=== WelshDragon is now known as YDdraigGoch [01:15] hi. i reported a DoS security bug against ubuntu's packaging of apache. there's been no apparent activity, a week later. is this typical? [01:15] neilv: have you subscribed the security-team? [01:16] they are subscribed. i flagged it as a security bug from the start [01:16] https://bugs.launchpad.net/ubuntu/+source/apache2/+bug/394350 [01:16] Ubuntu bug 394350 in apache2 "RLimitCPU has no effect in Apache" [Undecided,Incomplete] [01:18] hmm, you seem to have done everything correctly. try #ubuntu-hardened or #ubuntu-server and see if a security-team member is around. [01:18] thanks [06:40] hello!!! [06:40] i know that this is not a support channel but do you think someone could help me figure out a problem with the repositories?? [07:09] Good morning [10:03] mvo: did you change apt to hold back stuff that breaks something rather than removing? [10:03] i rmember that seb wanted you to do that last sprint [10:03] asac I have not changed that code, its a fiddling beast. if you have a good example situation I can look at it again === asac_ is now known as asac [10:03] now would be a good time I guess [10:03] mvo: please look at NM ppa [10:04] mvo: i have added a breaks libmbca0 to mobile-broadband-provider-info [10:04] but it seems taht apt is holding the upgrade back now [10:04] instead of removing [10:04] mvo: should i have used a conflicts rather? [10:05] https://edge.launchpad.net/~network-manager/+archive/ppa [10:05] 11:00 [freenode] [torgny_j(n=quassel@94.191.168.84.bredband.tre.se)] The following packages have been kept back: [10:05] 11:00 [freenode] [torgny_j(n=quassel@94.191.168.84.bredband.tre.se)] mobile-broadband-provider-info network-manager-gnome udev-extras [10:05] i forgot to remove libmbca0 from the Recommends of network-manager-gnome [10:05] i how thats not the reason ;) [10:07] asac: what is the output of -o Debug::pkgProblemResolver=true ? [10:10] mvo: i asked him [10:10] what does "Enhances: " mean? [10:10] asac: not a lot currently, its something like a reverse suggests [10:10] asac: but its not used by apt [10:10] (or any other tool) [10:10] good. then why are folks using it ;) [10:11] just to confuse me ;) [10:23] probably [11:15] pitti, hey [11:15] pitti, can you build ppa builds priorities too? [11:15] bump priority rather [11:17] seb128: hey [11:17] seb128: yes, I can [11:17] pitti, https://edge.launchpad.net/~rodrigo-moya/+archive/ppa [11:18] pitti, can you bump the json-glib build there? [11:18] 11:27 < asac> ppa weather: http://identi.ca/notice/6188970 [11:18] ;) [11:18] pitti, some people need that to do a demo at GUADEC after lunch [11:18] seb128: done [11:18] pitti, thanks [11:19] wow, the PPAs have a huge backlog [11:19] right [11:19] http://www.sofaraway.org/ubuntu/tmp/ppa-builders.png [11:19] http://www.sofaraway.org/ubuntu/tmp/ppa-builders2.png [11:19] now we have 6 builders again at least on i386 [11:23] pitti, can you do the same for the karmic version there? [11:24] seb128: bumped [11:24] pitti, danke [11:24] seb128: jaunty all built [11:24] excellent [11:25] pitti, btw gnome bug #586708 might be the bump you were looking for about gvfs [11:25] Gnome bug 586708 in general "User Choice of VFAT Mount Option" [Enhancement,Unconfirmed] http://bugzilla.gnome.org/show_bug.cgi?id=586708 [11:26] mvo: http://paste.ubuntu.com/212636/ [11:27] mvo: so situation is: old mobile-broadband-provider-info depends on libmbca0 ... new one breaks libmbca0; new udev-extras seems to conflict with hotkey-setup [11:30] mvo: what does the number mean after the package name? === Ari-x is now known as Ari|pause === tkamppeter_ is now known as tkamppeter [12:07] pitti - just looking at the GDM source - do you agree that the only reason GDM needs to depend on gnome-session right now is because it needs to tell the session manager what autostart directory to use, and other session managers don't support this? [12:08] chrisccoulson: well, gdm itself is half of a GNOME session now [12:08] so it certainly needs gnome-session-bin [12:08] but we don't need gnome-session's dependencies and shipped .desktop files [12:08] (which is what they split out in Debian, AFAIK) [12:09] i'm just wondering if it's possible to use other session managers to load the GDM session, but they would need to support the "--autostart" option to specify what autostart directory to use [12:10] the only reason i mention it is because it looks like xubuntu will have to have both gnome-session and xfce4-session by default now, unless i'm misunderstanding something [12:11] if that's the case, perhaps the right thing to do is to adapt xfce4-session so that it can load the GDM session [12:11] ah, perhaps yes; I haven't looked into it that deeply [12:11] chrisccoulson: but as a first mitigation, I think merging with Debian and just depending on -bin should work [12:11] then this at least stops pulling in gnome-panel, nautilus, compiz, etc. [12:12] gdm is the GNOME display manager, I don't see the issue with depending on gnome-session [12:12] i had a look at the xfce4-source, and it seems possible to do that - as long as that is the only reason GDM needs gnome-session [12:12] xfce can use xdm they want or something else [12:13] hi seb128 [12:13] hey chrisccoulson [12:13] having a good week? [12:14] yes! [12:14] yes! [12:14] out of the internet connection there [12:15] GUADEC core is over now though [12:19] hey o/ [12:19] same internet connection at RMLL :( [12:23] hello didrocks [12:23] didrocks, how is that going? [12:27] seb128 - were you involved with any tracker discussions at GUADEC? ;) [12:28] not really, they have a hacking session this afternoon but I'm flying back today [12:28] anything you wanted to know about it? [12:28] I don't think there is much new out of what has been on planet recently [12:28] it was just more for general interest really [12:33] ok === Ari|pause is now known as Ari| [12:38] udev (142-2) breaks dmsetup (<= 2:1.02.27-4ubuntu5) [12:38] pitti: do you in which way it breaks it? [12:39] * asac tries to get udev >= 142 to jaunty [12:40] I just re-enabled desktop effects (had been unstable earlier in karmic), and now gnome-session is continuously respawning metacity (which fails because compiz is running) [12:42] mdz, right several people have been running into this issue [12:44] seb128, is bug 389686 the best one for this issue? [12:44] Launchpad bug 389686 in metacity "compiz --replace fails to kill metacity, resulting in cpu overload" [Undecided,New] https://launchpad.net/bugs/389686 [12:45] mdz, I've not looked to bugs too much this week but I think you can use this one, the other comments I did read are confusing "gconf is using too much cpu" [12:45] asac: the number is the score that apt assigns to the package (how important it thinks it is) [12:45] yeah [12:46] so seems udev-extras isnt important enough to make hotkey-setup go away [12:46] same for libmbca0 breaks [12:46] mvo: so conflicting is better than breaks maybe? [12:47] seb128, ok, I'm marking that one triaged and moving it to gnome-session (since that seems a more likely culprit than metacity) [12:47] mdz, thanks [12:48] chrisccoulson, do you think you could have a look to this one? [12:49] yeah, i can take a look at that. although, gnome-session hasn't really changed yet in karmic [12:49] but, i suppose that doesn't mean it isn't a gnome-session bug [12:49] right [12:54] asac: what do you want to archieve that libca0 gets removed from the users system? the effect should be the same, in both cases the score is calculated the same way [12:55] mvo: well. new mobile-broadband-provider-info would break libmbca0 parser [12:55] so it doesnt work again [12:55] lunch time bbl [12:55] also the new nm-gnome applet doesnt need libmbca0 anymore [12:55] so it can go away [12:56] mvo: we could just keep libmbca0 installed ... that wouldnt hurt, because noone is using it. but it wouldnt be just to not conflict/breaks on it i think [12:56] asac: well, if there is a new version that supports the new format the lib should be updated. [12:56] asac: it will be picked up as a automatic deepndency if nothing is using it anymore [12:56] mvo: there is no version [12:56] asac: so that should be fine, at most the next release upgrade will rmeove it automatically [12:57] its supposed to die imo [12:57] so only solution is to make libmbca0 an empty package? [12:58] mdz - i can't test this metacity respawing issue until i get home, but, jut out of interest - what is the output of "gconftool-2 -g /desktop/gnome/session/required_components/windowmanager" just after you've enabled desktop effects? [12:58] mvo: and let it slowly die by dist upgrader? [12:58] does it still say metacity? [12:58] well, or making sure that its not used by anything anymore and using the break or conflict [12:58] but the score of 1 indicates that something is still depending on it [12:59] chrisccoulson, it says compiz right now (after letting it spin for a long time and then running killall metacity) [13:01] mdz - thanks. i'll take more of an in depth look at it later, but the fact that it stops if you run "killall metacity" makes it look like it's actually a metacity bug - it's probably asking to be restarted [13:02] chrisccoulson, ok, feel free to move it back, I was guessing [13:02] mvo: are you talking about libmbca0? nothing uses it afaik [13:02] mvo: at least after the upgrade [13:02] no problem. i'll leave it as it is for now [13:02] mvo: only thing i forgot is a recommends in the -gnome package [13:02] thanks [13:02] mvo: is that a problem? [13:02] e.g. do recommends bring a score of 1? [13:04] asac: in the past it did not, but I fixed that recently, not sure if you version is affected or not. I think I need to re-create the situation in a chroot to check - do you get this on your machine or is it from a user bugreport? [13:05] mvo: so you say recommends now bring in a score? that would explain it [13:05] no need to reproduce if it doesnt go away after removing recommends ill let you know [13:26] asac: breaks> no idea, I'm afraid; Keybuk? [13:27] hmm? [13:28] asac| udev (142-2) breaks dmsetup (<= 2:1.02.27-4ubuntu5) [13:28] asac| pitti: do you in which way it breaks it? [13:28] -*- asac tries to get udev >= 142 to jaunty [13:41] Keybuk: do i need to backport devmapper 2:1.02.27-4ubuntu6 from karmic or was the break introduced because it needed a rebuild? [13:57] asac: the udev rules format changed [13:57] and we moved from using vol_id to blkid [14:18] * hyperair grumbles about gdm's user switcher being incredibly bulky [14:20] at least fusa was shrinkable in its preferences [14:50] Would it make sense for a developer to add an Alpha 5 milestone to this bug, Update manager stopped showing the number of updates available https://bugs.launchpad.net/bugs/330439 ? Avoid hitting the KK UI Freeze like it did with JJ... [14:50] Ubuntu bug 330439 in update-manager "[jaunty] update manager stopped showing the number of updates available" [Medium,Triaged] [14:53] nerd_bloke: I have a look [14:54] mdz - i'm failry certain that the metacity respawning issue looks like a metacity bug rather than a gnome-session one, so I've re-assigned it back now. i can't test it until i get home from work, but i've had a look at the code and i've got a rough idea of what is happening now [14:55] chrisccoulson, thanks for looking at it [14:56] you're welcome [14:57] chrisccoulson, you are doing some good work as usual ;-) being accepted as motu yet? ;-) [14:57] been [15:04] seb128 - not yet. i've put myself on the motu-council meeting agenda for 23/7 [15:04] mvo: thanks [15:04] the next meeting is actually friday, but i couldn't make that one else i'd be late for work;) [15:06] chrisccoulson, ok [15:28] Does anyone know if we have (even officious) stats about the adoption rate of security updates over time ? [15:37] http://www.techzoom.net/publications/firefox-update-dynamics/ === Richie is now known as YDdraigGoch === crevette__ is now known as crevette [16:16] * hyperair fumes [16:17] who was the brilliant one who decided that it would be a good idea to stop gdm while upgrading?! [16:17] * hyperair (im)patiently waits for apt to finish its stuff before gdm can come back on [16:20] hyperair: see #ubuntu-devel /topic; it was a bug [16:20] and u-d-a@ [16:20] hmm [16:20] is there a bug about mass storage devices no longer automounting? [16:20] i see [16:20] * Laney assumes yes [16:21] mass storage devices not automounting? [16:21] it does for me now =\ [16:21] WFM [16:21] it doesn't auto-popup nautilus, though [16:21] that's a bug [16:21] nope, didn't mount at all [16:21] i disabled that anyway, so i didn't notice anything [16:22] just tried with my phone though, I'll check with a usb key later [16:29] * hyperair curses [16:29] i can't even launch a browser now because it won't give me anything but a bloody black screen [16:34] have to run out early today, cu tomorrow === cking_ is now known as cking [18:22] pitti: there? so NM maintainer aims for Fedora 12 (end of october/begin of nov) for 0.8 release. so if we want to do that, we need to accept that we will get a final version SRU post-release [18:24] if thats ok, i will ensure that he doesnt break abi after our final release cut (like he did for intrepid) ... and then i would now start preparing the 0.8 bits === dpm__ is now known as dpm === WelshDragon is now known as YDdraigGoch === WelshDragon is now known as YDdraigGoch