[12:40] <mjg59> BenC: The Broadcom stuff seems really quite solid now
[03:11] <cjb> BenC: Do you maintain binary compatibility with upstream, or could I try to persuade someone to carry the fs events connector/netlink patch?
[03:11] <mjg59> cjb: Userspace or kernelspace?
[03:11] <mjg59> (Compatibility)
[03:13] <cjb> Kernelspace; new netlink entries are analogous to new syscalls.
[03:13] <mjg59> Then there's no real effort
[03:13] <mjg59> So you just need to talk Ben into believing it's useful :)
[03:15] <cjb> Cool.  I wonder what kind of convincing one makes about whole fs event notifications being useful.  Seems like a you-get-it-or-you-don't thing.
[03:16] <cjb> rlove doesn't like the patch; he dislikes that you need to have a daemon up to catch the events, so he wants an event log to go into ext3/jbd instead.
[03:25] <infinity> I'm with rml, but only because I'm a frothing fanboy.
[03:26] <infinity> (And I hate having 372 helper daemons running just so my kernel can talk to itself)
[03:30] <cjb> infinity: Stephen/tytso are never going to accept such a patch to ext3, though.  And I don't think the jbd gives early enough context for it to stay in there without hitting ext3.
[03:31] <cjb> I'm a frothing fanboy too, and did take the time to write up rml's thoughts:  http://blog.printf.net/articles/2006/03/31/filesystem-notifications-revisited
[03:33] <cjb> But that doesn't mean I want to give up on the feature because neither set of maintainers think it's quite right for their subsystem.  The netlink version requires you to keep the daemon running in order not to miss events, but doesn't break anything?  I'm for that, then.
[08:46] <lamont> (since debian-installer wants it and all that...)
[08:57] <infinity> lamont: Already discussed with Kamion, who claimed he'd sort out whose bug it is. :)
[08:57] <lamont> oh, even better.  danke
[11:42] <dholbach> hey guys!
[11:43] <dholbach> BenC: what could we put on a wiki page to make bug triage for Kernel bugs easier?
[11:45] <dholbach> http://wiki.ubuntu.com/DebuggingUSBStorage http://wiki.ubuntu.com/DebuggingIRQProblems http://wiki.ubuntu.com/DebuggingHardwareDetection are not really it, are they? :/
[02:53] <zul> heylo
[04:24] <BenC> dholbach: Check the CategoryKernel stuff
[04:24] <dholbach> i see
[04:26] <dholbach> um...
[04:27] <dholbach> i don't want to build a kernel... i was more looking for information for bug triagers
[04:27] <dholbach> questions they can ask, stuff we can point people to, if they have kernel problems
[04:27] <dholbach> if you want people to triage kernel bugs next week's friday - it'd be great to have something like that
[04:29] <BenC> there's a lot of that under CategoryKernel
[04:29] <BenC> debugging stuff
[04:30] <BenC> the main thing that needs is questions like "Was this under dapper"
[04:30] <dholbach> what else? lspci -v, lshal, dmesg - is that kind of stuff of any use?
[04:33] <BenC> lspci -vv
[04:33] <BenC> lspci -vvn
[04:34] <BenC> dmesg
[04:34] <BenC> https://wiki.ubuntu.com/MemoryTest
[04:34] <BenC> https://wiki.ubuntu.com/DebuggingSystemCrash
[04:34] <BenC> https://wiki.ubuntu.com/DebuggingIRQProblems
[04:34] <BenC> https://wiki.ubuntu.com/DebuggingProcedures
[04:34] <BenC> those are the four URL's I use to point people
[04:35] <BenC> also, bug submitters need to be explicitly told "attach the output to these commands, do not paste them into a comment"
[04:35] <BenC> otherwise we get a huge unreadable comment listing
[04:36] <BenC> Sound problems need to have alsa-team subscribed
[04:37] <BenC> 100 new emails in my bug folder just since last night
[04:37] <BenC> The problem for me is, a lot of those are bugs getting closed, but I still have to read all the emails to find out
[04:38] <BenC> a lot of them can be closed, but I still have to go through the emails to verify it is a problem I fixed
[04:41] <dholbach> same here
[04:41] <dholbach> for gnome land we get millions of bug reports :/
[04:41] <dholbach> but i'll knock up a webpage with the stuff you mentioned
[04:41] <dholbach> DebuggingKernelProblems?
[04:42] <dholbach> does that sound good?
[04:42] <BenC> yeah
[04:55] <infinity> Would be nice if Malone could put the bug status in the Subject line, now that I think about it.
[04:56] <infinity> So you could skip mails with "[Fix Released]  foo bar baz" if you're in a hurry and catch up on them later.
[04:57] <mjg59> Hrm.
[04:58] <mjg59> Why is hibernate really really really slow on battery and fast on ac?
[05:02] <mjg59> And why has it just worked properly for me? IT BURNS
[05:05] <BenC> mjg59: where's the script that allows people to add modules that should get unloaded/reloaded on suspend/resume?
[05:05] <BenC> s/script/config/
[05:09] <mjg59> BenC: /etc/default/acpi-support
[05:13] <BenC> thanks
[05:14] <mjg59> Hm. Now I seem to have fixed it, but I'm not sure how...
[09:22] <zul> BenC: ping..
[09:33] <zul> bored bored bored..