[12:00] <yuval> Kamion, sivang, https://bugzilla.ubuntu.com/show_bug.cgi?id=1231
[12:03] <mdz> is there any way to play a Vorbis stream from firefox, e.g. as background audio for a web page?
[12:04] <Kamion> quick review of https://bugzilla.ubuntu.com/show_bug.cgi?id=2123 please?
[12:04] <Kamion> yuval: thanks, I noticed your updates; I'll have a look tomorrow
[12:04] <Kamion> (since most of my other RC stuff is out of the way now)
[12:05] <Kamion> yuval: it seems to be a d-i bug too
[12:05] <Kamion> he_IL.UTF-8 UTF-8
[12:05] <Kamion> he_IL ISO-8859-8
[12:05] <Kamion> but languagelist has:
[12:05] <Kamion> Hebrew;he_IL;he_IL;he;IL;he_IL:he:en_GB:en;kbd=iso08.f16(utf8)
[12:06] <Kamion> ... in fact, why don't I just have a look tonight ...
[12:06] <yuval> Kamion: I checked debian installer and there is no problam there.
[12:06] <Kamion> really? languagelist seems to be the same
[12:07] <yuval> ubuntu have a more update base-config, maybe something changed.
[12:07] <thom> Kamion: that patch looks reasonable at plain reading
[12:08] <Kamion> thom: actually there's a minor bug in the LENOF macro (insufficient parentheses at one point), although it doesn't affect this case
[12:09] <Kamion> yuval: don't see anything relevant in the base-config changelog
[12:10] <thom> g'night
[12:12] <Kamion> thom: but thanks
[12:12] <Kamion> will upload now
[12:12] <thom> np
[12:12] <yuval> Kamion: What is he_IL;he_IL;he;IL;he_IL:he:en_GB:en ?
[12:12] <thom> i'm almost at matchsticks to hold eyes open here :-)
[12:13] <Kamion> yuval: too complicated to explain here, look in the languagechooser source ...
[12:13] <yuval> I'll do that.
[12:14] <Kamion> I'm trying it out with the first two he_IL entries changed to he_IL.UTF-8
[12:15] <Kamion> not that I speak Hebrew, but fortunately I can drive the installer almost blind :)
[12:15] <yuval> lol
[12:16] <mdz> Kamion: how confident are you in this archive-copier change?
[12:16] <Kamion> mdz: seems to work when I test package-cache-names at the command line
[12:16] <mdz> Kamion: it's tempting to revert to copying everything, considering the late hour
[12:17] <Kamion> that seems overkill; this is a really easy change
[12:17] <yuval> Kamion: Do you know how "Hebrew" looks in Hebrew?
[12:17] <Kamion> yuval: not even that
[12:18] <mdz> Kamion: ok, if you're confident, it's OK with me
[12:18] <Kamion> mdz: if you're really uncomfortable with the change to C code I can fix it in debian/postinst, but that's harder to test
[12:18] <Kamion> (since I have to boot d-i to test it in that case)
[12:19] <mdz> Kamion: it looks bigger than it is, due to the renaming
[12:19] <mdz> I think it's fine
[12:21] <Kamion> uploaded
[12:22] <yuval> I'll be back in half an hour. I go out with my dog.
[12:25] <sivang> Kamion : hand me over an iso when you set it up, I'll test it on my USB kbd
[12:25] <sivang> Kamion  : however if the languagle list is the same like debian's , I don't see why it should break.
[12:27] <Kamion> that does puzzle me
[12:27] <sivang> I mean, you havn't touched anything on the lists right?
[12:28] <Kamion> we are somewhat behind on languagechooser/countrychooser/base-config, and the interactions there are subtle
[12:28] <sivang> BTW : how can I svr or else get U-I's source?
[12:28] <Kamion> (and quick to anger)
[12:28] <Kamion> sivang: from the source packages in the archive. we have no revision control yet.
[12:28] <sivang> ok
[12:29] <sivang> apt-get source u-i/d-i ? ;)
[12:29] <Kamion> it's easy for me, since I uploaded most of it and can just look in ~/src/ubuntu/ :-)
[12:29] <Kamion> no, individual source packages
[12:29] <Kamion> so e.g. apt-get source languagechooser
[12:29] <sivang> oh
[12:29] <sivang> much better ;)
[12:30] <sivang> for iso testing..;)
[12:31] <sivang> Kamion : the languagechooser scripty is the whole source?
[12:32] <Kamion> sivang: it has complex interactions; if you're interested in figuring out how d-i works it's probably much easier to check out the upstream source from subversion, at the moment
[12:32] <sivang> Kamion : oh, you removed all the remarks and stuff? :)
[12:33] <Kamion> no
[12:33] <sivang> ok
[12:33] <Kamion> but it's easier to navigate the upstream tree since it's all in one piece
[12:33] <Kamion> erk, he_IL.UTF-8 doesn't even work a little bit for me
[12:33] <T-Bone> Kamion: btw, i've been unfortunately delayed on the ia64 side, but I haven't dropped d-i stuff there. It's just a 1week delay, most likely
[12:37] <sivang> Kamion : you on the USB kbd?
[12:38] <Kamion> sivang: that's not relevant for this part of the bug
[12:39] <Kamion> sivang: don't get too distracted by the keyboard thing; it's connected, but it won't be the thing causing text to be unreadable
[12:40] <sivang> oh , _that_ bug
[12:40] <sivang> :)
[12:40] <sivang> I thought we were discussing the one that fails right when d-i starts.
[12:40] <sivang> *that makes d-i fail on start
[12:41] <Kamion> no
[12:41] <Kamion> hm
[12:41] <Kamion> well, there seem to be two conflated things in that bug
[12:42] <Kamion> I'm talking about Yuval's report of corrupted text in the second stage
[12:42] <Kamion> which is really a different bug and probably shouldn't be part of #1231
[12:42] <sivang> right. 
[12:43] <sivang> I never went on with the hebrew install fearing it would break other thigns, so I never got to see this ;-)
[12:43] <yuval> sorry... he_IL.UTF-8 doesn't change anything?
[12:43] <sivang> moreover, I don't think I could survive a complete hebrew d-i install ;-))
[12:44] <yuval> sivang: Why not?
[12:44] <Kamion> yuval: it's corrupted differently; I'm trying to investigate
[12:47] <sivang> yuval : ah, it's a long story. I am so used to english on computers by now, it freaks me sometimes people have "enabled" setups. I'm using english to correspond even with my closest freinds, so I don't really need it. Plus if you have ever tried KDE/GNOME with hebrew, it's not the most pleasent experience trying to figure out where everything is put. issue of taste and experience. 
[12:48] <sivang> yuval : using a bank's site or anything close does require me to have the suitable fonts for hebrew display on the browser, however all interactions with them are numbers and english password letters - so everything's fine.
[12:49] <Kamion> oh, we don't have jfbterm in base, that might explain it ...
[12:49] <sivang> yuval : But I sure like to see Ubuntu works well with it, just for the sake of hebrew users.
[12:50] <Kamion> maybe it would be easier to attack the Greek bug first
[12:50] <sivang> boy it uses iconv
[12:50] <sivang> Kamion : there's a related greek one?
[12:53] <Kamion> I think so, if only I can find it
[12:54] <Kamion> maybe I'm mixing up Debian and Ubuntu bugs
[12:55] <Kamion> oh, there it is, #1683
[01:00] <sivang> Kamion : I can't explain why, but base-config sure attracts me. reading the sources is the _only_ way to get aroud it?
[01:01] <Kamion> I don't understand; could you rephrase?
[01:03] <sivang> Kamion : ahhm. sorry. I meant, in general reading the package's srcs would be the best way to understand base-config and it's inner workings?
[01:03] <Kamion> that is the case for software in general
[01:04] <sivang> Kamion : true, however having a pre background about it can help understanding it. When I do so, I usually come up with new questions, that usually are left unanswered unless I bug someone on IRC ;-)
[01:04] <sivang> Kamion : "When I do so" = when I read the source
[01:31] <sivang> do we have a script to automagically set up nvidia? (think i've seen that on the wiki somewhere)
[01:44] <Kamion> daniels: OK, I'm a substantial way through investigating this locale problem, so I can take the Greek bug back if you like; it is indeed a subtle set of interactions ...
[01:46] <Kamion> we need languagechooser to set up /etc/environment in order for /etc/init.d/console-screen.sh to work in order for UTF-8 to have any chance of being set up correctly in order for ...
[01:47] <sivang> having hebrew, greek and other  display correctly for ... :)
[01:47] <Kamion> quite
[01:48] <Kamion> ow, gums bleeding, this can't be good
[01:48] <Kamion> think I'll go to bed
[01:48] <T-Bone> heh
[01:48] <T-Bone> have a good night
[01:48] <sivang> Good night Kamion
[01:53] <yuval> Good night (01:53 in israel)
[01:55] <sivang> I think he's still away, he said he had more of a couple of things to do ;)
[02:05] <jdub> mdz: so i have some interesting bugs to figure out
[02:06] <jdub> ubuntu, on pia's oldish p2 toshiba satellite is incredibly slow
[02:06] <jdub> to install and to run
[02:06] <jdub> sid runs at full speed
[02:06] <mdz> is DMA disabled?
[02:06] <jdub> however, running sid with the ubuntu kernel demonstrates the slowness
[02:06] <jdub> nup
[02:06] <jdub> doesn't seem to be i/o related
[02:07] <jdub> everything runs slowly
[02:07] <mdz> what kernel did you use under sid?
[02:08] <mdz> I had a similar machine (though k6 rather than p2), and ubuntu performed similarly to debian
[02:08] <jdub> she's currently using a 2.4 kernel; will have to try a different 2.6 when she comes back from cairns
[02:08] <mdz> did you try the -686 kernel?
[02:09] <jdub> hrm
[02:09] <jdub> i think she might have
[02:09] <jdub> but i'll check that too
[02:13] <mdz> jdub: you have a powerpc machine handy?
[02:13] <mdz> (running ubuntu)
[02:13] <mdz> someone please try installing libopenhbci-dev on powerpc
[02:16] <jdub> hrm, it's totally out of date ;)
[03:06] <jdub> "Disable challenge-response authentication?"
[03:10] <mdz> jdub: Kamion
[03:10] <mdz> jdub: #1586
[03:11] <jdub> ahr
[04:16] <daniels> Kamion: sure, you can have it; seems you figured it out while I was asleep :)
[04:25] <tseng> rc bugs dropping like drunken college chicks
[04:25] <tseng> way to go.
[04:48] <sivang> jdub : i have had witnessed horroble slowness with ubuntu using kerlen 2.6.8-386 on a dell inspiron 8200
[04:49] <sivang> jdub : I tried swithicng to 686 kenrel - didn't help either
[04:51] <sivang> jdub : this is a 1.8Ghz machine, 7200 rpm 256MB ram
[05:12] <T-Bone> past 5AM here, time to get some sleep. Bye all.
[05:54] <fabbione> morning guys
[06:05] <fabbione> mdz: permission to upload fontconfig for 2122, s/|/, in debian/control ofr laptop-detect
[06:06] <fabbione> jdub?
[06:06] <fabbione> or any other devel
[06:14] <mdz> fabbione: yes, that's trivial, no need to ask
[06:15] <fabbione> usual peer-review ;)
[06:15] <fabbione> never lose the good habits :)
[06:16] <fabbione> mdz: i have a wacom driver for the submitter to test and some keyboard layout almost pending (waiting Denis Barber to commit)
[06:16] <fabbione> the latter ARE important for us to have
[06:16] <mdz> ok
[06:19] <lamont> moof
[06:19] <fabbione> hey lamont 
[06:24] <fabbione> lamont: uh what happened?
[06:24] <lamont> must. learn. to. fall. better.
[06:24] <lamont> oof.
[06:26] <fabbione> ahi
[06:28] <mdz> fabbione: do you need a ppc login to build a driver for that guy?
[06:30] <doko> morning
[06:31] <fabbione> mdz: oh right... he has a ppc
[06:31] <fabbione> mdz: no i can build in the chroot
[06:31] <mdz> ok
[06:34] <lamont> mdz: 2113....  should we add cupsys-driver-gimpprint* to supported?
[06:34] <lamont> (works for the user...)
[06:35] <mdz> I think it's too late for that
[06:37] <lamont> certainly want them for hoary...
[06:41] <fabbione> ubuntu-users is way too high traffic now
[06:43] <mdz> yep
[06:51] <lamont> 12MB in the last half of Sept, 7MB in the first 6 days of Oct.  sheesh
[06:55] <sivang> mdz : I'm occasionaly stung when I see you respond almost every user with support request. How do you keep up? ;)
[06:57] <fabbione> sivang: mdz is a bot
[06:57] <fabbione> that's the truth
[06:57] <fabbione> we created an AI living form even able to walk on walls
[06:57] <fabbione> like spideman
[06:57] <sivang> he's so fast also,
[06:57] <fabbione> spiderman even ;)
[06:58] <sivang> amazing
[06:58] <sivang> :))
[06:59] <sivang> are we going to have to bug system tp accept http://bugzilla.u.c/bug# ?
[06:59] <sivang> *the
[06:59] <mdz> sivang: it already does
[06:59] <mdz> fabbione: :-P
[06:59] <sivang> oopst
[06:59] <sivang> ok.
[06:59] <sivang> mdz : thanks :)
[07:00] <sivang> mdz : are you going to leave fabbione get away with this? :-)
[07:00] <mdz> sivang: my revenge is swift and without warning
[07:00] <mdz> when he least expects it
[07:01] <mdz> the black helicopters will hover overhead...
[07:01] <sivang> haha
[07:01] <fabbione> ahhah
[07:01] <sivang> boy, i'm after a sleepless night, almost dropps and can't stop
[07:01] <sivang> fabbione : remeber the drug?
[07:01] <sivang> I think I am in it bad....
[07:04] <fabbione> sivang: yes i do :-)
[07:04] <fabbione> mdz: a patch for 2111 is very very simple
[07:04] <fabbione> mdz: the driver code is nice and dandy ;)
[07:04] <mdz> why do I get the feeling you are not entirely serious?
[07:05] <fabbione> mdz: i am serious
[07:05] <fabbione> it's a 3 lines change
[07:05] <fabbione> right now is:
[07:05] <fabbione> use_bios = TRUE;
[07:05] <fabbione> it will be:
[07:06] <fabbione> if CARD == foo || CARD == bar { use_bios=FALSE
[07:06] <fabbione> and so on
[07:06] <fabbione> the line after will load the override (if any) from XF86Config-4
[07:06] <fabbione> as simple as this ;)
[07:07] <fabbione> there are only to ProSavageDDR models defined/known in the driver
[07:07] <fabbione> that makes things easier
[07:07] <mdz> I hate NFS
[07:07] <mdz> oh I hate it so
[07:11] <lamont> mdz: just repeat after me 'NFS stands for "Not a File System"'
[07:11] <sivang> night everybody - hitting bed after 20hrs of uptime ;-) wish I had some auxilary power source to connect to..
[07:11] <sivang> NFS stands for "Not a File System"
[07:13] <mdz> I am finally caught up on email
[07:13] <mdz> from taking sunday off
[07:13] <mdz> -users is out of control
[07:13] <lamont> mdz: definitely.  No slow s-curve of adoption here...
[07:14] <fabbione> i don't think we expected such a boom
[07:14] <fabbione> did we?
[07:14] <lamont> fabbione: hoped is the word, I think.
[07:15] <mdz> aw, ubuntu-traffic didn't make LWN this week
[07:16] <fabbione> we hoped for a boom... but i don't think we were ready for one of this size.. that's my impression at least
[07:16] <lamont> mdz: so do you want a new cupsys upload with the file then, or we defer moving cupsys-driver-gimpprint's move til hoary and downgrade the bug?
[07:17] <mdz> lamont: that bug isn't RC anyway, is it?
[07:17] <mdz> if it is, that's wrong
[07:17] <plovs_work> i want to thank you guys, my gnome-login-bonobo-hang-error was no longer there this morning, thanks!
[07:17] <mdz> 3 of the remaining 14 bugs are powerpc bugs :-/
[07:18] <lamont> normal
[07:18] <mdz> doko: what is the status of tetex-base?
[07:18] <mdz> (#2066)
[07:20] <doko> mdz: removing the references in the documentation to the removed .html file. should be done in some minutes
[07:20] <mdz> ok
[07:48] <jdub> fabbione: pong
[07:50] <fabbione> jdub: too late :-)
[07:52] <fabbione> good night lamont 
[07:55] <lamont> anyrate, sleep until morning..
[08:19] <fabbione> is "reject_nondigits" something that debconf uses?
[08:20] <fabbione> no
[08:20] <fabbione> never mind
[08:22] <mdz> jdub: this new firefox is supremely buggy
[08:27] <Mithrandir> mdz: should we consider switching to gcc 3.4 for it?  Works on AMD64. :)
[08:28] <tseng> seems to work on x86 now for the most part as well
[08:28] <tseng> besides a pretty major b0rkage with sse2
[08:29] <Mithrandir> tseng: do you have a p4?
[08:29] <tseng> i have a p4 and a p-m
[08:29] <Mithrandir> could you try reproducing 1854?
[08:30] <tseng> i have no HT
[08:30] <doko> mdz (or someone else): somebody wants to review tetex-base (#2066) before upload? chinstrap:~doko/tetex/
[08:30] <tseng> and no apt-get lameness on gzip either
[08:30] <Mithrandir> tseng: ok. :/
[08:30] <mdz> doko: attach a diff to the bug?
[08:30] <Mithrandir> tseng: it doesn't show without HT, it seems.
[08:30] <tseng> Mithrandir: my friend joe is using an HT p4 w/ no serious problems either
[08:31] <tseng> Mithrandir: smp 686 kernel
[08:31] <doko> mdz: ok, but debdiff doesn't show removed files.
[08:31] <Mithrandir> tseng: hmm, ok.  I find it somewhat strange that I'm the only one able to reproduce it.
[08:32] <tseng> ah, you can?
[08:32] <Mithrandir> yes, it's I who filed it.
[08:32] <Mithrandir> :)
[08:32] <fabbione> foo() { echo "" }
[08:32] <fabbione> bar=$(foo)
[08:32] <fabbione> what can cause something like that to hang?
[08:32] <Mithrandir> and I'm hunting for somebody else who can confirm it, if nothing else to prove I'm not on serious amounts of crack.
[08:33] <fabbione> in a debconf environment
[08:33] <tseng> this happens on any apt-get upgrade?
[08:34] <Mithrandir> tseng: no :/  ; apt-extracttemplates seems to hang randomly about 10% of the time.
[08:34] <tseng> mmmm, random
[08:35] <Mithrandir> of course, stracing the problem makes it go away.
[08:35] <tseng> is apt-extracttemplayes heavily threaded?
[08:35] <Mithrandir> which probably means it's some sort of a race issue
[08:36] <Mithrandir> I don't know, mdz could tell?
[08:36] <fabbione> Mithrandir: can you try to call sync between each call to apt-extracttemplates?
[08:36] <tseng> well mdz alluded to nptl there
[08:36] <tseng> we fought some funny bugs with nptl and apps making alot of threads just crapping out
[08:37] <tseng> but it didnt seem to affect distros that ship a hybrid glibc
[08:37] <Mithrandir> fabbione: same problem.
[08:37] <Mithrandir> tseng: can I list the threads somehow, or do they still show up as separate processes in ps?
[08:38] <tseng> not under nptl
[08:38] <tseng> they look like a single process
[08:38] <fabbione> Mithrandir: do you have any idea for the sh script above?
[08:38] <Mithrandir> I doubt apt-extracttemplates uses many, if any, threads.
[08:38] <tseng> but that doesnt explain much about the non-SMP kernel
[08:38] <Mithrandir> fabbione: are you sure you aren't talking to debconf?
[08:38] <tseng> ya..
[08:38] <tseng> the race idea is valid also
[08:39] <fabbione> Mithrandir: it's the config.in for xserver-xfree86
[08:39] <Mithrandir> since it only happens once in about ten times.
[08:39] <fabbione> Mithrandir: it loads debconf that's clear
[08:39] <fabbione> Mithrandir: it hangs in a function similar to the one i wrote above
[08:39] <fabbione> Mithrandir: see https://bugzilla.ubuntu.com/attachment.cgi?id=355
[08:40] <Mithrandir> fabbione: yes, but the foo=$(bar) might try to read from debconf since stdout is redirected.
[08:40] <doko> mdz: done
[08:40] <Mithrandir> tseng: and I have to kill -9 the gzip processes.
[08:41] <Mithrandir> hmm, weird.
[08:41] <Mithrandir> if I reproduce the problem, it doesn't seem to go away until I start another process (ls, dmesg, whatever)
[08:41] <Mithrandir> oh, it does
[08:41] <Mithrandir> it's just slow
[08:52] <mdz> doko: debdiff seems to show the removed file OK
[08:53] <mdz> doko: the diff looks fine to me
[09:01] <mdz> doko: will you upload it?
[09:02] <doko> mdz: done
[09:08] <doko> mdz: the libnet-ph-perl upload discussed in email: does the control file section need to be changed to universe/*
[09:13] <mdz> doko: no, it is overridden
[09:13] <doko> ok, uploading
[09:22] <daniels> mdz: now down to 12 RCs?
[09:23] <mdz> 1080 is getting downgraded as well; it's way too late to mess with fixing bug-buddy
[09:24] <mdz> doko: you forgot to close #2066?
[09:34] <Kamion> jdub: hmm, I guess that message is going to show up for everyone who installed the preview and then upgrades to final
[09:35] <Kamion> that kind of sucks
[09:35] <Kamion> um, guys
[09:35] <Kamion> mdz: wvdial is being interactive on me during the initial installation
[09:35] <Kamion> mdz: can't we stop it doing that? debconfization should be two hours' work for somebody who knows debconf reasonably well
[09:35] <mdz> Kamion: pitti fixed that yesterday, I thought
[09:36] <mdz> maybe it didn't make the CD build?
[09:36] <Kamion> ah, maybe it missed my CD
[09:36] <Kamion> I think I have yesterday's here
[09:36] <mdz> Kamion: #2069
[09:36] <Kamion> OK, fine
[09:37] <Kamion> but hey, SHINY GREEK TEXT
[09:37] <doko_> kamion: should the pmac sound problem be fixed on yesterday's powerpc cd?
[09:38] <mdz> Kamion: any idea what's going on with #1997?
[09:38] <mdz> doko_: <mdz> doko: you forgot to close #2066?
[09:38] <Kamion> doko_: should think so, yeah
[09:39] <Kamion> mdz: absolutely no idea; I'm tempted to ask the submitter to try yet another daily just to be sure
[09:39] <Kamion> but the CD he claims to have really should have all the required code
[09:40] <mdz> Kamion: it doesn't look like we've confirmed that the module is being loaded
[09:40] <doko_> mdz: done :)
[09:40] <mdz> maybe it is in fact being loaded, but something else is wrong
[09:40] <Kamion> it's rather hard to tell without a keyboard
[09:41] <mdz> true
[09:41] <fabbione> brb
[09:41] <mdz> I thought ohci-hcd was built in on powerpc for just this reason
[09:42] <Kamion> it's not; it really shouldn't need to be
[09:42] <mdz> hmm, so there's no keyboard until hotplug runs?  that's rather scary
[09:42] <mdz> what happens if fsck fails?
[09:43] <Kamion> dunno :)
[09:43] <Kamion> mdz: fix in #1683 OK by you, in principle? I need to do more testing
[09:43] <Kamion> for Greek it works for me all the way up to a desktop
[09:43] <Kamion> incidentally I really must learn Greek, because it is so much prettier
[09:44] <mdz> Kamion: so the bug is that it wasn't generating the locale?
[09:44] <Kamion> it wasn't generating it early enough
[09:44] <mdz> ah
[09:44] <Kamion> it was being generated in base-config or something, but that's too late; it needs to be there by the time /etc/init.d/console-screen.sh runs
[09:44] <mdz> and the fix is to have prebaseconfig generate the locale in the target?
[09:45] <mdz> if so, sounds fine
[09:45] <Kamion> right
[09:46] <Kamion> anyway, I need to shut down, taxi coming soon; back in some hours
[09:57] <seb128> morning
[09:59] <sjoerd_> pitti: ping
[09:59] <seb128> hey sjoerd 
[09:59] <sjoerd> seb128: morning :)
[10:10] <pitti> sjoerd: pong
[10:11] <pitti> sjoerd: sorry, I was away for half an hour
[10:11] <sjoerd> pitti: haha
[10:11] <pitti> sjoerd: ?
[10:11] <sjoerd> pitti: don't apply davidz version of my patch, it's wrong
[10:11] <pitti> sjoerd: the one for the first hotplug event?
[10:11] <sjoerd> pitti: that you say sorry for being away half an hour :)
[10:11] <sjoerd> pitti: yes
[10:11] <pitti> sjoerd: last_seq_number or so
[10:13] <sjoerd> pitti: it can kill off hotplug processing of hal, if events come at a bad timing
[10:13] <pitti> sjoerd: well, since the current version works great, I rather not change anything unless I find a bug
[10:13] <sjoerd> pitti: i'll send a mail explainign why, after i finish math homework :)
[10:13] <pitti> sjoerd: but thanks for letting me know. Did you already talk to David about this?
[10:13] <pitti> sjoerd: ah, okay. 
[10:14] <sjoerd> david is probably still sleeping anyway
[11:15] <thom> Mithrandir: hrm, do you know if firefox-locale-no has a PR1.1 release?
[11:16] <Mithrandir> thom: it does, I just haven't made a package yet.
[11:16] <Mithrandir> at least pr1.0 release.
[11:18] <thom> oh, rock
[11:18] <thom> we might get it locaalised to one language then
[11:18] <thom> afaics, french and german hasn't happened yet
[11:19] <Mithrandir> I can look at fixing it up and dropping the package into experimental tonight
[11:19] <daniels> thom: it's so nice
[11:20] <daniels> thom: remind me to buy you a beer some day
[11:20] <thom> *g*
[11:20] <thom> told ya
[11:21] <daniels> thom: except mine is superior to yours
[11:21] <pitti> jdub, mdz: IMHO we can downgrade #2131, do you agree?
[11:21] <daniels> thom: they must've fixed the keyboard layout in a later revision :)
[11:22] <daniels> thom: your keyboard LED is white, no?
[11:22] <thom> no, orange
[11:22] <daniels> ahr, mine's orange too
[11:23] <daniels> I thought yours was white for some reason
[11:23] <daniels> strikes me as kind of odd to put an orange LED there
[11:23] <daniels> when white provides more light and IMO looks better
[11:23] <daniels> ho hum
[11:24] <daniels> dinner time
[11:28] <sabdfl> Kamion: fun and games with openssh-server this morning
[11:30] <sabdfl> Kamion: i installed openssh-server a month ago and never touched the config
[11:30] <sabdfl> this morning'supdate is asking me about passwords being disabled and challenge response...
[11:30] <sabdfl> seems like that quesiton assumes the default is passwordauth yes, so if your config is "no" then it assumes you must have edited it
[11:31] <sabdfl> but i think your default is passwordauth no
[11:31] <sabdfl> so this question is getting asked
[11:31] <sabdfl> can we fix that please?
[11:38] <sabdfl> seb128: morning
[11:38] <sabdfl> Setting up desktop-file-utils (0.9-0ubuntu1) ...
[11:38] <sabdfl> File '/usr/share/applications/gnumeric.desktop' contains invalid MIME type 'comma-separated-values' that is missing a slash
[11:38] <sabdfl> ?
[11:38] <seb128> hello sabdfl 
[11:38] <seb128> sabdfl: yeah, the new desktop-file-utils reports problems instead of just ignoring the files
[11:38] <sabdfl> seb128: cool, lint!
[11:39] <sabdfl> or... lint-free!
[11:39] <sabdfl> nice that there was only one glitch when that got turned on. neat work, thank you
[11:39] <seb128> sabdfl: so gnumeric.desktop has a problem, I put that on my todo list for today :)
[11:44] <thom> great, now there are people filing dups upstream of the firefox crasher
[11:44] <thom> they're never gonna fix it in a week :/
[11:45] <seb128> thom: you're going to fix it, aren't you ? :)
[11:46] <thom> i stared at it till my eyes blead, and pitti's dug into it too
[11:49] <daniels> eh? surely the mime type for csv would be something like text/csv or text/comma-separated-values
[11:51] <daniels> mdz: dbus-monitor seems to be completely non-functional in some cases. amircornot?
[11:51] <daniels> or ratemybugseverity, whichever you like
[11:54] <daniels> wow, IE really is terrible.
[11:54] <daniels> konq was cooler about three years ago
[11:54] <thom> get netbooting, kid
[11:55] <daniels> thom: er
[11:55] <daniels> thom: so we're on a 4GB plan, right.  but Dad didn't see fit to tell me that he forgot to ask for unlimited and that we were on 4GB until I'd sucked 2.3GB in the first 24-48h.
[11:56] <Kinnison> pitti: ping
[11:56] <daniels> thom: and that was about a month ago, and we haven't rolled over to a new month yet.
[11:56] <daniels> thom: so I'll download it at uni tomorrow.
[11:56] <thom> oops.
[11:56] <daniels> yeah
[11:56] <daniels> hence the note on StaffCalendar
[11:59] <thom> oh man. i saw gpl'd flash impementation on ubuntu-devel and nearly died
[11:59] <daniels> pitti: hal is being really weird on Kinni's system
[11:59] <daniels> pitti: the first word that comes to mind when describing his system starts with 'b' and ends in 'ong'
[12:01] <pitti> Kinnison, daniels: pong
[12:01] <pitti> back from lunch
[12:01] <Kinnison> pitti: hal is being very odd at me
[12:01] <pitti> Kinnison: some details?
[12:01] <Kinnison> pitti: Well; on first boot; if I plug a usb device in; hal knows nothing about it
[12:01] <daniels> pitti: if you have binaries or sources around for -1ubuntu4, that may help isolate Kinni's problem
[12:01] <Kinnison> pitti: but restarting it seems to help
[12:02] <pitti> daniels, Kinnison: I can reconstruct ubuntu4, I just need to remove four patches
[12:02] <pitti> Kinnison: did it work with ubuntu4?
[12:03] <pitti> Kinnison: most of the patches fix only segfaults
[12:04] <Kinnison> pitti: I don't remember
[12:04] <pitti> Kinnison: you can try removing the patch fix_first_hotplug.patch
[12:04] <pitti> Kinnison: Shall I build you a package or do you want to build it yourself?
[12:04] <Kinnison> pitti: As I said; it only manifests itself when I first boot
[12:04] <Kinnison> pitti: so if I restart hal; it'll most likely work
[12:04] <sjoerd> Kinnison: if you unplug your usb device, restart hal, plug it in again
[12:04] <pitti> Kinnison: exactly, the mentioned patch fixes sort of a race condition
[12:04] <sjoerd> Kinnison: what happens then
[12:06] <pitti> Kinnison: also, does the second plug work? We had problems only with the first plug in after hal start
[12:07] <sjoerd> pitti: the race is still there, it's just less likely :(
[12:07] <Kinnison> sjoerd: I'll try the unplug/restart/plug now
[12:07] <pitti> Kinnison: you shouldn't need to reboot
[12:07] <pitti> Kinnison: /etc/init.d/dbus-1 restart should do
[12:07] <sjoerd> pitti: only restarting hal should be enough (then your gvm isn't dead)
[12:08] <pitti> sjoerd: no, g-v-m was recently fixed to reconnect :-)
[12:08] <Kinnison> Right; so if I unplug/restart/plug I get:
[12:08] <sjoerd> pitti: ooooh... patch patch patch :)
[12:08] <Kinnison> g-v-m notices and mounts
[12:08] <Kinnison> hal-device-manager doesn't notice
[12:08] <pitti> what? that's odd, never seen this
[12:09] <daniels> jesus, how hard is it to do transparent png?
[12:09] <pitti> there is no device and volume in hal-device-manager?
[12:09] <Kinnison> pitti: indeed; but if I restart hal-device-manager it shows up
[12:10] <plovs_work> pitti, just to let you know, my camera now automatically opens gtkam and I get my photo's, *all* my  devices are working now! Thank!
[12:10] <sjoerd> h-d-m is easily confused if hal restarts
[12:10] <pitti> Kinnison: well, you must restart device manager anyway
[12:10] <pitti> Kinnison: for me it crashes if hal is stopped
[12:11] <Kinnison> pitti: so; unplug/dbus-restart/plug works okay
[12:11] <pitti> Kinnison: this is indeed the "first hotplug gamble" bug
[12:12] <pitti> Kinnison: neither David, nor Sjoerd, nor me found a clean solution for this
[12:12] <thom> fabbione: dude, soon your head will be so big you won't fit through doors :-)
[12:12] <pitti> Kinnison: it usually works from the second plug on, but the first one is still a problem sometimes
[12:12] <sjoerd> pitti: no there is a ``clean'' solution, but it requires kernel support and isn't mainline yet.....
[12:12] <pitti> thom: and he will soon get X-shaped legs :-)
[12:13] <pitti> Kinnison: no problem, but it is good to know that there isn't yet another bug
[12:13] <pitti> sjoerd: what will change?
[12:13] <sjoerd> Kinnison: what happens if you start hal manually after the first boot
[12:14] <sjoerd> pitti: the hotplug seqnum will be published in /sys
[12:14] <daniels> fabbione: wow, bong
[12:14] <pitti> sjoerd: ah, sounds nice
[12:14] <sjoerd> pitti: so it's initialised correctly
[12:14] <daniels> fabbione: should've CCed me on that bug
[12:14] <daniels> fabbione: what was the problem?
[12:14] <pitti> sjoerd: was that the reason for the uniniialized random sequence number?
[12:15] <pitti> sjoerd: so if the kernel publishes the sequence number, hal can just sort them?
[12:15] <Kinnison> sjoerd: how do you mean?
[12:15] <fabbione> daniels: you botched the backport of the driver
[12:15] <daniels> fabbione: yeah ... how?
[12:15] <sjoerd> Kinnison: put a exit 0 at the start of the hal init script, reboot, start hald --daemon=no --verbose=yes
[12:15] <daniels> fabbione: i'd be interested in the interdiff
[12:16] <sjoerd> Kinnison: see what happens then, should be interesting
[12:16] <pitti> sjoerd: I think the whole concept of partition detection in hal is somewhat flawed
[12:16] <pitti> sjoerd: why the kernel can't just look at the devices exported by the kernel?
[12:16] <fabbione> daniels: you added only one .h file out of several files required. and forgot to modify the Imakefile to reflect the changes..
[12:16] <pitti> sjoerd: why do partition detection again?
[12:16] <Kinnison> sjoerd: I'll try to remember that for my next reboot
[12:16] <fabbione> thom: ahahah
[12:16] <daniels> fabbione: weird, I could swear I did that
[12:16] <daniels> fabbione: although if I just forgot to redo the patch in DBS, I'll kick myself
[12:16] <sjoerd> pitti: hmmm, partition detection in hal isn't really a problem
[12:17] <pitti> sjoerd: it was
[12:17] <daniels> fabbione: oh well.  next time something like that comes up, please CC me.
[12:17] <fabbione> daniels: well.. one way or the other.. stuff was missing
[12:17] <daniels> fabbione: yeah
[12:17] <sjoerd> pitti: yeah, because it did it differently from the kernel
[12:17] <pitti> sjoerd: and I heard from David (or some other guy) that it is still hackish
[12:17] <sjoerd> pitti: hal's logic behind it is weird.. the volume_id stuff is nice
[12:18] <sjoerd> pitti: but as you know by now, hal's concept is great, it's code however....
[12:19] <pitti> sjoerd: all these segfaults I fixed recently showed that hal does not care a s**t about checking externally acquired values ...
[12:19] <pitti> sjoerd: so I expect more segfaults in the future :-(
[12:20] <pitti> sjoerd: but right, the concept is nice
[12:20] <pitti> sjoerd: maybe David can be convinced to rewrite the stuff in python :-)
[12:20] <sjoerd> pitti: i don't think that helps
[12:20] <sjoerd> pitti: you can write shitty code in any language
[12:20] <pitti> sjoerd: yes, but you cannot produce segfaults so easily
[12:20] <sjoerd> some parts of the design are flawed to begin with
[12:21] <sjoerd> if the freebsd guys want to port hal they basically need to rewrite it :(
[12:21] <pitti> sjoerd: so if they do, maybe they design it better
[12:22] <pitti> sjoerd: the problem is that I personally only need a fraction of hal's features
[12:22] <pitti> sjoerd: we only need it for detecting removable media, but hal does a thousand other things
[12:23] <sjoerd> well need.. you only currently use those 
[12:23] <pitti> sjoerd: OTOH, if we had devised our own hotplug agent, we would not have such nice features as gphoto camera autohandling
[12:24] <fabbione> daniels: i also have a pending patch for 2111
[12:24] <fabbione> daniels: want to review?
[12:24] <sjoerd> pitti: hal's concept for a hw database of your system is great. but nobody is going to use it if the info they isn't at least mostly  there
[12:24] <pitti> but anyway, on many systems it runs quite nicely now
[12:25] <sjoerd> pitti: so currently hal is publishing a lot of info, which basically nobody uses that
[12:25] <pitti> sjoerd: right. Network devices are a field that can still be exploited
[12:25] <sjoerd> not only that, video device detection is in current cvs which could be used by some apps
[12:25] <pitti> sjoerd: in Hoary, we also want proper device labels instead of sdasomething
[12:26] <fabbione> daniels: http://people.ubuntulinux.org/~fabbione/
[12:26] <pitti> sjoerd: hal's ID detection/assignment is probably good for this
[12:26] <sjoerd> pitti: in sarge+1 i want al the hotplugging mounting stuff to work out of the box :)
[12:26] <pitti> sjoerd: great!
[12:26] <fabbione> daniels: 036 for the wacom, 991 for the savage
[12:26] <sjoerd> pitti: and that includes nice mount points
[12:26] <pitti> sjoerd: it would be nice if Ubuntu and Debian used the same system
[12:26] <daniels> fabbione: sure, I'll review it in a sec
[12:27] <fabbione>   307346 total
[12:27] <pitti> sjoerd: my idea was to have g-v-m maintain a database volume id -> device name and use this as mountpoint and label
[12:27] <fabbione> we are close to 310000 lines of patches for X
[12:27] <pitti> sjoerd: not device name, device label
[12:27] <sjoerd> pitti: i'm currently figuring out how svn-buildpackage works, to put debians hal and gvm under version control
[12:27] <pitti> sjoerd: pmount already supports device labels, BTW
[12:27] <daniels> fabbione: uh
[12:27] <fabbione> it will take AGES to review them for X.org
[12:28] <daniels> fabbione: don't really like 911
[12:28] <pitti> sjoerd: BTW, would you consider using pmount in Debian? Instead of fstab-update?
[12:28] <fabbione> daniels: what do you think is wrong?
[12:28] <sjoerd> pitti: ofcourse, i'm telling everybody that i need to really look at it sometime :)
[12:28] <daniels> fabbione: I would prefer that you used the first block that you commented out to do the detection
[12:28] <fabbione> daniels: i moved it after the chipset probe
[12:28] <daniels> so, set the default according to which chipset it is
[12:28] <pitti> sjoerd: well, I would maintain it for both Ubuntu and Debian
[12:28] <sjoerd> pitti: not changing fstab with al it's problems is a big plus 
[12:28] <daniels> fabbione: oh, there's a chipset probe in between there. bong.
[12:28] <pitti> sjoerd: but it makes no sense to put it in Debian if nothing uses it
[12:28] <fabbione> daniels: i need to have the pciId stuff
[12:28] <fabbione> daniels: otherwise it's the same
[12:29] <daniels> fabbione: ok, I'm happy with that
[12:29] <pitti> sjoerd: the other big plus is hal-running-as-user
[12:29] <fabbione> daniels: yes, but i still want a user test
[12:29] <fabbione> daniels: do you have a normal savage to test on?
[12:29] <daniels> fabbione: could you please give me an interdiff between the original #036 and the new one?
[12:29] <sjoerd> pitti: i don't really find hal running as user very usefull currently, but my view is somewhat different
[12:29] <daniels> fabbione: maybe, but it's still packed away; I can test later on tonight
[12:29] <fabbione> daniels: nope.. i fried away the old 036
[12:30] <sjoerd> pitti: although i've got the design in my head to make hal run mostly as user, while still doing everything
[12:30] <fabbione> daniels: the old 036 was adding xf86wacom.h and that's it
[12:30] <pitti> sjoerd: what's wrong with hal in Ubuntu? It works fine
[12:30] <fabbione> daniels: there was nothing more than that
[12:30] <sjoerd> pitti: don't know if i've got the time to code it
[12:30] <pitti> sjoerd: the only real reason to run hal as root were the callout scripts
[12:30] <pitti> sjoerd: like fstab-update
[12:30] <sjoerd> pitti: i know, for the stuff ubuntu uses it it's great
[12:31] <pitti> sjoerd: but we want to keep hal policy free, so we do not want that anyway
[12:31] <sjoerd> pitti: don't forget filesystem label reading
[12:31] <pitti> sjoerd: hal can read plugdev devices
[12:31] <sjoerd> pitti: yes, but not the ones on my harddisk
[12:31] <pitti> sjoerd: just not fixed disk partitions, but we do not want hal to be able to mess them up
[12:31] <pitti> sjoerd: you can put hal into 'disk', but that is equivalent to root
[12:32] <pitti> sjoerd: actually we need a read-only 'disk' group
[12:32] <sjoerd> pitti: yes, that's why i say that i find running hal as user useless :)
[12:32] <pitti> sjoerd: in Hoary we will probably solve that with mandatory access control
[12:32] <daniels> fabbione: sorry, but the only S3 cards I have are early Trios, and a Virge
[12:32] <sjoerd> pitti: when using gnome-vfs with hal it's really nice to get your filesystem labels read
[12:32] <sjoerd> pitti: so my os X partitions shows ``OS X volume'' instead of idedisk1 ;)
[12:33] <pitti> sjoerd: if I look at the many segfaults and stack corruptions of hal, I won't ever allow hal to run as root
[12:33] <daniels> what I do have here is Cirrus, Tseng, i740, r200, Voodoo (probably Voodoo2?), GeForce2, soon a Number9 card, and a G400
[12:33] <daniels> and the i855 in here
[12:33] <sjoerd> pitti: the design in my head has only a bare minimum of hal code running as root and the others as user
[12:33] <pitti> sjoerd: as I said, in a MAC system we could allow read-only access to disks
[12:33] <pitti> sjoerd: this sounds better
[12:33] <ross> ooh, my ipod's volume is labelled "ross' ipod"
[12:33] <pitti> sjoerd: so you have two hald processes then?
[12:34] <sjoerd> pitti: yes, split one off before dropping prives
[12:34] <pitti> ross: pmount /dev/sda2 "ross' ipod" :-)
[12:34] <sjoerd> pitti: with communition though a pipeline.. to do request like ``read this devices label''
[12:34] <ross> pitti: sweet
[12:35] <ross> e2label works on mounted filesystems, right?
[12:35] <pitti> ross: pmount support is ready, we just need g-v-m support for this
[12:35] <pitti> ross: I don't know
[12:36] <sjoerd> pitti: do you want to do it in gvm or some other layer ?
[12:37] <sjoerd> for debian i think another layer is nicer, so k-v-m could use it too
[12:37] <pitti> sjoerd: actually I would have liked to do all this magic in a DE agnostic layer
[12:37] <pitti> sjoerd: so I could even benefit from it in fvwm or the command line
[12:37] <pitti> sjoerd: but in Ubuntu, I think g-v-m is the right place
[12:38] <ross> pitti: hal needs to monitor the filesystems for label changes ;)
[12:38] <pitti> sjoerd: the database can be DE agnostic, just the invocation must be rewritten for KDE etc.
[12:38] <pitti> ross: good catch, I will ask upstream about that
[12:38] <sjoerd> pitti: you know that rml plans to changes g-v-m's design completely
[12:38] <pitti> ross: OTOH, it's difficult to detect
[12:38] <sjoerd> ross: that would mean polling for now
[12:39] <sjoerd> when/if the kernel gets the extra event stuff via netlink it should be easy though
[12:39] <sjoerd> thatis if the kernel itself knows the volume has changed
[12:39] <pitti> sjoerd: oh, then it might be worth discussing that with him
[12:39] <pitti> sjoerd: if g-v-m supported device labeling upstream, this would be nice
[12:40] <ross> yeah, i can see it being a pita
[12:40] <pitti> ross: I think it would be enough to read the label on plugin
[12:40] <pitti> ross: (which is done anyway)
[12:40] <sjoerd> pitti: let's see if i can find the mail
[12:40] <pitti> ross: so a unplug/replug would suffice
[12:40] <pitti> brb
[12:41] <sjoerd> pitti: Message-Id: <1095697479.3666.16.camel@betsy.boston.ximian.com>
[12:42] <sjoerd> pitti: on gnome's utopia-list
[12:42] <pitti> sjoerd: thanks, I will look at it
[12:44] <pitti> sjoerd: damn, the mail archives don't recognize the Message-Id as search term
[12:46] <ross> pitti: unplug/replug doesn't work on / :)
[12:46] <pitti> ross: it's tricky :-)
[12:46] <pitti> ross: well, unplug certainly works :-)
[12:48] <pitti> sjoerd: any idea how I can search for a message id?
[12:49] <sjoerd> hmmm no ;)
[12:49] <sjoerd> Date: Mon, 20 Sep 2004 12:24:39 -0400
[12:49] <sjoerd> should be easier if it's not in your mailbox :)
[12:50] <pitti> sjoerd: thx
[12:51] <sjoerd> pitti: i'm planning to create an alioth project, to do hal and g-v-m's maintaince under version control and probably an utopia discussion list
[12:52] <sjoerd> pitti: do you feel like eventually discussing these ideas on a list there ?
[12:52] <pitti> sjoerd: of course
[12:52] <pitti> sjoerd: just read the mail, in fact I already saw it in my mbox
[12:52] <pitti> sjoerd: I like the notification area plan, we urgently need this anyway
[12:52] <pitti> sjoerd: a small icon in the panel which allows to unmount devices
[12:53] <sjoerd> a sort of eject icon or something..
[12:53] <daniels> fabbione: sorry, little sister was demanding a story
[12:55] <pitti> sjoerd: right
[12:55] <pitti> sjoerd: https://bugzilla.ubuntulinux.org/show_bug.cgi?id=980
[12:56] <ross> hm, why doesn't my ipod disk appear as a removable device in hal?
[12:57] <pitti> ross: hmm, it should
[12:57] <pitti> ross: works fine for carlos and other guys
[12:57] <ross> pitti: well, i can't see a "removable" key in the hal device manager
[12:58] <pitti> ross: oh that. Just ignore it.
[12:58] <pitti> ross: this key has never worked for me correctly
[12:58] <pitti> ross: but is it automatically mounted and such?
[12:58] <ross> yeah
[12:58] <ross> pitta: but i can't unplug it without manually ejecting in a terminal
[12:58] <pitti> ross: this thing again...
[12:59] <pitti> ross: what happens with unmount in the context menu?
[12:59] <sjoerd> pitti: i don't know how exacly ubuntu handles drives
[12:59] <sjoerd> pitti: their not on the desktop are they ?
[12:59] <ross> pitta: it unmounts, but the ipod wants to be ejected
[12:59] <pitti> sjoerd: we reenabled the desktop icons shortly before the preview
[12:59] <pitti> sjoerd: since we needed an easier way to unmount than the "Disks" window
[12:59] <ross> ah the ipod device itself is removable
[01:00] <pitti> ross: oh yes, the ipod still thinks it's connected. Same problem as with Carlos
[01:00] <pitti> ross: however, it it safe to unplug it
[01:00] <pitti> ross: you can also unload the spb2 module
[01:00] <ross> silly me was looking at the partition node
[01:00] <ross> pitti: really nautilus needs unmount and eject entries
[01:01] <pitti> ross: right. Would that be easy to implement, I would have done it long ago
[01:01] <ross> (in the device context menu)
[01:01] <ross> ha
[01:01] <pitti> ross: the problem is that this would involve a fair amount of new code
[01:01] <pitti> ross: actually there is only one option and a low level determines whether to unmount or eject
[01:01] <ross> i was hoping the gnome-vfs+hal code would make it easy
[01:02] <pitti> ross: the problem is not the backend stuff, but the user interface
[01:03] <pitti> ross: actually we should use eject for all removable media; it works fine for usb sticks
[01:03] <pitti> ross: so maybe it is just a matter of calling eject unconditionally
[01:03] <ross> i'm tempted to write a nautilus-python plugin which looks up the mount point in hal and adds an eject entry
[01:03] <pitti> ross: if we can convince Matt to allow the "always eject" change in gnome-vfs, we would be done with it
[01:04] <pitti> ross: I will prepare an updated package and have it reviewed by seb128, mdz and you. What about this?
[01:05] <sjoerd> pitti: the big problem with device eject/umount stuff it to get the user to actually do it...
[01:05] <sjoerd> pitti: most people are used to just pulling usb sticks out
[01:06] <pitti> sjoerd: well, hal will lazily unmount removed devices, that's not the problem
[01:06] <sjoerd> pitti: yeah i know
[01:06] <ross> pitti: you mean call eject on unmount?
[01:06] <pitti> sjoerd: the real problem with this approach is https://bugzilla.ubuntulinux.org/show_bug.cgi?id=1959
[01:06] <pitti> ross: yes
[01:06] <pitti> ross: this works even better than just unmounting with my usb stick
[01:06] <pitti> ross: the device is actually shut down, the LED goes off
[01:07] <sjoerd> pitti: autch
[01:07] <pitti> sjoerd: yes, I told everybody proudly that you can just rip off devices in Ubuntu, and then this report came
[01:07] <sjoerd> pitti: but i mean the problem is getting people to understand that umount manually first is really the right way
[01:08] <sjoerd> apple stuff does that nicely by putting a warning on the device
[01:08] <pitti> sjoerd: see ross' problem
[01:08] <sjoerd> yes
[01:09] <pitti> ross: can you please file a bug report about this? Then I have a place to send a patch to
[01:15] <pitti> ross: I have a minimal patch ready (just changes a FALSE to TRUE)
[01:18] <ross> k
[01:18] <ross> pitti: what component? nautilus?
[01:18] <pitti> ross: gnome-vfs2
[01:19] <pitti> ross: just assign it straight to me (martin.pitt@canonical.com)
[01:20] <ross> pitti: #2134
[01:20] <pitti> ross: thanks!
[01:29] <daniels> fabbione: dude
[01:29] <daniels> fabbione: look in the patches in people.ubuntu.com/~daniels/xorg/
[01:29] <daniels> fabbione: i've already done the X11R6 migration :)
[01:29] <daniels> there are a few patches in there related to it
[01:32] <fabbione> daniels: i know that.
[01:33] <daniels> fabbione: ah, ok
[01:33] <daniels> fabbione: just making sure :) you have to change the man dir and stuff, iirc
[01:34] <fabbione> daniels: yup...
[01:34] <fabbione> wartylog:
[01:34] <fabbione> Start killing X11R6.
[01:34] <fabbione> s/warty/
[01:35] <fabbione> ^^ --->> start <<---
[01:35] <daniels> yeah :)(
[01:35] <daniels> just making sure
[01:42] <fabbione> daniels: i need to play around to understand more stuff
[01:42] <fabbione> daniels: and see how things interact with each other
[01:42] <fabbione> daniels: it's not only question of grabbing a patch here and one there
[01:46] <daniels> fabbione: yeah, fair enough
[01:46] <daniels> fabbione: if you ever need a hand, give me a yell, because I got it pretty much all worked out
[01:46] <daniels> took a while
[01:47] <fabbione> daniels: ain't my fault if you slow :P
[01:47] <fabbione> daniels: but i have your patches.
[01:47] <fabbione> i want to use them as last resource if i don't get a clue
[01:49] <daniels> fabbione: heh heh, I've already done it, buddy; where are your packages? :P
[01:49] <thom> oh man. fedora core 3 has gtk filechooser patches for firefox
[01:50] <thom> they're so sexy
[01:50] <fabbione> daniels: where are yours? do they work? do they actually install?
[01:50] <fabbione> daniels: no.. sorry.. do they actually exists?
[01:50] <fabbione> ;)
[01:50] <pitti> ross: new package is available, see bug 
[01:50] <fabbione> thom: argh... include them :-)
[01:51] <pitti> ross: I would greatly appreciate some testing :-)
[01:51] <thom> tempting, but i might get killed
[01:51] <fabbione> thom: where is that SECURITY updated to the gtk filechooser in firefox?
[01:52] <fabbione> thom: you forgot to add it last upload!
[01:52] <ross> pitti: sure
[01:54] <thom> heh
[01:55] <thom> what the heck is tr_TR ?
[01:55] <pitti> thom: Turkish?
[01:55] <Mithrandir> turkey?
[01:56] <thom> yeah, i guess so
[01:56] <daniels> fabbione: haha
[01:56] <daniels> fabbione: they're on my desktop (well, my portable hard drive)
[01:57] <daniels> fabbione: they're not *entirely* complete, but they work fine with just fd.o xlibs + xorg
[01:57] <daniels> they were running on mdz's laptop for a while
[01:57] <daniels> fabbione: i'll go somewhere where I can upload all the source packages (probably uni) tomorrow and throw them up on fooishbar
[01:57] <daniels> fabbione: they're not entirely complete in terms of separating the .orig.tar.gz and stuff and retargeting to warty, but they're a very good base imo
[01:58] <daniels> fabbione: remember how mdz kept stalking around looking for me until I unbroke his laptop at Oxford? :)
[01:58] <fabbione> daniels: are they printed on paper or something that they lay on your desktop? ;)
[01:58] <fabbione> daniels: no because i was too busy fixing other stuff breaking on xfree86
[01:59] <daniels> fabbione: heh heh. they're on my portable hard drive, which is connected to my desktop right now.
[01:59] <daniels> fabbione: oh, fair enough
[01:59] <daniels> fabbione: but stuff like xresprobe needs to be probed
[02:00] <fabbione> daniels: pkgnum is going to be ... hmmm.. high :-)
[02:00] <fabbione> daniels: i wonder if elmo will hunt me down
[02:02] <daniels> fabbione: yeah, last I saw it was up to about 52 source packages
[02:02] <daniels> i think
[02:02] <daniels> yeah
[02:04] <fabbione> daniels: yes.. that's YOUR split...
[02:04] <fabbione> daniels: mine start with 1 source -> 171 packages -> 171 source packages -> N binary packages
[02:04] <daniels> !!!
[02:04] <fabbione> but 171 is a virtual number
[02:04] <daniels> i think elmo just booked a flight to Copenhagen
[02:04] <Mithrandir> fabbione: don't scare us that much.
[02:05] <fabbione> since it needs to be stripped of all the crap
[02:05] <fabbione> Mithrandir: i am not kidding
[02:05] <fabbione> i am going to reduce X.org in small little tiny packages
[02:05] <fabbione> a lot of them
[02:05] <Mithrandir> one package per file :)
[02:05] <fabbione> daniels: for ex: xorg/xc/libs/zlib <- we can kill this one
[02:06] <fabbione> daniels: there are plenty like that one
[02:06] <fabbione> Mithrandir: almost ;)
[02:07] <daniels> fabbione: yeah
[02:07] <daniels> fabbione: we can kill a fair few things in programs as well
[02:08] <daniels> not to mention a lot of ancient libs
[02:08] <fabbione> daniels: exactly
[02:08] <daniels> tell me you're not building a pex5 source package :P
[02:08] <fabbione> daniels: check my commits
[02:08] <fabbione> daniels: right now i am only building one package
[02:08] <fabbione> daniels: so i mean.. it's not like i am going to build much atm
[02:09] <daniels> fabbione: yeah, I'm watching your commits roll in to debian-x
[02:10] <daniels> it's been a long time since I've checked out any part of the XSF SVN repo
[02:10] <fabbione> daniels: as soon as warty is out i will put up .orig.tar.gz
[02:10] <daniels> fabbione: i think we should co-ordinate a little more on this
[02:10] <fabbione> right now i am working on it only in my (almost inexistent) spare time
[02:11] <daniels> fabbione: because we've got two pretty diverse sets of packages here, so putting out an .orig.tar.gz might be a little hasty ... they tend to stick around
[02:11] <fabbione> daniels: it's not like they can do anything with it
[02:11] <fabbione> daniels: it doesn't build any binary
[02:11] <fabbione> daniels: it's the *-source-* stuff
[02:12] <fabbione> daniels: that is nothing more than grabbing X11R6 from freedesktop :-)
[02:12] <daniels> mmm, if you want to
[02:12] <daniels> if you're going to do that, if you at least grab the xorg-6.8.1.tar.gz (or whatever kem renamed that to -- probably X11R6.8.1-src.tar.gz) from ~xorg/X11R6.8.1/src-single/
[02:13] <fabbione> i did already
[02:13] <fabbione> that's the one i am using
[02:14] <fabbione> but it still needs to be sanitized
[02:14] <fabbione> there are the non-free code that we want to get rid of
[02:14] <daniels> ah, rad
[02:14] <daniels> yeah, mainly fonts
[02:15] <daniels> the autoconfig stuff is fine, remember
[02:15] <fabbione> dude...
[02:15] <daniels> it's basically stuff which is non-dfsg-free -- there's no XFree86 1.1 code
[02:15] <daniels> yeah?
[02:15] <fabbione> read debian-x :-)
[02:15] <ross> pitti: so far so good with new gnome-vfs. it ejected the ipod as expected
[02:15] <fabbione> daniels: -r19
[02:15] <pitti> ross: nice to hear, thanks! :-) Can I quote you on a bug followup
[02:15] <fabbione> daniels: and thread
[02:15] <pitti> ross: ?
[02:15] <ross> sure
[02:16] <tseng> ross: hmm, does the ipod realize its been unmounted also?
[02:16] <ross> tseng: yes, that what was good
[02:16] <tseng> hm neat
[02:16] <tseng> :D
[02:16] <daniels> fabbione: oh right, I just saw Roland waffling about some crap about Xprint and sort of tuned out
[02:16] <ross> pitti: would be nice if in the end nautilus knew if the device was ejectable, like it does with cdroms
[02:17] <fabbione> daniels: i am off for a little while
[02:17] <fabbione> daniels: try to test the savage driver if you can
[02:17] <daniels> fabbione: take cure dude :)
[02:17] <fabbione> daniels: kinda asap :-)
[02:17] <daniels> fabbione: i have no savages, sorry, but I'll try to find someone with one
[02:17] <daniels> fabbione: heh yeha
[02:17] <daniels> also, 'yeah'
[02:18] <pitti> ross: Well, actually an iPod cannot be "ejected"; it will not be spat out of the firewire plug AFAIK :-)
[02:18] <fabbione> daniels: you said you have one in a box
[02:18] <fabbione> daniels: find it and plug it
[02:18] <fabbione> ;)
[02:18] <pitti> ross: so I think the current term "unmounting" is better
[02:18] <fabbione> or ask on #ubuntu or something
[02:18] <azeem> how about 'Remove device'?
[02:18] <azeem> 'unmounting' sounds quite techy
[02:18] <daniels> fabbione: i said I probably did, then I realised it was only Virge and old Trios
[02:18] <pitti> azeem: remove from your desk?
[02:19] <ross> pitti: hm. problem is that ipods are "removable", whereas an external jazz drive is actually removable. in HAL they'll both appear as removable
[02:19] <pitti> ross: you mean the ZIP and cd-roms are ejectable?
[02:19] <ross> cd-roms are from nautilus, zip should be
[02:19] <pitti> ross: right, this probably requires HAL support
[02:19] <tseng> my old usb zip drive actually rockets disks at me ocassionally
[02:19] <tseng> its humorous.
[02:20] <pitti> tseng: but ZIP drives can actually be ejected
[02:20] <tseng> yep.
[02:20] <pitti> but there's nothing to eject on an iPod (at least I would not want it to :-) )
[02:20] <tseng> itunes does call it "ejecting" the ipod however
[02:20] <pitti> hmm. The German translation is far better
[02:21] <azeem> ok, so what about 'disconnect'?
[02:21] <pitti> sounds much better
[02:56] <pitti> jdub: can you please take a look at #2135 (and approve, if appropriate)?
[02:57] <thom> jdub: the firefox crasher was reported upstream on the 3rd of august :(
[03:09] <jdub> thom: good lord
[03:10] <jdub> pitti: approved (and commented)
[03:10] <thom> jdub: we also have basically no translations available for 1.0
[03:13] <pitti> jdub: thanks, uploaded
[03:18] <jdub> thom: do you have a landline?
[03:19] <thom> jdub: wait one
[03:20] <jdub> (thom is stringing up some tin cans and string to the overhead cables AS WE SPEAK)
[03:20] <thom> pretty much
[03:20] <thom> it does mean finding the phone :-)
[03:20] <tseng> the first trans-pacific tin can conversation
[03:21] <jdub> tseng: the ISS is just one big tin can, dude ;)
[03:21] <tseng> hmz
[03:38] <ross> yo yo jdub
[03:39] <tseng> hey ross, ive been thinking about a stupid "feature" on s-juicer if youd entertain
[03:41] <tseng> seems to be an unpopular idea, but i encode both flac + mp3 ... i have a home theater box and an ipod. totem wants to rip the cd twice instead of encode the tracks twice
[03:41] <tseng> jdub: someone is working on packaging beagle/dashboard now :)
[03:41] <ross> tseng: i doubt i'll implement multiple encoding formats
[03:42] <ross> you could hack it with a weird profile in sj-profiles-branch
[03:42] <tseng> hm
[03:47] <jdub> tseng: rock :)
[03:47] <jdub> tseng: so we really should get your mono stuff into universe
[03:47] <jdub> pants off ross
[03:47] <tseng> hm, for warty?
[03:47] <jdub> ross: 0.9 is in :-)
[03:47] <jdub> tseng: sure
[03:48] <tseng> hm ok
[03:48] <jdub> tseng: how much of it is different to sid versions now?
[03:48] <tseng> monodevelop only
[03:48] <ross> jdub: so i heard, pants off!
[03:48] <tseng> which i need to poke here again, missing deps
[03:48] <ross> jdub: did you get my app install mail?
[03:48] <jdub> ross: been travelling for a couple of days, so will get to it soon ;)
[03:48] <tseng> oh and tomboy still isnt in sid afaik
[03:48] <ross> jdub: ok. good time?
[03:49] <jdub> ross: lovely. did an ubuntu presentation in daniels's home town.
[03:49] <ross> jdub: sweet.  i just plugged ubuntu to a techie doing due diligence here
[03:49] <jdub> ross: put up embarrassing pictures of him, pipka and bob2 up on the big screen etc. ;)
[03:49] <daniels> jdub: sorry about the state of the house, btw
[03:49] <ross> ha
[03:49] <tseng> jdub: there is a very minor bug also in my monodevelop (upstream as well) ... missing some file for nermle template, warning issued on New
[03:50] <daniels> jdub: we were all pretty much out for everything but dinner for the last week; particularly me, as I'd get home, and need to clear a day's worth of work after that
[03:50] <jdub> daniels: i didn't find any WMD, so i'm not too fussed.
[03:50] <tseng> but I think in universe that isnt a real blocker
[03:50] <daniels> jdub: heh
[03:50] <daniels> jdub: or catch cooties
[03:50] <jdub> daniels: in communist ivanhoe, the cooties catch you.
[03:50] <daniels> jdub: ahr, I am the cootie monster
[03:50] <ross> should totem in ubuntu be able to play the fluendo stream?
[03:51] <tseng> seb just posted new gst* and totem packages this morning if that helps
[03:51] <ross> hm, good point. i'm, using totem-xine atm
[03:51] <tseng>  deb http://people.ubuntulinux.org/~seb128/gstreamer/  /
[03:52] <tseng> the fluendo guys have been working like mad lately.
[03:52] <ross> yeah
[03:52] <jdub> oh good
[03:55] <jdub> ross: totem+gstreamer in ubuntu should be able to play the fluendo stream now
[03:55] <jdub> now like, "already" not "as of just now"
[03:55] <seb128> no
[03:55] <seb128> current version is bugged
[03:55] <jdub> seb128: oh
[03:55] <seb128> I've asked if I should apply the patch to gstreamer and you said it's not a priority so no
[03:55] <pitti> seb128: I tried the new packages; mpeg2 decoding got a slight bit better, but is still unusable; totem crashed after closing it
[03:56] <pitti> Hi carlos!
[03:56] <carlos> pitti: my hero!!
[03:56] <carlos> :-P
[03:56] <pitti> carlos: Can I bother you to take a look at #2134?
[03:56] <seb128> if we don't have regression I'm all for using the new packages
[03:56] <carlos> let me see it
[03:56] <pitti> carlos: your iPod should finally unmount fine with this
[03:56] <jdub> seb128: i thought i said "because the new releases with BBB's hacking love will be out soon anyway"?
[03:56] <carlos> pitti: oohh, that one
[03:56] <carlos> pitti: I answered it already :-P
[03:56] <carlos> that's why you are my hero :-)
[03:57] <pitti> carlos: oh, I did not notice. Thanks
[03:57] <seb128> jdub: no no, you just said it was not a priority :)
[03:57] <seb128> it was like 3 weeks ago
[03:57] <Kamion> sabdfl: there's already a bug for that; suffice to say there was a rationale for the behaviour I chose and it wasn't a mistake for Debian (long story, basically the config options were shuffled by upstream and *then* the PasswordAuthentication default was changed by me, which confused things muchly), but it's not appropriate for Ubuntu and I'll be changing it
[03:57] <pitti> carlos: I feel honored :-)
[03:57] <ross> seb128: totem from that source is hanging
[03:58] <ross> seb128: i think its a HAL problem
[03:58] <tseng> jdub: btw, is there an env i can set so that dch gets my proper email?
[03:58] <Kamion> with only a little hacking
[03:58] <seb128> ross: hum ... totem uses hal ?
[03:58] <Kamion> jdub: how scared will you be if I tell you that we need to drag jfbterm into warty?
[03:59] <daniels> Kamion: ?!?
[03:59] <ross> seb128: the bacon stuff will if its around
[03:59] <ross> seb128: call yourself a packager? ;)
[03:59] <jdub> Kamion: first, i will be pensive, and raise my eyebrows.
[03:59] <Kamion> thought so
[04:00] <carlos> pitti: you are doing a good work about removable devices in Ubuntu that are really good for normal users and we will be the first with those features :-P
[04:00] <Kamion> needed for certain languages, I'm afraid
[04:00] <ross> seb128: well, i think that is what is happening anywya
[04:00] <Kamion> otherwise they won't be able to read the text in base-config at all
[04:00] <jdub> Kamion: which ones?
[04:00] <pitti> carlos: do you happen to have any other USB devices to test it with?
[04:00] <daniels> jdub: presumably chinese and japanese
[04:00] <Kamion> CJK
[04:00] <daniels> ah, also Korean
[04:00] <daniels> cjk
[04:00] <ross> seb128: watch:
[04:00] <ross> $ totem
[04:00] <ross> libhal.c 2213 : Error sending msg: Service "org.freedesktop.Hal" does not exist
[04:00] <ross> see
[04:01] <daniels> ross: cool!!
[04:01] <Kamion> and possibly a few random stragglers where the Linux console doesn't do their UTF-8 well
[04:01] <carlos> pitti: yes, let me look for my syster's usb pen
[04:01] <daniels> ross: is HAL running?
[04:01] <Kamion> we get away without it for Hebrew apparently; haven't tested Farsi yet although we're probably missing most of the d-i translations for that anyway
[04:01] <ross> daniels: it got upgraded so probably not
[04:01] <jdub> Kamion: this won't matter so much when we're-- yeah -> utf-8 impact?
[04:01] <pitti> carlos: beware, now it is ejected; after unmounting, it will stick in your eyes :-)
[04:01] <seb128> ross: $ grep -i hal totem-0.99.17/configure.in
[04:01] <seb128> $
[04:01] <Kamion> jdub: hmm?
[04:02] <ross> seb128: fun.
[04:02] <jdub> Kamion: when we switch to utf-8 for hoary, what will we use?
[04:02] <Kamion> jdub: base-config actually tries to use it even for UTF-8; apparently it does a better job than the console for some languages
[04:02] <seb128> ross: isn't it ? :)
[04:02] <Kamion> however turning on the framebuffer is *probably* enough for most people
[04:02] <carlos> pitti: it's not mounted automatically...
[04:03] <ross> seb128: now i am getting confused
[04:03] <Kamion> I'll fix up Greek and Hebrew now
[04:03] <pitti> carlos: hmm. What does the device manager say?
[04:03] <carlos> and dmesg shows me it
[04:03] <pitti> carlos: BTW, this can't have anything to do with my gnome-vfs changes
[04:03] <carlos> I know 
[04:03] <pitti> carlos: does device manager show a volume for it?
[04:03] <carlos> I saw the patch
[04:04] <carlos> which one was the command?
[04:04] <ross> seb128: aha, maybe its gnome-vfs?
[04:04] <pitti> carlos: hal-device-manager
[04:04] <pitti> carlos: should also be in the menu
[04:04] <pitti> carlos: unless you did not install the package hal-device-manager
[04:05] <carlos> seems like it's not installed
[04:05] <seb128> ross: the hal supported is turned off in warty
[04:05] <seb128> for gnome-vfs
[04:05] <ross> ross: i bet pitta turned it on for his ipod fixing gnomevfs
[04:05] <seb128> ross: oh, that's nautilus-cd-burner 
[04:05] <seb128> * Use CD selection widget from nautilus-cd-burner, instead of our copy
[04:05] <seb128> for totem
[04:05] <ross> ah ok
[04:05] <seb128> and n-c-b has the hal support on
[04:06] <carlos> pitti: Could not get device list. Make sure hald is running!
[04:06] <ross> fune
[04:06] <ross> i must restart i guess
[04:06] <carlos> seems like hald died
[04:06] <tseng> hald is good at that
[04:06] <pitti> carlos: so your hal does not run. /etc/init.d/dbus-1 restart
[04:06] <carlos> but It was running becuase the iPod was mounted automatically
[04:06] <carlos> hmm
[04:07] <carlos> pitti: it's running
[04:07] <carlos> hal       3196  0.0  0.7  7616 5644 ?        Ds   15:02   0:02 /usr/sbin/hald --drop-privileges
[04:07] <pitti> carlos: just because you restarted it or was it running before?
[04:08] <ross> carlos: i've got one of those! Ds is bad
[04:08] <fabbione> daniels: the savage patch seems to work
[04:08] <daniels> fabbione: awesome.  did you find someone to test?
[04:08] <carlos> pitti: it was running before
[04:08] <ross> pitti: me too
[04:09] <pitti> carlos: so you just cannot communiate with hal any more? Try lshal
[04:09] <fabbione> daniels: the submitter :-)
[04:09] <daniels> fabbione: ah, rad :)
[04:09] <ross> pitta: "Service "org.freedesktop.Hal" does not exist"
[04:10] <carlos> pitti: no, it hangs
[04:10] <daniels> ross: 'pitti'
[04:10] <daniels> ross: as in, da fool
[04:10] <ross> doh
[04:10] <ross> i'm hungry ;)
[04:10] <carlos> carlos@frodo /var/log $ lshal
[04:10] <carlos> lshal version 0.2.98
[04:10] <carlos> libhal.c 696 : org.freedesktop.DBus.Error.NoReply raised
[04:10] <carlos> "Message did not receive a reply"
[04:10] <carlos> *** [DIE]  lshal.c:dump_devices():70 : Couldn't obtain list of devices
[04:10] <pitti> what the hell...
[04:11] <pitti> ross,carlos: can you please restart dbus-1 and see whether this is reproducible?
[04:11] <daniels> hal is totally shot
[04:11] <sjoerd> pitti: if hald is in D state your screwed
[04:11] <daniels> s/hal/&d/
[04:11] <pitti> sjoerd: right
[04:11] <pitti> didn't notice that
[04:11] <carlos> pitti: after restarting dbus I get:
[04:11] <pitti> carlos,ross: so I guess you have to reboot
[04:11] <daniels> oh wow, hald in D
[04:11] <carlos> carlos@frodo /var/log $ lshal
[04:11] <carlos> lshal version 0.2.98
[04:11] <carlos> libhal.c 696 : org.freedesktop.DBus.Error.ServiceDoesNotExist raised
[04:11] <carlos> "Service "org.freedesktop.Hal" does not exist"
[04:11] <carlos> *** [DIE]  lshal.c:dump_devices():70 : Couldn't obtain list of devices
[04:11] <ross> pitti: this is the 3rd time i've got hald in D after upgrading
[04:11] <pitti> daniels: if I had a penny for every time... nevermind
[04:12] <daniels> ross: probably trying to read from a non-existant device
[04:12] <daniels> huzzah.
[04:12] <sjoerd> the only times i've seen hald in D state it where hardware issues
[04:12] <carlos> wow hald does not die
[04:12] <daniels> yeah, same
[04:12] <ross> ooh mine died!
[04:12] <daniels> carlos: yeah, it's in D, which means it's waiting on the kernel
[04:12] <daniels> ross: woo!
[04:12] <ross> hm
[04:13] <ross> now lshal hangs instead
[04:13] <carlos> ross: :-P
[04:13] <daniels> ross: if you can get debugging up the hizzle, knowing just what it's doing might be nice
[04:13] <sjoerd> ross: does dmesg show strange things ?
[04:13] <daniels> ross: yay! has hald gone back into D?
[04:13] <ross> and its back in Ds
[04:13] <daniels> ross: yeah, that's what I thought
[04:14] <sjoerd> ross:  ls -l /proc/$(pidof hald)/fd 
[04:14] <ross> hm, i blame the ipod
[04:14] <sjoerd> ross: what device files does it have open 
[04:14] <ross> sjoerd: null, null, null and null
[04:14] <pitti> @all: this might be because devices are now ejected, which means to power them down
[04:14] <ross> sjoerd: some pipes, 2 sockets and the pci id files
[04:14] <pitti> so it may take forever to open the device
[04:15] <ross> blocking in scsi_wait_req says wchan
[04:15] <ross> i'm totally blaming a powered-down but connected ipod
[04:15] <pitti> ross: try to stick your iPod back in
[04:15] <daniels> ross: urgh
[04:15] <pitti> ross: no, that won't help, it will get a new device node
[04:16] <sjoerd> hmm why are the pci id's files still open..
[04:17] <ross> go sjoerd
[04:17] <pitti> odd, works for me
[04:18] <sjoerd> could be a ieee1394 bug
[04:19] <pitti> sjoerd: probably; carlos already told me about some module unloading magic
[04:19] <pitti> sjoerd: but this should not hang hal
[04:19] <sjoerd> pitti: imho the kernel should say, dead device, -EINVAL or something
[04:20] <sjoerd> your process should stay in D so long because you open() or something like that
[04:20] <pitti> sjoerd: yes, I saw this on an user with a faulty CD rom drive
[04:22] <sjoerd> pitti: my own cdrom drive is quite nasty during bootup, causing hald to stay in D for some time (untill the kernel shuts off dma for it)..
[04:22] <sjoerd> hehe
[04:23] <pitti> So I guess we have to revoke the eject patch for now
[04:23] <sjoerd> i don't think their is a way to work around this
[04:23] <pitti> carlos: Hi! Back?
[04:23] <pitti> sjoerd: well, open() should have a timeout
[04:23] <carlos> pitti: hi
[04:23] <pitti> sjoerd: or even better, should fail immediately
[04:23] <pitti> carlos: we just discussed that you are lost. Your computer was destroyed completely, go and buy a new one
[04:23] <pitti> carlos: just kidding
[04:23] <carlos> pitti: :-P
[04:24] <thom> jdub: mail away
[04:24] <pitti> carlos: we probably have to revoke the patch because the kernel stinks
[04:24] <carlos> pitti: I was calling to Apple, I'm having problems with my iPod battery
[04:24] <pitti> carlos: it hangs forever on opening a non-existing device
[04:24] <carlos> btw, it works with the pen drive
[04:24] <sjoerd> pitti: you can open with O_NONBLOCK.. 
[04:24] <pitti> Interestingly, this does not crash on USB
[04:24] <pitti> sjoerd: does this also apply to open() itself?
[04:25] <pitti> sjoerd: I thought it only was meaningful to further read() calls
[04:25] <sjoerd> from the manpage: Neither the  open  nor  any subsequent operations on the file descriptor which is returned will cause the calling process to  wait.
[04:25] <carlos> pitti: also, If I boot with the iPod connected, it's not mounted :-(
[04:25] <sjoerd> but i think it's non-trivial to patch hal for that 
[04:26] <pitti> carlos: https://bugzilla.ubuntulinux.org/show_bug.cgi?id=1499
[04:26] <pitti> sjoerd: are there so many places which open a device?
[04:26] <carlos> carlos@frodo ~ $ eject /dev/sda
[04:26] <carlos> pitti: that's the problem
[04:26] <carlos> eject hangs
[04:26] <carlos> with the iPod
[04:26] <pitti> ah, THIS is the hanging problem
[04:26] <pitti> there was a bug about that recently, but it was closed again
[04:27] <carlos> pitti: I'm not using latest kernel
[04:27] <ross> AHA
[04:27] <carlos> need to execute a dist-upgrade 
[04:27] <carlos> so perhaps that's the problem
[04:27] <ross> carlos: that is the bug i had last week
[04:27] <pitti> carlos: if you can reproduce the hanging problem, I would like to patch eject with O_NONBLOCK and let you test it
[04:27] <sjoerd> pitti: on my friends G5 the daily snapshot cd's give the same error:  
[04:27] <sjoerd> christian: https://bugzilla.ubuntu.com/show_bug.cgi?id=2137
[04:27] <carlos> pitti: sure
[04:28] <carlos> pitti: but could we do it later?, give me the URL and I will work on it tonight, I just started to work today and I have lots of things to do...
[04:28] <sjoerd> pitti: more problamatic that hal's logic works synchronous everywhere 
[04:28] <pitti> carlos: sure, I just provide a new package and put it on my webserver
[04:28] <carlos> pitti: perfect
[04:28] <carlos> I will reboot now to "fix" the hang
[04:28] <pitti> carlos, ross: I reopen the eject bug
[04:29] <carlos> pitti: perfect
[04:31] <pitti> carlos,ross,sjoerd: bad news. eject() already uses O_NONBLOCK
[04:32] <sjoerd> pitti: i assume your on the hal@fd.o list 
[04:32] <pitti> sjoerd: yes
[04:32] <sjoerd> k
[04:33] <pitti> sjoerd: just saw your patch
[04:33] <pitti> sjoerd: will it help to fix these hangings?
[04:33] <sjoerd> that's fast i send it about 10 seconds ago
[04:33] <sjoerd> pitti: no, it just plugs an innocent fd leak
[04:34] <tseng> i have an ipod to play with if anyone needs more test
[04:34] <pitti> throw it away
[04:34] <pitti> no, please don't
[04:35] <pitti> throw hal away
[04:35] <tseng> no i use it every day
[04:35] <tseng> its nice.
[04:35] <sjoerd> if you throw it away, throw it in this direction :)
[04:35] <tseng> except for not using OGG
[04:35] <tseng> which is suck.
[04:35] <ross> tseng: join the gang and send a complain to apple
[04:35] <pitti> tseng: isn't there any hacker out there with updated firmware?
[04:35] <tseng> i think when it first came out there were some guys that installed ucLinux on it
[04:36] <tseng> ive never seen any real clean solution
[04:36] <tseng> its very proprietary all around
[04:36] <daniels> the interesting thing with decoding Ogg on iPod is that they only ever got it to ~90% realtime
[04:37] <daniels> when the Vorbis guys claim Tremor should be able to do 100%, even in pure integer mode
[04:37] <tseng> so whats that make for, a little pause between tracks?
[04:38] <daniels> assuming you pre-buffer the entire thing to PCM
[04:38] <pitti> sjoerd: I cannot find any other open() option for us. Once you do the open(), the process hangs forever without any signal being able to recover it
[04:38] <daniels> you'd need a 10% track length pause just between the tracks, assuming you could manage to decode the entire thing to PCM and store it
[04:38] <daniels> not to mention spinning up the hard drive, access, streaming the file, et al
[04:38] <tseng> =/
[04:39] <daniels> (decoding to PCM being non-trivial -- you'd need something like ~60MB of solid state for most tracks)
[04:39] <sjoerd> pitti: that's not easily fixed then...
[04:40] <pitti> sjoerd: no, it's a kernel bug.
[04:41] <sjoerd> pitti: exactly 
[04:41] <sjoerd> pitti: although it would be nice if hald worked somewhat less async
[04:42] <pitti> sjoerd: indeed.
[04:42] <sjoerd> uhm somewhat less sync or somewhat more async :) 
[04:42] <pitti> sjoerd: but regardless of how it works, if a single open() on the wrong device kills it, you are lost anyway
[04:42] <sjoerd> pitti: yes that's what i said earlier :)
[04:44] <Kamion> so; any objections to me changing the default Hebrew locale to he_IL.UTF-8? it works for me.
[04:45] <Kamion> I think the same might be needed for Arabic (ar_EG.UTF-8) too; checking
[04:47] <seb128> jdub, mdz: ok to upload eagle-usb instead of eagle-adsl ? We already have the kernel module from eagle-usb, we should have the user tools in sync, I've forgotten to do this before ...
[04:53] <zepo__> hi everybody...someone has installed Ubuntu on Acer travelmate laptop,coz i've problem
[04:54] <zepo__> or tell me something about, tnx
[04:55] <fabbione> zepo__: these questions should go to #ubuntu. please see /topic
[04:56] <zepo__> ah ok... i've seen in the ubuntu homepage : http://www.ubuntulinux.org/community/teams/laptop/view?searchterm=laptop
[04:56] <zepo__> and nobody can help me on #ubuntu
[04:58] <zepo__> am i wrong?
[04:58] <thom> zepo__: check http://wiki.ubuntulinux.org/HardwareSupport
[04:58] <zepo__> tnx
[04:59] <carlos> pitti: could we add to #2134 a dependency to #1891 ?
[04:59] <pitti> carlos: good idea
[04:59] <carlos> pitti: thanks
[05:00] <tseng> mono fans: considering tweaking muine package to use gstreamer rather than xine.. thoughts?
[05:04] <carlos> tseng: gstreamer for sound works really good
[05:05] <tseng> yep.
[05:05] <tseng> the debian sid pkg defaults to xine atm
[05:05] <Kamion> wow, something is really screwed with Arabic
[05:05] <Kamion> (d-i)
[05:06] <Kamion> the translations for partman are sufficiently broken to make the mount point selector fall over and refuse to allow anything but /
[05:06] <daniels> Kamion: ... impressive
[05:06] <Kamion> I don't believe I've ever seen that before ...
[05:06] <Kamion> daniels: hey, you look like a volunteer ... :)
[05:08] <Kamion> there's also something weird with the countrychooser shortlist generation for Arabic, not sure about that yet
[05:09] <daniels> Kamion: how much arseclownery is required to build a netinst?
[05:09] <Kamion> for warty?
[05:09] <Kamion> theoretically not too much, debian-cd has support, you set INSTALLER_CD=2 or some such; but I've never actually tried it for warty
[05:10] <Kamion> partly 'cos Mark wanted the Only One CD approach
[05:10] <Kamion> I was vaguely thinking about chucking out a totally unofficial netinst for hoary
[05:10] <fabbione> Kamion: netinst ?
[05:10] <fabbione> we have netinst, don't we?
[05:10] <Kamion> daniels: the debian-cd tarball I'm using is at http://cdimage.ubuntulinux.org/code/
[05:10] <fabbione> or are we talking about 2 different netinst?
[05:10] <Kamion> fabbione: no, netinst = CD with installer and base system only
[05:11] <Kamion> you're thinking of netboot
[05:11] <Kamion> netinst is badly misnamed really
[05:11] <daniels> oh, netboot, sorry
[05:11] <daniels> something I can install my X40 with :)
[05:11] <fabbione> Kamion: ok :-) isn't that the miniiso in netinstall dir on the mirror?
[05:11] <Kamion> daniels: aha! we have netboot, look in install/netboot/ on CD images
[05:11] <fabbione> daniels: we have netboot punk
[05:11] <Kamion> or /dists/warty/main/daily-installer-i386/netboot/ I think
[05:12] <daniels> Kamion: right ... so no buggery required with debian-cd?
[05:12] <Kamion> nope
[05:12] <Kamion> businesscard would actually be more useful to produce than netinst; that's a CD with just all the udebs on it, fetches base system etc. from network
[05:13] <daniels> Kamion: cool
[05:13] <fabbione> Kamion: we can expand the mini.iso
[05:13] <Kamion> but we have the netboot mini.iso for most of that kind of use case I guess
[05:13] <Kamion> fabbione: nooooooooo
[05:13] <daniels> Kamion: hm
[05:13] <Kamion> fabbione: upstream totally vetoed that :)
[05:13] <daniels> Kamion: would it be possible to just netboot the standard cd? :)
[05:14] <Kamion> daniels: 'fraid not, the CD doesn't actually have enough udebs to support netboot; there's a bug about it
[05:15] <daniels> Kamion: ahr, heh
[05:15] <daniels> Kamion: ok, I'll try mini.iso, and construct a mirror here
[05:15] <daniels> Kamion: any tips on constructing a mirror?
[05:15] <daniels> Kamion: (well, basically gotchas that I'd need to look out for)
[05:18] <Kamion> I basically just do debmirror --section=main/debian-installer
[05:18] <daniels> Kamion: rad
[05:18] <Kamion> make sure that gives you a Release file
[05:18] <Kamion> I wish I knew why my powerbook's screen occasionally goes into flickery nightmare mode
[05:19] <daniels> hm
[05:19] <daniels> Kamion: i was kind of hoping to do it from local sources
[05:19] <Kamion> it's like it's shuddering a bit under a fifth of a screen from the right
[05:19] <Kamion> s/from/to/
[05:19] <daniels> Kamion: is apt-move with a bit of poking sufficient?
[05:20] <zepo__> may i suggest to consider, if it's possible,clearly,also Acer laptop
[05:20] <Kamion> daniels: you could copy all the udebs from the CD to an archive structure and debmirror the rest
[05:20] <daniels> zepo__: basically, try it and see if it works; the reason it won't be listed is because no-one has it
[05:20] <daniels> zepo__: even a failure report is better than unknown, because we then have a far better idea of what to fix
[05:20] <Kamion> there won't be all that much beyond what's on the CD
[05:21] <daniels> Kamion: cheers
[05:25] <zepo__> i tried,i have on my acer
[05:25] <zepo__> daniels: sorry, and i 've video problems
[05:26] <zepo__> daniels:if i will resolve i'll post on ubuntu page,but i don't know if i'll do
[05:27] <thom> why is someone using ndiswrapper for an ipw2200?
[05:29] <thom> Kamion: 2137 looks strangely familiar
[05:29] <thom> :-)
[05:35] <Kamion> thom: indeed :)
[05:35] <Kamion> guess why it didn't take me long to work out what the problem was ...
[05:35] <Kamion> (likely to be)
[05:36] <daniels> ok, seems apt-move has mostly done the trick
[05:37] <Kamion>    * Load firmware from standard locations (me):
[05:37] <Kamion>      . drivers/net/wireless/acx/acx100_helper.c
[05:37] <Kamion>      . drivers/net/wireless/acx/acx100_usb.c
[05:37] <Kamion> ROCK!
[05:37] <daniels> Kamion: btw, $MIRROR in update-cd is still n-n-y
[05:37] <tseng> yay for dh_netlibs !!!
[05:37] <tseng> + cdbs
[05:38] <Kamion> daniels: right, we haven't changed the directory in /srv yet
[05:38] <Kamion> thom: hm, do you have a minute to coordinate moving that to /srv/cdimage.ubuntu.com, or have you not bothered yet for other services?
[05:40] <daniels> Kamion: fair point
[05:41] <daniels> Kamion: hm
[05:41] <daniels> Kamion: do I get to set a local mirror to download stuff from, or will it have a bash at archive.ubuntu.com?
[05:41] <daniels> Kamion: (that's a d-i rebuild if no, correct?)
[05:42] <thom> Kamion: sorry, move what?
[05:42] <Kamion> thom: little:/srv/cdimage.no-name-yet.com
[05:42] <Kamion> daniels: you'll get to set a local mirror when using netboot
[05:43] <thom> Kamion: most haven't bothered yet
[05:44] <Kamion> fair enough, no hurry then
[05:45] <ross> thom: amusing bug i've just found in IIS
[05:45] <ross> thom: which being an apache guy you'd find amusing
[05:46] <tseng> jdub: would you mind double checking tomboy sometime? I fixed it
[05:46] <ross> thom: domain names with the final period don't get matched in the virtual hosts.. www.180sw.com/ works fine, www.180sw.com./ fails
[05:47] <thom> yeah, seen that one
[05:47] <thom> it's... irritatiing
[05:48] <ross> fools. our company signature has the dot at the end...
[05:48] <daniels> Kamion: ahr, crap
[05:48] <daniels> Kamion: any way to build it without source?
[05:48] <thom> ross: it does? ugh
[05:48] <daniels> Kamion: (my /var/cache/apt/archives doesn't extend to sources, sadly)
[05:49] <thom> and why on earth are you using IIS?
[05:49] <thom> hosted by bluewank?
[05:49] <ross> thom: business, so telewest thankyou very much ;)
[05:49] <daniels> Kamion: forgive the dumb questions
[05:49] <thom> ah
[05:51] <Kamion> daniels: --nosource, if you mean debmirror?
[05:52] <Kamion> daniels: no problem
[05:52] <daniels> Kamion: debian-cd
[05:52] <daniels> Kamion: i've apt-moved /var/cache/apt/archives into a pool structure
[05:53] <Kamion> uh; do you still need to use debian-cd?
[05:54] <Kamion> I thought you were going to use netboot now
[05:54] <daniels> Kamion: oh
[05:54] <daniels> how do I generate netboot, if not debian-cd?
[05:54] <Kamion> I think we're talking at cross-purposes :)
[05:54] <Kamion> you download it from warty?
[05:54] <daniels> oh, man
[05:54] <daniels> yeah
[05:54] <daniels> imagine I'm sitting here slapping my forehead.
[05:54] <Kamion> the reason you need the debmirrored archive is because netboot will need to download from it
[05:54] <daniels> yeah :0
[05:54] <daniels> :), even
[05:56] <daniels> it was an entertaining sojurn into debian-cd, anyhoo
[05:57] <Kamion> heh :)
[05:57] <Kamion> an ugly piece of code, that
[05:57] <daniels> yeah
[05:58] <seb128> grrrr
[05:59] <seb128> I just spent 1 hour to understand that wanadoo rejects all the mail with a subject starting by "new ..."
[05:59] <daniels> seb128: wtf?
[05:59] <seb128> I was trying to send a mail about the new totem/gst packages
[05:59] <seb128> daniels: apparently an anti-spam system
[06:05] <carlos> seb128: man, you need a real mail account...
[06:05] <seb128> I need a real provider :p
[06:08] <thom> seb128: rock on
[06:11] <daniels> hm, the biggest challenge might actually be finding a hub ;)
[06:13] <tseng> hmm, pkgconfig is uninstallable?
[06:13] <tseng> or am I smoking crack.
[06:17] <thom> tseng: works fine for me
[06:17] <tseng> thom: good deal
[06:17] <daniels> i'm with thom
[06:18] <tseng> says no installation candidate
[06:18] <daniels> try putting a warty repo in sources.list :P
[06:18] <tseng>  deb http://archive.ubuntu.com/ubuntu warty main restricted universe
[06:18] <tseng> smartarse
[06:19] <thom> apt-cache policy pkg-config ?
[06:19] <tseng> oh.. sorry
[06:19] <tseng> i have depend pkgconfig =/
[06:20] <thom> that'd be why then :-)
[06:20] <tseng> ya, i r dumb
[06:20] <tseng> hacking on the monodevelop pkg some more
[06:21] <thom> cool
[06:22] <tseng> i broke some stuff when i converted to cdbs
[06:22] <tseng> the sid package was debhelper
[06:24] <thom> horrible thing :-)
[06:24] <seb128> bah
[06:25] <daniels> thom: in my case, cdbs wasn't doing me wrong; it was autotools the whole time
[06:25] <mdz> pitti: agreed, 2131 is not RC
[06:25] <seb128> jdub, mdz: ok to upload eagle-usb instead of eagle-adsl ? We already have the kernel module from eagle-usb, we should have the user tools in sync, I've forgotten to do this before ...
[06:25] <pitti> Hi mdz!
[06:25] <mdz> sabdfl: already a bug filed about that question
[06:26] <mdz> daniels: dbus-monitor is not critical functionality
[06:26] <daniels> mdz: 'kay
[06:26] <mdz> seb128: what's the difference between eagle-usb and eagle-adsl?
[06:26] <daniels> mdz: so absolutely no changes, even if dbus-monitor is totally dead to the world?
[06:26] <pitti> mdz: okay, but thom was faster with downgrading it :-)
[06:26] <mdz> daniels: give me a bug #
[06:27] <seb128> mdz: the package has been renamed. We have the module from eagle-usb, the old -adsl doesn't work with 2.6
[06:27] <daniels> mdz: none filed
[06:27] <seb128> dunno what has changed exactly in the user tools part
[06:27] <daniels> bonza! ubuntu me up, baby
[06:29] <mdz> 13018: arguments to dbus_connection_send_with_reply_and_block() were incorrect, assertion "(error) == NULL || !dbus_error_is_set ((error))" failed in file dbus-connection.c line 1999.
[06:29] <mdz> This is normally a bug in some application using the D-BUS library.
[06:29] <mdz> Failed to set up match "dbus-monitor": 
[06:29] <mdz> free(): invalid pointer 0x40257edc!
[06:29] <mdz> daniels: that?
[06:29] <daniels> mdz: ya-huh
[06:29] <mdz> seb128: so we need to replace -adsl with -usb?
[06:29] <seb128> mdz: yes
[06:30] <daniels> Kamion: where does 'anna' come from?
[06:30] <Kamion> the name?
[06:30] <daniels> Kamion: 'user.warn anna[2451] : WARNING **: bad d-i Packages file'
[06:30] <Kamion> no, Anna Hess, Joey's sister
[06:30] <daniels> Kamion: grep turns up nothing in d-i
[06:31] <Kamion> daniels: there's a source package named thus
[06:31] <daniels> hrm
[06:31] <daniels> arh, maybe if I learnt to type
[06:33] <mdz> daniels: if the breakage is in dbus-monitor itself, and fixing it doesn't risk breaking dbus itself, I don't mind if you want to fix it
[06:33] <daniels> mdz: thanks
[06:38] <daniels> ARGH
[06:43] <daniels> Kamion: hum
[06:59] <tseng> anyone using the new totem pkg?
[06:59] <sivang> just upgraded. dpkg asks about hal config file, replace?
[06:59] <daniels> Kamion: 
[06:59] <tseng> if i start totem while muine was using the dsp, it crashes
[06:59] <daniels> daniels@nanasawa:~/mirror/warty-local/dists/warty/main/debian-installer/binary-i386% for i in $(zgrep '^Filename:' Packages.gz | sed -e 's/^Filename: //;'); do (cd ~/mirror/warty-local/$(dirname $i) && wget http://archive.ubuntu.com/ubuntu/$i); done
[07:00] <Kamion> hoorah
[07:00] <Kamion> looks reasonable to me, if gross :)
[07:00] <Kamion> you might want a mkdir -p in there
[07:00] <daniels> already done a mkdir -p loop
[07:00] <Kamion> ok
[07:00] <daniels> it's happily downloading all the udebs now
[07:00] <Kamion> debmirror for dummies. :)
[07:01] <daniels> it was the easiest way to not do a debmirror ;)
[07:01] <daniels> heh
[07:01] <daniels> the problem was that I moved my local cache over with apt-move (which has every package I need -- I checked)
[07:01] <daniels> so I ran debmirror to just get the main/d-i section, and it decided to clobber the Release file and delete all of my debs
[07:02] <daniels> so I wget'ed the Packages.{gz,bz2} from auckland, hacked the Release file by hand, and now I'm wget'ing the udebs
[07:03] <daniels> ghetto mirror 101
[07:09] <thom> speaking of ghetto, it appears i'm moving
[07:22] <daniels> thom: oh?
[07:25] <daniels> thom: where to?
[07:29] <daniels> lamont_r: hey dude
[07:33] <daniels> Kamion: (turns out linux-image hadn't been updated -- go figure)
[07:34] <daniels> Kamion: bloody good job on d-i btw; it should, by rights, be absolutely refusing to install (hacked-up setup, plus about twenty failures and other crap)
[07:38] <Kamion> d-i's pretty good at coping with weird shit :)
[07:39] <daniels> (mercifully)
[07:45] <daniels> second stage ...
[07:59] <lamont_r> any rc bugs for me to help with?
[08:36] <amu> lamont_r: firefox ? ;)  
[08:40] <lamont_r> I thought they decided that firefox was a heisenbug
[08:41] <sivang> lamont_r : how did it go with postifx eventually?
[08:42] <sivang> ah, just noticed debian branding issue in emacs
[08:43] <sivang> "modified by debian project"
[08:43] <azeem> what's wrong with that?
[08:44] <sivang> ah nothing
[08:44] <tseng> it was modified by debian project, no?
[08:44] <sivang>  I just noticed it was removed from openoffice,
[08:45] <sivang> which was also modified by the debian proejct..
[08:45] <sivang> so figured this should be consistent.
[08:48] <lamont_r> sivang: considering upload now, although I'd like _one_ more set of eyes first...
[08:53] <sivang> lamont_r : you've fixed up the postfix-doc oopsy ?
[08:53] <amu> lamont_r: hehe ( heissenbug ) 
[08:53] <lamont_r> sivang: the .gz issue?
[08:54] <sivang> lamont_r : yes. that missing file.
[08:54] <lamont_r> fixed that by pointing to the html version, which is never compressed.
[08:54] <sivang> lamont_r : page me I will give it another run also
[08:54] <amu> i found something nice to fix, ex. if i install german env, i got a english openoffice, english *spellcheckers ... should it german instead of english ;) ? 
[08:54] <lamont_r> see bug 2022
[08:55] <lamont_r> amu: and lots of "misspelled" words, eh?
[08:58] <amu> lamont_r: thats no prob and impossible to fix i18n from gnome, but as a new user, i expect choosing german or any other lang, my programs and depends setuped in german, just a little more useability
[09:01] <mdz> Kamion: around?
[09:05] <Kamion> mdz: yep
[09:07] <lamont_r> amu: yeah
[09:10] <fabbione> hey amu
[09:11] <amu> fabbione: buona sera ;) 
[09:11] <fabbione> amu: heheh
[09:13] <amu> fabbione: what up, you GF is again away ? *ducked* 
[09:13] <lamont_r> fabbione: gmt+3 there?
[09:14] <Kamion> mdz: will be gone for dinner in about 15 minutes
[09:15] <fabbione> amu: ehehehe
[09:15] <fabbione> lamont_r: gmt+2
[09:15] <fabbione> (we are still summer time)
[09:15] <lamont_r> right.  so london + 1
[09:16] <fabbione> amu: she is sleeping.. all this changing house stuff is killing us a bit
[09:16] <fabbione> lamont_r: yeps
[09:16] <amu> hehe, summertime, but 8C here 
[09:17] <fabbione> amu: yeah... nothing new about it :-)))
[09:17] <myk> is there documentation on the correct way to rebuild debian packages for ubuntu?
[09:18] <Kamion> myk: the only things that build our packages are our build daemons ...
[09:18] <mdz> Kamion: have a moment to review ubuntu-meta for me?
[09:18] <Kamion> mdz: sure
[09:18] <Kamion> what is it? :)
[09:18] <mdz> mailed
[09:18] <mdz> the thing which generates ubuntu-base and ubuntu-desktop
[09:19] <myk> Kamion, i'd like to use robert love's netapplet, and there's already a debian package for it.  can i just install it?
[09:19] <Kamion> ah
[09:19] <Kamion> myk: should be able to yes
[09:19] <Kamion> myk: failing that just build the package in the normal Debian way on an Ubuntu system
[09:19] <myk> mkay
[09:20] <lamont_r> myk: for rebuilding, the trivial answer is: have current build-depends installed and say dpkg-buildpackage -rfakeroot -b
[09:20] <myk> lamont_r: thanks
[09:21] <lamont_r> I guess there's and apt-get version that does that for you, but I can't ever remember it... :-)
[09:22] <myk> i'm new to the debian system... still thinking in gentoo
[09:24] <lamont_r> myk: the build daemons actually us a chroot to build things, and 'sbuild' to deal with getting in and out of the chroot properly.
[09:25] <lamont_r> but that's overkill for most things.
[09:26] <myk> is there a way for users to submit rebuilt packages to say, universe?
[09:27] <lamont_r> myk: for that, generate a diff vs the ubunut source that fixes the issue, and file a bug with that.
[09:27] <myk> lamont_r: well, this is a debian package I just downloaded that isn't yet available in ubuntu
[09:27] <lamont_r> as time permits, we'll review it and upload
[09:28] <lamont_r> then the buildds build it in the pristine environment
[09:30] <thom> mdz: i guess you and jdub need to work out a sane strategy for firefox, but the current release isn't sustainable
[09:30] <lamont_r> thom: ouch
[09:35] <mdz> thom: it does seem a bit on the shitty side
[09:35] <mdz>    * Add ipw2100 and ipw2200 to nic-extra-modules.
[09:35] <tseng> hmm
[09:35] <Kamion> thought you'd like that
[09:35] <mdz> Kamion: that includes the firmware as well?
[09:35] <Kamion> dunno why I'd forgotten it
[09:36] <Kamion> no, not the firmware, that involves more work
[09:36] <Kamion> I do want to do that before warty though, if I can
[09:36] <mdz> I thought you were intentionally leaving it out because the firmware wouldn't work yet
[09:36] <Kamion> might as well have it in there, then at least even if I don't get the firmware in people can copy it over on a USB stick or a floppy
[09:36] <lamont_r> does this mean that we have those stupid d-link cards working now?
[09:36] <Kamion> that would suck, but only gently
[09:37] <Kamion> the d-link cards need ndiswrapper, don't they?
[09:37] <Kamion> I haven't included that, that seems like lots more pain
[09:37] <elmo_> the word your look for is WORLDDESTROYINGCRACK
[09:38] <lamont_r> no issue here.
[09:38] <lamont_r> ndiswarpper evil.
[09:38] <lamont_r> and warping, apparently
[09:41] <lamont_r> fabbione: daniels: how do I tell X to switch bits-per-pixel without restarting X?
[09:41] <sivang> lamont_r : ok, i see it's accepted. have a look at it? or when it's accepted that means it's uploaded already?
[09:41] <lamont_r> accepted means that the source is in the archive.  given thebuild time, the binaries will be available in just over 23 minutes
[09:42] <sivang> lamont_r : ok
[09:43] <lamont_r> that reminds me - time to close the bug.
[09:43] <fabbione> lamont_r: you mean depth?
[09:43] <lamont_r> yeah
[09:43] <fabbione> lamont_r: you can't.. 
[09:43] <fabbione> you need to restart X
[09:43] <lamont_r> feh
[09:44] <lamont_r> ok.  something to do after the upgrade finishes,then.
[09:44] <lamont_r> xresprobe doesn't give depth, edid fails, and the card claims that it does 24-bit.
[09:44] <lamont_r> for whatever reason, 16 bit was chosen back in the dawn of time for this system
[09:45] <lamont_r> fabbione: there was a keystroke sequence for changing resolution though, yes?
[09:46] <thom> ctrl+shift++
[09:47] <thom> and ctrl+shift+-
[09:47] <fabbione> yes
[09:47] <lamont_r> hrm.. metacity seems to keep those grabbed, or the LCD doesn't want to change, or something.. :-)
[09:51] <thom> mdz: if we go back to 0.9.3 we'll need to sync the firefox language packs from unstable
[09:51] <mdz> thom: that sounds relatively painless
[09:51] <mdz> thom: certainly compared to the profile nightmares that would likely ensue
[09:51] <thom> yeah
[09:52] <thom> i'm sure downgrading profiles is even less supported than upgrading them
[09:52] <mdz> echo "rm -rf /home/*/.mozilla/firefox" >> mozilla-firefox.postinst
[09:52] <thom> the firefox wrapper script does something similar, actually. it may or may not be enough
[09:59] <mdz> Kamion: any feedback on ubuntu-meta?
[09:59] <lamont_r> gnumeric.desktop contains invalid MIME type 'comma-separated-values'
[09:59] <lamont_r> (missing slash)
[10:00] <mdz> lamont_r: how is multiverse looking in terms of builds?
[10:01] <lamont_r> it was chunking along - let me go look
[10:01] <elmo_> 1.1G    ftp/pool/multiverse
[10:01] <elmo_> quinn-warty.amd64-old:149
[10:01] <elmo_> quinn-warty.i386-old:136
[10:01] <elmo_> quinn-warty.powerpc-old:139
[10:01] <elmo_> ^-- multiverse entries in q-d
[10:02] <lamont_r> 209 total multiverse packages (i386), 75 Installed.
[10:03] <elmo_> a lot of the marillat stuff probably won't build - some of it wasn't even signed (woot)
[10:03] <lamont_r> 76 dep-wait, 4 needs-build, 53 building (==failed)
[10:04] <elmo_> heh half the world dep-waiting on java
[10:05] <lamont_r> not surprising. :-(
[10:05] <thom> since world+dog have now asked
[10:22] <lamont_r> hrm.. telling xchat to open a url in a new tab of mozilla was working with firefox before the upgrade..
[10:24] <Mithrandir> lamont_r: http://freerelay.raw.no/setup-postfix ?
[10:26] <lamont_r> um, that just says how to use freerelay.raw.no as a relay...
[10:26] <lamont_r> although I must wonder why you have relayhost twice...
[10:26] <lamont_r> also, I think you really want [freerelay.raw.no] :2525
[10:26] <lamont_r> without the []  it does MX lookups
[10:27] <lamont_r> time to go get kids from school.  bbiab
[10:49] <sivang> I have a user wanting to make a custome kernel, is this applicable to ubuntu already? (e.g, k-src pageas, build-deb it etc)
[10:49] <sivang> (just like it goes for debian?)
[10:54] <mdz> mako: ping?
[10:56] <mdz> sivang: it is almost the same; the packages are named linux-* rather than kernel-*
[10:57] <sivang> mdz : yes, someone else noted it on #ubuntu, I checked so it's cool. afterwards it's make xconfig/menuconfig and dpkg-builddeb..;)
[10:58] <mdz> sivang: generally, when someone is new to the system and is asking to compile the kernel, the question is "why?" :-)
[10:58] <sivang> hmmm
[10:58] <sivang> :)
[10:58] <sivang> didn't even think of asking him
[10:58] <sivang> but he said himself,
[10:58] <mdz> so many people are accustomed to compiling the kernel just because they've always done it
[10:58] <sivang> "I think I need for booting my USB drive"
[10:59] <sivang> his kenrel panics on /dev/console missing || cannot open
[10:59] <mdz> same guy from the mailing list?
[10:59] <mdz> if so, I think he's in over his head
[10:59] <sivang> he told me he used to do this on gentoo (compile the kernel with uid support built in)
[10:59] <sivang> and it worked for him.
[10:59] <sivang> sec, I'll search
[11:00] <sivang> oh god..He's gone to IRSSI, I am afraid something there is going to break..shall I shout him to stop? :)
[11:03] <mako> mdz: hey there
[11:04] <mdz> mako: hey, I noticed that LWN didn't pick up this week's traffic, do we need to synchronize with them or something in order for that to happen?
[11:04] <mako> mdz: whats up?
[11:04] <sivang> hey mako
[11:04] <mako> sivang: hey
[11:04] <mako> mdz: i didn't realize they had picked up other traffics
[11:05] <mdz> mako: http://lwn.net/Articles/103368/
[11:05] <mdz> mako: last week picked up traffic #5
[11:05] <mdz> which is awesome
[11:05] <mdz> it would be even better if they did it every week from now on :-)
[11:06] <mdz> so 0923 had a headline article about ubuntu, 0930 had traffic #5, but this week we're a zero :-(
[11:07] <mako> mdz: excellent.. do we have a contact there or should i just email lwn@lwn.net?
[11:07] <mdz> mako: I think lwn@lwn.net is the place
[11:07] <mako> i'm happy to just BCC traffic to them every week then
[11:07] <mdz> mako: I think we probably want to set up an ubuntu-news, a la debian-news
[11:07] <mako> mdz: i was thinking of that
[11:07] <mdz> I'm pretty sure lwn@lwn.net subscribes, and that's how they pick up DWN
[11:07] <mdz> mako: mail jdub on it?
[11:07] <mako> mdz: sure, i can do that
[11:08] <mako> mdz: i'm going to try to hook up with kernel traffic i think too 
[11:08] <mako> mdz: so we're in their header
[11:08] <mdz> mako: once it's set up, we should announce it to -announce for those folks who are ready for more traffic than -announce, but less than -users :-)
[11:08] <mako> mdz: absolutely
[11:10] <mako> mdz: can you verify that you can you regen your pw in shipit and such without getting internal server errors for me?
[11:11] <mdz> mako: yep, works
[11:11] <mdz> at least, doesn't throw an error
[11:11] <mako> brilliant
[11:11] <mako> ok
[11:13] <mako> kiko pointed out you can order negative cds :)
[11:13] <mako> if you do this, i will flag it when shipping and send someone to your house to steal your cds
[11:14] <mdz> good plan
[11:18] <m_tthew> mako: worked for me 100%
[11:18] <m_tthew> mako: the new shipit stuff
[11:18] <Mithrandir> mdz: would you be unhappy if I sat down and implemented good amd64 nvidia support over the weekend, wrt stability?
[11:18] <mdz> Mithrandir: no, I would be quite pleased in fact
[11:18] <mako> m_tthew: awesome :)
[11:18] <mdz> I started on it, but it required more time than I could devote
[11:18] <mdz> Mithrandir: if you want, I'll send you what I have
[11:19] <Mithrandir> mdz: the student society has some amd64 cards, so I'm going to borrow it.
[11:19] <Mithrandir> mdz: yes, please.
[11:19] <mdz> argh, I can't
[11:19] <mdz> apparently I deleted it
[11:19] <mdz> it was a mess anyway; I had pretty much decided I should have rewritten debian/rules instead
[11:19] <Mithrandir> ok
[11:19] <Mithrandir> it uses the nvidia-kernel-source thingy, or how is all this structured?
[11:20] <mdz> Mithrandir: as long as you can verify that the existing i386 package is exactly the same after your modifications, I'm OK with it
[11:20] <Mithrandir> I haven't looked at it at all
[11:20] <mdz> Mithrandir: it's all part of linux-restricted-modules
[11:20] <Mithrandir> including the X module bit?
[11:20] <mdz> yes
[11:20] <Mithrandir> ok, coolie
[11:20] <Mithrandir> verification should be easy.
[11:20] <mdz> it builds sboth the modules and nvidia-glx
[11:33] <thom_> daniels: ping?
[11:41] <sabdfl> erm... i can't believe i even found the words "ndiswrapper" and "installer" in the same SENTENCE in scrollback tonight :-)
[11:43] <sabdfl> kamion: did I read it right, we have fw in the installer? awesome, great work indeed.
[11:45] <thom_> using bleeding unreleased features in dbus
[11:45] <sabdfl> thom: careful dude, russian joke about condoms calls them "red hats", and this is a very multicultrual channel
[11:46] <mdz> sabdfl: we don't yet have firmware in the installer, but Kamion has done much of the work and we may have it for Warty
[11:46] <sabdfl> fantastic, that's a big one for me, worth leaving one or two other warts in for if it comes down to a choice
[11:47] <thom_> sabdfl: thanks, i'm now disturbed :-)