[00:55] <dashua> nicee
[02:52] <verb3k> In checkinstall, what does the --backup option do exactly?
[05:56] <psusi> if there is a bug in the toolchain, what should it be filed against?  gcc?
[05:59] <persia> psusi, Depends which part of the toolchain :)
[06:00] <psusi> I *THINK* it's ld
[06:00] <psusi> I filed this bug all the way back in 2005 and it has not gotten any attention yet the problem has actually grown worse... I think it's because I never targeted it at a specific package.  bug #24692
[06:01] <psusi> in maverick it seems we now have a default alighment of 2 mb, not just 1
[06:03] <persia> psusi, Might track down whoever uploaded the alignment change.
[06:03] <psusi> I don't even know where the setting is ;)
[06:03] <persia> Might be nobody else does, hence the uncommented bug report :)
[06:04] <psusi> hehe... maybe I check the logs and see who usually uploads gcc... someone has to be an expert on it ;)
[06:05] <ebroder> Ugh. We need a duplicate bug finder for dkms failures
[06:06] <ebroder> I'm up to...37 dupes of this fglrx bug?
[06:13]  * psusi is glad he doesn't use fglrx anymore.... hooray open source drivers!
[06:14] <persia> ebroder, I think the bugsquad has something called "recipes" that helps merge mass-duplicates.  Ask someone in #ubuntu-bugs.
[06:14] <persia> I strongly believe it's harmful to claim bugs are duplicate without deep information.
[06:15] <ebroder> Deep information? What do you mean? I'm reading the description (making sure they're installing fglrx, whether they know it or not) and checking the build log to make sure it's the same build error
[06:16] <micahg> ebroder: the Ubuntu X team likes a separate bug for each piece of hardware
[06:16] <persia> It's the making sure it's precisely the same error because one understands how the software works that is the deep information to which I refer.
[06:16] <persia> micahg, Only where it's hardware-specific (and it very often is)
[06:16] <micahg> persia: right and unknown = don't dupe
[06:16] <ebroder> persia: I'm fairly confident on this one - the build-logs look *identical* with the exception of timestamps and l10n
[06:17] <ebroder> And it's a build error - how can it be hardware specific?
[06:17] <ebroder> (the bug I'm using as a parent for everything is bug #642518)
[06:17] <persia> I don't mean to criticise your desire to duplicate them: I'm uninformed, and presume you to be informed.
[06:17] <micahg> ebroder: same with me, just providing info here
[06:18] <persia> I mean to explain why I think it's a good idea for the reporters to have filed the duplicates in the first place, and point you at a resource so you can clean up more easily.
[06:18] <ebroder> persia: I understand; sorry for getting defensive. I did probably jump in a bit aggressively, but this bit me indirectly, and seemed like something I could take care of
[06:18] <persia> So, please, mark them as duplicates.  You may find "bug recipe"s useful in doing so.
[06:18] <ebroder> I think I've got them all by now, but I'll keep that in mind next time :)
[06:18] <persia> No worries.  Just wanted to make sure we understood each other :)
[07:19] <ebroder> Ok, this sucks. The security update made compat_alloc_user_space GPL-only, when it wasn't before
[13:13] <mdke> grateful if a release team member could have a look at / approve ubuntu-docs 10.10.3 when convenient
[14:17] <lucidfox> Okay, bug #615300 just feels like one slap in the face too many
[14:17]  * lucidfox considers filing "Evolution should *not* have a default signature enabled"
[16:23] <ScottK> ebroder: Sounds like a bug that should be reported and brought to the attention of the security team.
[16:49] <kees> ebroder: are you on maverick or an earlier release?
[16:58] <ebroder> kees: Lucid. The issue is that the new compat_alloc_user_space is EXPORT_SYMBOL_GPL, while it used to be an inlined function declared in a header
[16:59] <ebroder> And that's breaking fglrx - bug #642518
[16:59] <ebroder> I have a fix, although I don't really like it, and I want someone more familiar with fglrx than me to pass judgement before I start pushing for sponsorship
[17:01] <kees> ebroder: okay, thanks. this is a bug in the kernel update and we'l get it fixed asap; sorry for the glitch!
[17:02] <ebroder> kees: Hmm, really? Are you guys allowed to decide that the symbol should be exported to non-GPL modules? I thought there was legal mumbo-jumbo surrounding all that
[17:03] <kees> ebroder: it's a backported fix; it just wasn't backported correctly. it'll still be a problem for maverick.
[17:05] <ebroder> Ok, great. fglrx will still need a patch, because it's looking for the function in the wrong header file, but that's easier to deal with
[17:18] <ebroder> kees: Do you want me to file a bug about this?
[18:29] <kees> ebroder: I think the existing uninstallability bug is enough, but thanks!
[19:49] <lucidfox> So
[19:49] <lucidfox> is bug #615300 going to be set to something other than "Fix Released" and reopened for discussion?
[19:52] <fagan> lucidfox: I presume there will be more talk tomorrow about it when the canonical crowd are around
[20:01] <vish> lucidfox: hmm , did anyone test the change?
[20:01] <vish> i need to use evolution in Maverick
[20:05] <lifeless> wheee
[20:05] <lifeless> thats a fun bug :P
[20:08] <fagan> lifeless: ill check it
[20:10] <fagan> its not showing up on the bottom of my messages
[20:11] <fagan> I sent a few messages today but its not showing up
[20:11] <vish> fagan: might it just be a preference ?
[20:11] <lifeless> AIUI its in your signatures lists you can choose from
[20:11] <fagan> maybe
[20:11] <lifeless> so you can turn it on
[20:11] <lifeless> and for new accounts it may default on
[20:11] <vish> something that the user can change if they want
[20:11] <fagan> yeah its for new accounts
[20:12] <vish> it seems to be another huge fuss over nothing.. or just  a preference but need to actually check it in action though ;)
[20:12] <lifeless> it is minor; I don't really see the point though.
[20:12] <lucidfox> Making this default for new account would be a huge mistake.
[20:12] <lucidfox> No it's not minor
[20:12] <fagan> Well id like the wording to be a lot better
[20:12] <lucidfox> it's a recurring pattern
[20:12] <fagan> thats all
[20:12] <lifeless> lucidfox: snarfing passwords and mailing them to me would be major
[20:12] <vish> lucidfox: is it not possible to change it?
[20:12] <fagan> I dont mind it in there I just dont like the wods
[20:12] <fagan> words
[20:13] <vish>  to me it just sounds like using @"gmail"
[20:13] <lucidfox> It's possible. I still don't like this being by default. And I don't like Canonical forcing a closed-doors decision on the users yet *again*.
[20:13] <vish> but i wouldnt use it though .. i'd just change it to what i want to use..
[20:14] <fagan> well lucidfox they are a private company they dont really have to run most things past the community
[20:14] <fagan> its hard to dictate to someone thats giving you stuff for free
[20:15] <lucidfox> If they care about their reputation, yes they need to cooperate with the community at large
[20:15] <vish> lucidfox: so how to communicate?
[20:16] <vish> lucidfox: the bug was a public bug..
[20:17] <lucidfox> It wasn't. It went from filing to implementation in the span of... less than a day, I think. And by the form it was posted, it seems it was already decided behind closed doors by then.
[20:17] <vish> lucidfox: why do people always assume the worst? :)
[20:18] <lucidfox> I'm just judging by the history of Canonical's prior initiatives
[20:19] <lucidfox> Rick Spencer filed the bug, then Sebastien Bacher almost immediately assigned it to the beta milestone. Then a package was uploaded implementing it, with no comments in between.
[20:38] <fagan> lucidfox: well it might have been talked about on IRC beforehand
[20:42] <avi__> Hey guys, so I've written this cool app in PyGTK and Glade. How can I package this up and distribute it?
[20:42] <luca> hi everyone
[20:43] <fagan> avi__: you should ask that on #ubuntu-app-devel
[20:43] <avi__> awesome, will do. thanks!
[20:44] <luca> I would need some help, but none is available in @ubuntu+1 currently .. maverick, kernel 2.6.32-23 (latest) refuses to install
[20:45] <luca> currently I am without a workable kernel installed. Anyone with the same problem?
[20:46] <micahg> luca: maverick should have a 2.6.35 kernel unless this is special hardware
[20:46] <luca> sorry
[20:46] <luca> 2.6.35
[20:47] <luca> misspelled the kernel name
[20:47] <luca> there seems to be some problem in the transition between 2.6.35-22 to 2.6.35-23
[20:47] <luca> wondering if someone was having the same problem, maybe a fix.
[20:48] <mdke> lucidfox: I agree with you, thanks for raising it
[20:48] <ebroder> luca: Are you using proprietary drivers?
[20:48] <ebroder> There's a known bug compiling fglrx against the most recent Lucid and Maverick kernels
[20:49] <luca> broadcom for wireless, but it is not installed right now
[20:49] <ebroder> Ok, never mind, then
[20:49] <luca> (I know about the bug, I have been without fgrlx for some days now)
[20:58] <avi__> _is
[21:21] <kklimonda> lucidit would make more sense to bring this subject to u-d-d instead. Most developers don't ead forums nor use IRC at sunday.
[21:21] <kklimonda> meh
[21:21] <kklimonda> great, she's gone...
[21:22] <kklimonda> anyone else cares about it enough to raise it on ML? ;)
[22:34] <YokoZar> did ulimit -n change in maverick? It seems to be way too low 1024, iirc we had it much higher in previous releases
[22:47] <rishi> I have a stripped down Ubuntu 10.04 which I am using as an appliance for collecting behavioral data for research. Part of it involves recording audio/video from an USB camera. The recording is initiated from an initscript. Now I find that if I am not running a desktop, ALSA finds plughw:1,0 to be busy. The open system call reports EBUSY. It looks like something changed in 10.04's boot sequence which is causing this.
[22:48] <rishi> I ran a quick test on Ubuntu 9.10 and Fedora 13 and they seemed to be fine.
[22:48] <rishi> Can anyone suggest what might be going on?
[22:49] <rishi> I have run lsof, etc. on almost everything imaginable and there is no other process using the device.
[22:49] <rishi> Now the funny part is that if I try to record manually, ie. log in and then use a shell, then everything is fine.
[23:57] <bdrung_> doko: no time and no clue for fixing bug #643107.
[23:58] <doko> bdrung_: modularizing the ant script might help