36] slangasek: Right, I guess I need to re-upload meta then. [09:36] oh, wait, you get a bootprompt [09:36] TheMuso: correct [09:36] ogra: yes, it's the syslinux prompt, I assume [09:37] yeah, hmm, and there shouldnt be a partition table actually [09:37] * ogra just checked his image ... [09:37] slangasek: Ok so why are we removing kino? Is it pulling in libavcodec via another package as well? [09:37] TheMuso: yes, kino Recommends: ffmpeg [09:37] and the recommends: is logically correct [09:37] slangasek: Right, thanks. [09:38] ogra: hm, still works in KVM [09:38] ogra: perhaps the booting as "USB-CDROM" broke syslinux somehow [09:38] not sure, it shouldnt write to it, should it ? [09:38] ogra: and I agree, it'd probably be more robust to have a standard partition table and a bootable partition, then the bios woudl just recognize it as a normal hd [09:39] yes [09:39] we do that for the armel image [09:39] slangasek: ok unseeded, let me rebuild meta [09:39] TheMuso: yep, thansk [09:39] it has two partitions ... but the mobile and UNR images used to work that way still [09:40] so its a bit weird that your machine doesnt boot it [09:40] TheMuso: please close bug #364440 in the changelog [09:40] Launchpad bug 364440 in ubuntustudio-meta "edubuntu-desktop installs ffmpeg which depends on libavcodec52 but is not on install media" [Undecided,New] https://launchpad.net/bugs/364440 [09:40] slangasek: gotcha [09:40] * TheMuso gives meta five to rebuild. [09:41] ogra: I played with this silly Award bios some more; USB-CDROM/USB-FDD don't work, but USB-HDD works [09:41] ogra: s/works/boots/; but with -HDD I get the broken lilo prompt [09:42] ogra: usually (with a partition) I don't select any USB-* stuff, it just appears as a normal hard disk [09:42] * pitti tries to boot in his Dell, bbl [09:42] well, USB-HDD should dtrt [09:45] slangasek: would it not make sense to drop the ffmpeg recommendation from kino? [09:46] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=457988 was why recommends were added to kino [09:46] Debian bug 457988 in kino "kino: Please add Recommends: ffmpeg" [Normal,Closed] [09:46] ah [09:46] righto [09:47] ogra: hah, it boots on my Dell [09:47] yay [09:47] but given that there seem to be BIOSes out there that cant boot unpartitioned we should consider a partitioning scheme for KK for UNR [09:47] pitti, can you file a bug for that ? [09:47] StevenK, ^^^^ [09:48] we have the code on cdimage already for the armel live image so shouldnt be to hard to re-use that [09:49] ogra: will do, when I add my testing results to the ISO tracker [09:49] slangasek: woo hold on, we actually also seed ffmpeg. I guess I'd better unseed that as well. [09:49] great [09:49] TheMuso: yes [09:50] * TheMuso restarts meta build [09:51] slangasek: we also ship ffmpeg2theora. Is that also affected? [09:51] thats in universe however [09:52] I like that UNR starter, it's really nice [09:52] although it does depend on libavcodec52 so I guess its affected. [10:00] ogra: back to my normal system; do you have a LP project for the UNR image builder? [10:01] TheMuso: sadly, yes [10:01] pitti, ubuntu-cdimage should work for a start [10:01] slangasek: np removed, and meta is rebuilding seed lists. [10:02] * TheMuso is not bothered by all this removal, but knows the studio video user community may be less than happy. :) [10:07] ogra: done, bug 364506 [10:07] Launchpad bug 364506 in ubuntu-cdimage "UNR does not boot: no partition table" [Undecided,New] https://launchpad.net/bugs/364506 [10:07] thanks [10:08] also recorded in iso tracker [10:08] pitti: is apport already disabled in jaunty?? [10:09] mnemo: yes [10:09] ah ok [10:11] tjaalton: is jaunty-proposed already open? [10:12] asac, yes [10:12] * ogra uploaded to it already [10:13] ogra: thanks. [10:13] tjaalton: so yeah. i can push that to -proposed [10:15] pitti: there seems to be something really odd going on with the at_console thing [10:15] asac: --verbose ? [10:16] Hi there. I have experienced a very bad system blocker on the latest upgrade of xorg in jaunty. The server crashes on startup and keyboard is locked. I reinstalled and it was there again. I reported basic info in bug #364488. I need to go to work now but could some of you take a look at the bug, ask more information etc? It might go in the release notes if there's no time to fix the bug right now. But it's very serious. [10:16] Launchpad bug 364488 in xorg "After some day of installation, Xorg won't start anymore (black screen and locked keyboard)" [Undecided,New] https://launchpad.net/bugs/364488 [10:16] asac: excellent, thanks [10:17] pitti: sorry. my tbird hung. couldnt find the id so quickly: bug 360818 [10:17] Launchpad bug 360818 in network-manager-vpnc "NetworkManager.vpn fails -- nm-vpn-connection.c.900: NeedSecrets " [Undecided,Confirmed] https://launchpad.net/bugs/360818 [10:18] pitti: so both parties communicating have uid=0 [10:18] but dbus doesnt like that. interestingly, adding an allow for at_console suddenly makes that work [10:18] TheMuso: how's the upload coming? [10:19] slangasek: still waiting for meta to rebuild lists. Distance to the UK and bandwidth being used by downloaders is making things a little slower than normal. [10:19] bandwidth in this case is the DC bandwidth [10:19] my connection only has the rebuilding of lists using download bandwidth so my end is clear [10:19] pitti: is there a dbus log or soemthing i can enable to get more info about which rule matches for the user (and leads to a deny)? [10:19] TheMuso: oh; if that's an issue, I may be able to do it here faster, let's see [10:21] slangasek: I'm up to armel so it should be done in short order. ports.u.c is a little quicker than archive.u.c it seems [10:21] thats how it appears anyway [10:21] slangasek: done, packaging up and uploading. [10:21] TheMuso: ah, you win :-) [10:22] slangasek: uploaded [10:23] thanks [10:23] slangasek: thank you for bringing that to my attention. [10:23] no problem :) [10:24] Ok, unless there is antyhging else pressing, I'll test tomorrow, and I'll be off and on for the rest of the evening, but mostly off I'd say. [10:25] TheMuso: ok, later :) [10:28] ogra: do you know whether the OEM Services 8.04 UNR image was partitioned? (i.e. is bug 364506 a regression from that or not?) [10:28] Launchpad bug 364506 in ubuntu-cdimage "UNR does not boot: no partition table" [Undecided,New] https://launchpad.net/bugs/364506 [10:29] cjwatson, i dont think it was but i'm not sure [10:30] can you find out? [10:30] (or somebody on your team) [10:30] calc: i had something similar once, but that was a hal-state issue/bug i couldnt reproduce after reboot. [10:31] calc: if it persists for you after reboot, it would be precious if you could see which packages got updated in the last few days for you (as it worked before) [10:32] ogra: mdz says there's no partition table on the original UNR image either, so this can wait until karmic [10:33] cjwatson, oh, indeed, its wasnt intended for jaunty at all [10:33] cjwatson, we have code that produces partitioned images already and should be easy to port over, the bug was only thought as a reminder for KK [10:33] asac: you could try to start it in the foreground with sudo DBUS_DEBUG_OUTPUT=1 dbus-daemon --nofork [10:34] asac: (warning, untested) [10:34] cjwatson, another thing, what should happen if i boot a liveimage with oem-config/enable=true ? i seem to land in a normal live session here [10:35] pitti: ok. i ask [10:36] pitti: dan said that dbus might do funny things if root user is always "at_console" ... thats not the case right? [10:36] ogra: that's correct, but the installer will run in OEM mode when you start it; the first obvious visible effect of that will be that the title bar is different, and you get a text entry box on the first page asking you for an identifier for the current batch of machines [10:37] ah, k, i somehow was expecting a fullscreen installer [10:37] asac: no, root is not in fact "at_console" unless you have an open su - session [10:38] cjwatson, right, looks different, great [10:39] the extra text makes it a wee bit to big for 600 vertical pixels in german (half of the buttons are behind the panel) [10:39] pitti: interesting. [10:40] ogra: fullscreen is for "Install Ubuntu" mode, or if you use 'ubiquity --only' [10:40] ah, right [10:41] * ogra is impressed, seems to work so far on armel === foka_ is now known as foka [10:58] anyone know if the nvidia freeze bug is going to be fixed for jaunty ? or bug 355155 [10:58] Launchpad bug 355155 in linux "Computer hard locks randomly with ubuntu jaunty" [Undecided,Incomplete] https://launchpad.net/bugs/355155 [11:04] oly: unlikely [11:04] thats a shame, especially considering how many users have nvidia cards [11:05] does that mean we should use the ones from the nvidia site [11:05] oly: how about you contact nvidia about their drivers? [11:06] i could do, but really dont know what to report, its locks completely so no logs have errors [11:06] doesn't look like many people are hitting it, fortunately [11:07] oly: how are we supposed to be able to fix it, when we don't have the source to do so? [11:07] i'm not sure if/how you can get more logs though, seeing as i doubt nvidia provide debug symbols either? [11:07] Magic [11:07] yeah, thats true but devs generally have a better idea what to look for / report [11:08] oly: install the package "ssh" and connect into your machine from another machine... then after the lockup happens, save "dmesg" and "xorg.log" and send them to NVidia [11:09] ok, i will give that a try be intresting to see if that works [11:09] oly: beware though that nvidia purposely makes the log output suck because they don't print stacktraces etc because they dont have debug symbols... [11:10] nvidia-bug-report.sh is meant for sending bugs to them [11:10] mnemo: but it's a hard lock? [11:10] tjaalton: ahh, so there is a script! [11:11] Hobbsee: yes, it's their's and comes with the package [11:11] mnemo: i thought of suggesting ssh myself, but if it's properly hardlocked, it won't respond to commands by ssh either [11:11] do you run the script when i manage to ssh in or after i have had to reboot [11:11] tjaalton: one learns something new every day... [11:11] Hobbsee: :) [11:11] well by hardlock i mean keyboard / mouse does not work cant get to a terminal with ctrl + alt + f1 or anything like that [11:11] Hobbsee: xorg often hangs so that ssh still responds... but of course if the kernel hangs there isn't much to do [11:12] oly: it could be fixed in the new driver version (180.51) [11:13] is that making it into jaunty ? [11:13] guessing not now, synaptics shows 180.44 as latest in the repo [11:14] maybe as a backport [11:14] apparently it's a "prerelease" [11:15] okay, well i may wipe the machine anyway just to rule out something left over from a previous upgrade [11:25] anyway thxs for the ideas on how to debug :) [11:34] how odd. 7.10 desktop CD doesn't boot for me in kvm [11:35] bryce: where do I find info on what you need from i965 owners? [11:36] directhex, it probably knows that it's EOLed this week :) [11:36] ogra, i wanted to check something :/ [11:37] ogra, goes nuts & eats my CPU when isolinux should be doing something like shopwing a boot menu :/ [11:37] let's try the alt cd [11:37] Riddell: https://bugs.launchpad.net/ubuntu/jaunty/+source/xserver-xorg-video-intel/+bug/359392/comments/92 [11:37] Ubuntu bug 359392 in xserver-xorg-video-intel "[i965] X freezes starting on April 3rd" [Critical,Triaged] [11:38] Riddell: https://edge.launchpad.net/~ubuntu-x-swat/+archive/x-freeze-test also has descriptions === yofel is now known as yofel_ === yofel_ is now known as yofel__ === yofel__ is now known as yofel === Adri2000_ is now known as Adri2000 [12:30] mnemo: that's true. I had that for a while under intel [12:36] TheMuso: I'm tentatively pointing to http://ubuntustudio.org/9-04_release_note for Ubuntu Studio in the release announcement; does this look correct to you, and will content be available by Thursday? [12:37] slangasek: I have nothing to do with the website I'm affraid. [12:37] TheMuso: who do I talk to for that? [12:38] slangasek: I think either _MMA_ / Cory or luisbg whenever they are around. _MMA_ is not always in here, actually he hardly ever is in here however. [12:38] They should be around in a couple of hours if memory serves. [12:38] yes, I've noticed this :/ [12:38] luisbg: ping [12:39] TheMuso: well, luisbg is UTC-2 IIRC, so I guess he might be around (or at least awake) [12:39] slangasek: yeah [12:39] slangasek: and _MMA_ is on freenode, but not in any channels atm. [12:42] superm1: is http://mythbuntu.org/9.04/release [12:42] superm1: ...where you want me pointing for release? [12:46] TheMuso: something seems awry still with ubuntustudio; libavcodec52 has been successfully blacklisted, kino and ffmpeg2theora are off, but ffmpeg and the other ffmpeg libs are still there [12:51] TheMuso: however, the seeds look correct, so I guess this is my problem, not yours [12:57] slangasek: Germinate bug? [12:57] surely not [12:57] s/bug/problem/ [12:57] that, maybe [12:57] I did actually mean problem, silly tired brain [12:59] The following packages were automatically installed and are no longer required: [12:59] amarok-common libfftw3-3 libtunepimp5 libifp4 libruby1.8 libnjb5 libofa0 [12:59] Use 'apt-get autoremove' to remove them. [12:59] hrm, thats a freshly installed system ... (ubuntu-desktop) ... [13:02] ogra@oem-desktop:~$ LANG=C dpkg -l amarok-common [13:02] No packages found matching amarok-common. [13:02] mvo: why does it want to remove nonexisting packages ? [13:04] ogra@oem-desktop:~$ LANG=C sudo apt-get autoremove [13:04] ... [13:04] 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. [13:04] * ogra_babbage is confused [13:06] ogra_babbage: oh? hm, what kind of install is this? [13:07] thats an armel desktop install [13:07] (on the babbage board as you might guess by the nick) [13:07] i ran sudo apt-get install kubuntu-desktop ... that got me the above autoremove msg [13:08] i dont get that message on any other apt action though [13:09] ogra: could you please run "sudo apt-get autoremove -o Debug::pkgAutoRemove=true" ? the output is probably long [13:09] ogra: what command did produce the autoremove message? [13:09] sudo apt-get install kubuntu-desktop [13:09] its the first three lines before it lists what it will install [13:10] ogra: interesting [13:10] funny, I suspect its a side effect of the resolver [13:11] the autoremove code checks if something is installed or marked for install [13:11] ergh, the debug settin spills a lot of output [13:11] It's a feature! [13:12] ogra: -o debug::pkgdepcache::autoinstall=true as well please [13:12] ogra: fwiw, i just tried to reproduce it in a chroot but was not able to [13:12] damned ... /me curses pidgin ... your commands are full of :P [13:12] mvo: It could be an armel thing? [13:13] it likely is [13:19] mvo, http://people.ubuntu.com/~ogra/autoremove.log [13:20] the autoinstall=true option doesnt give any extra output [13:20] I forget... Where are the release notes? [13:22] soren: ttps://wiki.ubuntu.com/JauntyJackalope/ReleaseNotes [13:24] Reading state information... Done [13:24] Garbage: amarok-common [13:24] Garbage: libfftw3-3 [13:24] Garbage: libtunepimp5 [13:24] Garbage: libifp4 [13:24] Garbage: libruby1.8 [13:24] Garbage: libnjb5 [13:24] Garbage: libofa0 [13:24] aha [13:33] slangasek: ta [13:34] soren: there's also the 'ubuntu-release-notes' project that can be used if you need to bring something to attention (throughout the release cycle) [13:35] slangasek: Clever :) [13:36] TheMuso: stopmotion *also* recommends ffmpeg [13:36] slangasek: IIRC, sbeattie requested that bug #359338 be in the release notes. looking at the bug, I don't see where that would be flagged [13:36] Launchpad bug 359338 in linux "apparmor paths are broken when using encrypted home on jaunty" [High,Confirmed] https://launchpad.net/bugs/359338 [13:37] jdstrand: the 'ubuntu release notes' task at the top [13:38] * jdstrand still doesn't see the tag, but could be blind [13:38] oh, 'task' [13:38] duh [13:39] slangasek: anyway, as more people are using encrypted home and Private, more people will be affected by this [13:40] jdstrand: you don't need to advocate for it, I was just in the process of writing the release note for that one. :) [13:41] slangasek: :) [13:41] slangasek: thanks === xq_ is now known as xq [14:16] * TheMuso gulps. [14:16] slangasek: if I unseed, could you be so kind as to rebuild meta? [14:22] TheMuso: is unseeding the right thing here? That would seem to leave only dvgrab in the video seed [14:22] slangasek: I don't know what other option we have. Removing the recommends means diverting from debian. [14:23] and likely again removes functionality. [14:23] is that worse than having an empty video seed? [14:23] s/seed/task/ [14:23] I couldn't honestly say, since I'm not a user of such apps. [14:24] * TheMuso checks its source to see whether it actually uses ffmpeg in any way [14:24] TheMuso: point being that there's only one app left in the video seed after this change [14:24] it uses ffmpeg for exports [14:24] slangasek: Yes I know. [14:24] ok [14:25] my opinion is that it makes more sense to repopulate the video task, but drop the video task from the ISO [14:26] hrm ok thats a possibility, yes. [14:26] The only problem that might cause is the video task still appearing to be selectable at install time === asac_ is now known as asac [14:27] TheMuso: it would appear for installation IFF the ubuntustudio-video package is available, which will only be the case when network repos are available [14:27] slangasek: Right. [14:29] * TheMuso attempts to get in touch with other studio devs to see what they think. If not, my gut is we go for it, and deal with the fallout from the community later. [14:31] mvo: upgrading from 8.04 to 9.04 in kubuntu leaves guidance-power-manager (called kde-guidance-powermanager in hardy) installed, but it's marked as ForcedObsoletes in DistUpgrade tool, shouldn't it be removed? [14:32] Riddell: yes it should, could you mail me the main.log file please? === JayFo is now known as JFo [14:37] slangasek: If I don't get a response from anyone before meta gets rebuilt here, we go with the dropping of the video task, but still have it available in the repo etc. [14:37] TheMuso: ok; which way are you rebuilding meta, currently? [14:37] slangasek: which way? the way one always does with regenerating seed lists for a metapackage. [14:38] TheMuso: I mean, in which direction have you modified it [14:38] slangasek: in the seed branch, reverted the video seed back to what it was before I removed kino, ffmpeg etc, but removed video from a dependency of the ship seed [14:38] ok [14:41] jdstrand, sbeattie: please review https://wiki.ubuntu.com/JauntyJackalope/ReleaseNotes#Apparmor%20profiles%20incompatible%20with%20ecryptfs [14:43] slangasek: looks good except that it is ecrypted $HOME and encrypted Private [14:43] I'll get the changes made here, and announce it to the lists. [14:43] jdstrand: I don't understand [14:43] woops wrong channel [14:44] slangasek: there are two ways to setup ecryptfs in the installer: a) all of a user's home directory (encrypted $HOME) and b) a regular $HOME directory with an encrypted $HOME/Private directory [14:45] slangasek: granted, users of 'b' will have made a concious decision to put things in ~/Private and symlink, etc to be affected by the bug [14:45] cjwatson: If I have a line in a seed STRUCTURE file like this, "ship: boot desktop audio graphics audio-plugins video d-i-requirements" and I wanted to revent the video task from being on the disks, I remove video from that line correct? [14:45] TheMuso: right, but good grief it's late to be doing that? [14:46] cjwatson: see above, and the ffmpeg mess. [14:46] slangasek: thinking about it just now, 'b' may not be worth mentioning in the release notes [14:46] cjwatson: It was slangasek's suggestion. [14:46] what a is with that udev + netlink scare? [14:46] slangasek: I'll leave that up to you [14:46] jdstrand: are you saying both are affected and this isn't clear from the note? [14:46] slangasek: yes [14:47] cjwatson: 99% of our video packages need ffmpeg. [14:47] slangasek: I might be too 'inside' though, because I recognize the difference, where and average user probably wouldn't [14:47] cjwatson: the alternative is to have a totally useless video task on the ISO, because only one of 6 apps are left in the seed after getting rid of ffmpeg :( [14:47] TheMuso: ok, then yes that's the way to remove it [14:47] slangasek: perhaps s/encrypted home directories/encrypted directories/ [14:48] cjwatson: thanks [14:48] jdstrand: hmm - I'm personally content with what's currently there, but feel free to edit if you think it needs further improvement [14:49] slangasek: ok. kirkland may have a better handle on the language, since he knows the user's better than I [14:49] kirkland: can you review/comment? ^ [14:49] slangasek: ok one more meta upload coming your way, and the work to remove video from the disk is now in the seeds. [14:50] TheMuso: right; thanks [14:50] slangasek: thanksagain for the ehads up and suggestion. [14:50] * TheMuso is glad he checked in here before heading to bed. :) [14:51] me too! [14:51] s/user's/users/ [14:56] jdstrand: howdy, what am i looking at? [14:58] slangasek: ? [14:59] kirkland: https://wiki.ubuntu.com/JauntyJackalope/ReleaseNotes#Apparmor%20profiles%20incompatible%20with%20ecryptfs [14:59] slangasek: ack, looks good to me [15:00] Riddell: thanks for the guidance-power-manager bug, could you please run "sudo apt-get remove guidance-power-manager" on the upgraded system and let me know if that will remove something else? [15:02] slangasek, yes that's the URL we'll be using [15:02] superm1: thanks [15:03] slangasek, i take it you are only planning to accept ubiquity 1.12.13 in the event a livefs reroll is needed for all disks? [15:03] superm1: correct [15:04] slangasek, ok sounds good === zmrow is now known as zach_r [15:11] mvo: only kde-guidance-powermanager === yofel_ is now known as yofel [15:13] quadrispro: you reported in the iso tracker that bug 340160 was not fixed for you [15:13] Launchpad bug 340160 in linux "RTL8187SE not recognized by jJaunty kernel on EEEpc 701SD" [Medium,Fix released] https://launchpad.net/bugs/340160 [15:14] quadrispro: that bug says that the module is not built. however, it does seem to be built: filename: /lib/modules/2.6.28-11-generic/kernel/drivers/staging/rtl8187se/rtl8187se.ko [15:14] quadrispro: so if your wireless isn't working, that's likely a different bug. could you file it separately? [15:16] kirkland: what I was asking about was perhaps also including encrypted Private [15:16] kirkland: if you think it is not worth mentionign because of everything a user has to do to hit the bug with encrypted Private, that is fine [15:17] jdstrand: ah [15:17] jdstrand: well, anyone using ecryptfs might hit it [15:17] * jdstrand nods [15:17] jdstrand: i think that was fairly clear [15:17] jdstrand: let me read it again [15:17] kirkland: it said 'encrypted home directories' [15:18] jdstrand: right, i see that now [15:18] kirkland: I of course am very aware of the difference, but an average user might not be [15:18] kirkland: I thought maybe 'encrypted directories' would cover most use cases [15:18] When using encrypted home (or private) directories together with apparmor [15:18] jdstrand: ah, yeah, that's good [15:19] jdstrand: slangasek: fixed. [15:21] ubiquity seems to show the name of every package it removes from the target system; I wonder if that's a significant drain on cpu (the text changes too fast even under kvm for me to read, so I'm not even sure it's sensible) [15:22] I doubt printing the name is much of a drain [15:22] most of the work is in apt/dpkg [15:22] label1.markup = pkgname; [15:22] I think actually it's unusual for it to be too fast to read; it isn't for me, IME [15:22] (so in my case, the feedback is useful) [15:23] cjwatson, ah, kvm may actually have an edge here, since all disk I/O is actually in RAM [15:24] TheMuso: confirmed that the seed change has ffmpeg off the ISO now; just need to wait for another publishing run to get the -meta packages up-to-date. [15:25] we had ffmpeg on iso? its in main? [15:26] it's in main but forbidden from CDs [15:26] asac, without libavcodec. somehow. [15:26] cjwatson: why is it in main if its not on CDs? [15:26] there was a TB decision on it [15:26] did we need to provide support for that? [15:26] hmm [15:27] yeah, something like that - there's a date reference to the TB meeting in question in the seeds, if you want to look it up [15:27] ok. though i dont really get whats the difference from being in main vs. universe if not on CD === Sikon is now known as LucidFox [15:28] there was a reason for it, I just forget it :) [15:28] no, but good to know. i can tell gnash folks that they can focus on ffmpeg now (they always said gstreamer is a real pain) [15:28] only if we don't want gnash on the CDs [15:29] and gnash is in mobile-mid [15:29] really. wow [15:29] i should let the gnash folks know about that i guess ;) ... might help to motivate them [15:30] someone said that ffmpeg now has a similar mechanism as gstreamer that would allow on-demand codec install [15:32] asac, i think that's most plugin developers' view - ffmpeg is easy peasy, gstreamer is not [15:32] asac, e.g. moonlight uses ffmpeg, but they'd welcome a gstreamer port from someone with a high pain threshhold [15:35] Riddell: I think I found it, its a bit too careful. if guidance-power-manager is no longer useful at all, I can force the removal earlier (via PostUpgradeRemove) [15:37] Riddell: I updated the bug [15:37] Riddell: just let me know what is most appropriate [15:40] mvo: may as well just add to PostUpgradeRemove [15:40] Riddell: thanks, doing that now [15:42] Riddell: commited [15:43] thanks mvo [15:46] mdz: the module is build correctly, but it doesn't work [15:46] i must go away now, see you soon [15:46] np === sabdfl2 is now known as sabdfl [16:01] * kwwii leaves for the airport [16:02] cjwatson: as discussed: https://bugs.edge.launchpad.net/ubuntu/+source/ubiquity/+bug/364649 [16:02] Ubuntu bug 364649 in ubiquity "Please include installation media build number in installation logs" [Wishlist,Triaged] [16:20] O_o;, I'll be damned. The Ubuntu jaunty SPARC cd works. I thought we broke those [16:22] Try harder. === zul_ is now known as zul [16:31] NCommander: d-i hasn't built on sparc for ages, so they aren't releaseable [16:33] Does anybody other than me even use Winbind? [16:33] Seems like somebody would have realized that it breaks horrible between distro upgrades. Heh. [16:34] wasabi: Bug reports to that effect would be much appreciated. [16:34] I think I have one. [16:34] Just seems a bit odd. [16:37] The basic idea is that turning off Winbind for 5 minutes is not good. [16:39] lool, ogra: oh, iop32x got posted to the tracker, but that's DOA, right? [16:39] wasabi: yes, you did file a bug about that, 6 or 12 months ago... :) [16:39] i think so [16:39] wasabi, 225874 people according to http://popcon.ubuntu.com/main/by_inst [16:40] Do they actually use it for NSS? If you turn winbind off, nss starts failing. [16:40] And the desktop goes down pretty shortly after. [16:40] heh, no idea :) [16:40] So, like, I just did the Jaunty upgrade. [16:41] It shut down winbind early in the process... and everything broke. [16:41] Nautilus died, window manager went poof. [16:41] tried to respawn a few times... [16:42] all my open terminals were useless, because they didn't know who I was. [16:42] The time that winbind is down should be kept to an absolute minimum. [16:43] slangasek: It's likely it's dead, but I'll double check on a 100 MBits link to be sure [16:43] And then at some point during the upgrade X died. That was nice too. Probably a seperate problem. [16:43] or maybe gnome-session died because of lack of NSS. [16:45] why would gnome-session need NSS while it's running? [16:45] Got me. [16:45] Any usage of getent pretty much fails. [16:46] Basically, figuring out the home directory. [16:48] there's no reason the process should need to do that more than once [16:48] Well, lots of stuff pretty much broke horribly during hte upgrade. [16:48] cjwatson, well, the CD image itself is also hosed (I think its the same bug that bit PowerPC in intrepid), so thats pointless. Its on my karmic things-to-fix list now [16:49] It wouldn't suprise me that being unable to find the home dir, or the user name, or whatever, would be a handled error condition. [16:49] unhandled. [16:50] On that vein, winbind should also cache all records for any logged in user, and never even try to reconsult the domain for it. [16:50] s/on/in/ [16:50] wasabi: You could just install nscd [16:50] Oh goodness. [16:51] winbind is supposed to play that role. [16:51] and last I checked nscd was barely functional. [16:52] Yeah, so I hear, but I wouldn't expect NSS modules to write their own caching when there's this thing that's supposed to do it already. UNIX philosophy and all that [16:52] nscd is junk. gnscd works but needs-packaging [16:52] fwiw [16:52] Winbind is already a seperate daemon. [16:52] And it already does caching. [16:52] * ebroder shrugs [16:54] winbind is free to cache if it likes, but nscd has the specific role of caching for all of nss [16:54] caches on caches. lame? maybe. required? oh yes [16:56] nscd also has the side effect of not doing the nss request as the proper user... which in the case of Winbind doesn't matter much... but like, with LDAP, it matters. [16:56] Screws up anything you're trying to use kerberos with. [16:57] Not sure how I feel about that. Requires more thoughrt. [16:57] * ogra wonders about bug 364648 ... lool it was enabled by default for me [16:57] I'm...somewhat baffled that NSS+LDAP where you need krb auth works at all. Surely root needs to look up things sometimes? But that's not actually important [16:57] Launchpad bug 364648 in ubuntu-cdimage "UNR OEM mode: Autologin should be set by default" [Undecided,New] https://launchpad.net/bugs/364648 [16:58] (i'm 100% sure) [16:58] Well, the computer should have it's own kerberos ticket. [16:58] Which is a pain in the ass to make work right. [16:58] The computer should also change it's own password. :) [16:58] winbind/samba does all this. [16:58] But only for AD. [16:59] cjwatson, on the topic of releasable, are the PowerPC ones are? (I know they've been tested, and to my knowledge they work fine, both live and alternative). [16:59] I guess I do sort of think the userlist should be maintained by the machine, and not by the user. [16:59] I realize this isn't linux specific, but I thought someone in here might happen to know, is it possible to add two 128 bit ints in a single opcode on x86? [16:59] But that's a broken scenario too though. [16:59] NCommander: in principle, if they work ... [17:00] on any x86 cpu, I mean [17:00] cjwatson, does that mean they'd be on releases.u.c, or just in the release folder on cdimage? [17:00] they aren't unreleaseable in the same way as sparc, certainly [17:00] NCommander: not on releases.ubuntu.com [17:00] that's only for the most-downloaded images [17:00] wamty, I think there are some SSE opcodes that can, but I'd have to break out the book to check. [17:00] cjwatson, I thought as much; I just dislike how unpublized ports are at times :-/. At least my goal of PowerPC on jaunty was met. [17:12] anyone here have any experience with gconf? [17:12] i am trying to remove a directory [17:13] and i get the following message: "(process:16228): GConf-WARNING **: Directory `/apps/panel' was not being monitored by GConfClient 0x18eac60" [17:14] i realize that i am skipping a crucial step to remove the '/apps/panel' directory... but i have been unable to locate any documentation as to what that step is [17:23] apologies for my question... [17:27] cjwatson, Why does livecd-rootfs override the default showformat of dpkg-query to have a space between the name and version instead of a tab? [17:38] hi all, i noticed an update to evolution-mapi recently so i tried to test tit again... still fails, so i did a detailed apport bug 364708 [17:38] Bug 364708 on http://launchpad.net/bugs/364708 is private [17:38] s/tit/it [17:40] hi - it would be very good if bug 347361 could be fixed in Hardy, the corresponding debian package should have a patch [17:40] Launchpad bug 347361 in net-snmp "libsnmp-python segmentation fault" [Undecided,New] https://launchpad.net/bugs/347361 [17:40] I can confirm that the bug is real [17:44] cody-somerville: does it matter? [17:45] cody-somerville: I think it could use the default format, but it hardly seems important [17:45] cjwatson, You tell me if its important :P [17:45] it is not important [17:45] Okay, thanks. [17:45] regarding the bug above, it is reported against intrepid, must another bug be created for hardy? [17:46] cody-somerville: if you want it changed, though, feel free to file a low-priority bug [17:46] it would only be for code cleanup purposes [17:47] cjwatson, Okay === evand1 is now known as evand [18:03] how do i go about getting my password removed from an apport bug report stacktrace [18:06] jcole: you can try asking on #launchpad, they can edit comments [18:06] jcole: retracing are private bugs by default if that makes any difference, ie only selected people can access to those [18:07] thanks seb128 [18:07] you're welcome [18:08] seb128: im thinking i may need to sudo sed -i /var/crash/* from now on before submitting to apport [18:08] jcole: will that works in crashdumps too? [18:09] jcole: you can still retrace locally and clean the stacktrace if you don't want those infos to be in private bugs [18:09] jcole: is the password in an attachment? [18:09] james_w: yes, stacktrace [18:10] jcole: you can just delete the attachment then [18:10] on the right hand side is an "attachments" box [18:11] click the "edit" link next to it, and then delete [18:11] you can do that for each that contains it [18:11] james_w: oh ok, maybe i can edit then reupload [18:11] yeah, that would be useful if you still want this report worked on [18:11] is this bug still private? it will say in the top right [18:15] james_w: yes it is... but, i will make it public after i clean up the trace :) [18:16] just wanted to make sure you were vaguely protected [18:28] thanks guys, ive sanitized the attachments in bug 364708 and made it public... now ill just sit back and wait :) [18:28] Launchpad bug 364708 in evolution-mapi "4/21/2009 evolution-mapi still crashes after latest updates, calendar still does not work" [Medium,Confirmed] https://launchpad.net/bugs/364708 [18:31] jcole, does it work with SSL-demanding servers yet? [18:33] directhex: if i dont use jaunty's version and compile openchange/evolution-mapi/samba4 manually, everything works, even tls/ssl [18:34] jcole, hm. lame. [18:36] directhex: if i manually "fudge" evolution mapi configs into ~/.evolution/ and ~/.gconf.gconf/apps/evolution/ i can get mail to *partially* work with jauntys versions, but calendar does not work at all [18:37] s/.gconf.gconf/.gconf [18:44] jcole: can you figure if rebuilding only one of openchange, evolution-mapi, samba4 and see which one fix the issue? [18:51] seb128: well, ive tried that to a certain extent... the latest openchange does not like the samba4 dev in jaunty so i had to pull the latest samba4 dev from cvs for it to compile... then, the latest compiled openchange would not "cooperate" with jaunty samba4... i had to make install both openchange and samba4 [18:52] jcole: could you try if building openchange 0.8.2 is enough to fix the issue? [18:52] seb128: after that fiasco, i tried evo mapi in jaunty which still crashed fro me... so again, i compiled evo mapi from source which worked [18:53] seb128: i compiled openchange (using the latest samba4 dev) and did a make install... but jauntys samba4 version wont work with it [18:53] jcole: could you try just updating openchange to 0.8.2? [18:59] seb128: jaunty already has 0.8.2 and but svn has a later version... i will try to build again according to http://wiki.openchange.org/index.php/Debian [18:59] jcole: no, it has 0.8 [19:00] $ apt-cache show libmapi0 | grep Version [19:00] Version: 1:0.8-2ubuntu1 [19:00] ohh [19:01] 0.8-2 != 0.8.2 === dpm_ is now known as dpm [21:50] hi seb128, jcole [21:52] hey jelmer [21:52] jelmer: thanks for not updating openchange before jaunty [21:53] seb128: sorry, I just haven't had the time [21:53] upstream has nontrivially changed for 0.8.2 [21:53] that's ok, would be nice to let us know next time though [21:53] evolution-mapi doesn't work at all in jaunty [22:02] seb128, so perhaps considering that is it worthwhile to look into an SRU for 0.8.2? [22:02] superm1: if that's an update with non trivial changes that's not going to be easy to get sru-ed [22:02] superm1: we will try to backport fixes though [22:03] right, well maybe as soon as karmic opens; a backport at least yeah [22:03] superm1: but testing exchange fixes without an exchange server is not trivial so we rely on user comments [22:04] seb128, understood. i personally kept getting excited every time at the prospect that evolution-mapi might finally work whenever i saw the uploads for related packages coming through [22:05] I've been tried to get working for jaunty, but jelmer said he would update to 0.8.2 in debian which he didn't [22:05] then I backported some svn changes which were fixed a crasher issue according to upstream but there is still some bugs apparently [22:08] ah i see === JayFo is now known as JFo [22:54] Hey, I'm staring at a bug in notify-osd, where sometimes bubbles don't fade in properly, instead starting at 0 opacity until I mouse over them. (Especially when I am not moving the mouse, and ESPECIALLY volume confirmation ones although it could be because those are the ones I most expect to fade in properly). Anyone know about this one? [22:59] Picklesworth: https://launchpad.net/ubuntu/+source/packagenameyouareinterestedabout/+bugs [23:02] am i the only one confused by the new volume control behaving in reverse compared to the old one if you use the mouse wheel in it? [23:02] directhex: ditto [23:03] i have volume buttons on the keyboard, but habits die hard [23:03] directhex: to make it more confusing, open up the volume control and its reversed [23:05] wtf with hyperair [23:05] jcole, lame. file a bug! [23:06] there is a bug about the volume thing [23:06] that's okay then [23:06] bratsche looked a bit to it but said that's non trivial to change [23:06] I usually use the mouse scrolling on the icon without opening the popup dialog which works correctly ;-) [23:06] let me guess. some stupid thing to do with the gtk widget itself? [23:07] jcole: very sorry. i had a screen + irssi malfunction. [23:07] probably workaroundable in some hideous manner, but easier to ignore for now [23:07] right [23:08] or rather there is other higher priority bugs open [23:08] but patches are welcome as usual ;-) [23:08] gnome volume control also used to be "linked" to the volume control on the panel... now when you control the gnome volume control, panel volume doesnt update... [23:08] i said "for now"! [23:08] jcole: is the panel mixer displayed the same channel than the one you control? [23:08] seb128, i'm not accepting responsibility for patching gtk. too high profile. [23:08] displaying [23:09] jcole: if you set them both to modify the same stream then it will be linked. otherwise it won't [23:09] right click, preferences, see what is selected there [23:09] plus, C. yuck. [23:10] ie: open gnome volume control, then click the panel to show the volume control, move back over to the gnome volume control master and scroll on it [23:10] no problem, I am going to start leaning on luisbg_ more anyway./c [23:10] woops [23:10] the reverse works just fine [23:11] the popup is hidden when switching focus [23:12] seb128: just use the scroll [23:12] seb128: dont focus it :) [23:12] its no biggie, just something i noticed [23:13] right, the icon is not updated either when doing that [23:17] im out early... [23:25] Jaunty does have the fancy new gnome-volume-control-pulse in the repositories, now, which is vastly cooler if you use PulseAudio :) [23:25] (If you want a workaround) [23:33] seb128, directhex: Yeah, since the volume control went horizontal the mousewheel does the opposite effect. Because mouse-up translates into left in the range widget, which the applet interprets as "volume down". [23:33] funny that Ubuntu kept the horizontal thing but dropped the PulseAudio part. Not even Fedora is using the horizontal applet [23:34] unless my memory is foggy :P [23:41] the horizontal slider is the GNOME official 2.26 applet [23:41] the new one is a notification area icon and use pulseaudio only [23:41] we just build the optional upstream applet and use that without changing it [23:42] seb128: okay, thanks for clearing it up. Wasn't sure how official it was :) [23:51] seb128: notification area icon? how do i enable that? [23:51] i still have volume applet [23:52] alex-weej: install gnome-volume-control-pulse that's the GNOME 2.26 pulse only thing [23:52] is this in the default install? [23:52] alex-weej: it's only an icon in the notification area and a slider and can do pulse only [23:52] alex-weej: no, it does pulse only and we have the applet which does alsa, gstreamer, etc which quite some people rely on [23:53] ok [23:53] * alex-weej doesn't even have pulse installed, too much trouble for me right now :( [23:54] WTF can't I modify fields in read-only spreadsheet? I mean, it's not like I want to actually save them.... [23:57] lamont: because open office. [23:59] heh