[12:08] <Kamion> ogra: debian-installer failed to build, but that should be fixed now. rebuild your CDs.
[12:08] <nixternal> actually, NO to the NVidia drivers
[12:08] <ogra> oki
[12:08] <nixternal> thought they were in there
[12:08] <ogra> Kamion, thanks, sleep well
[12:09] <Nafallo> hehe, Kamion woke up only to tell ogra that :-)
[12:10] <ogra> Nafallo, he knows i'm short on time ... i'll be travelling by the time knot3 comes out and have to do the testing before ... ;)
[12:11] <Nafallo> aha :-)
[12:43] <Burgwork> http://blogs.adobe.com/penguin.swf/ <-- ubuntu!
[12:45] <jdong> Burgwork: yeah, I noticed... doesn't that make us feel good? :)
[01:40] <zul> Burgwork: I have updated my blog fyi
[02:36] <bddebian> Hwody
[02:36] <bddebian> Err Howdy even
[03:13] <jdub> mjg59: pcscd with a stupid plist (apple!) setting change in libccid works
[03:13] <jdub> though now i need a functional smart card
[03:24] <wasabi> I've been eatten alive by the XFS bug in 2.6.17 apparently. Fixed in 2.6.17.7.   The 2.6.17-7 kernel, is that this? Or is the - the ubuntu version?
[03:25] <bluefoxicy> o.o
[03:26] <bluefoxicy> doko?
[03:26] <wasabi> ALso somebody needs to upload xfsprogs 2.8.10 because it's related in this bug.
[03:28] <mjg59> wasabi: The Ubuntu kernel is 2.6.17.11
[03:28] <wasabi> mjg59, Where is that number shown?
[03:28] <bluefoxicy> hi Keybuk
[03:28] <mjg59> wasabi: The Makefile. Also the changelog.
[03:28] <wasabi> Heh.
[03:28] <Keybuk> heyhi
[03:28] <mjg59> Keybuk: Query about functional module dependencies
[03:29] <mjg59> Keybuk: tifm_711x binds to the PCI ID. tifm_sd provides the SD reader functionality and is required for the PCI device to be functional. This is trivially expressed in modprobe.d, but what package is the right place to put that?
[03:30] <Keybuk> shouldn't it either be a dep
[03:30] <Keybuk> or itself declare those aliases?
[03:30] <Keybuk> if not, module-init-tools can ship a modprobe.d file
[03:31] <mjg59> Is there a reasonable way to express inter-module dependencies outside that?
[03:31] <mjg59> tifm_sd relies on a ti flashmedia device
[03:31] <mjg59> The only current driver we have is for a PCI device, but that's not inherently true
[03:32] <Keybuk> in the modules itself?
[03:32] <mjg59> tifm_sd depends on tifm_711x by virtue of it currently being the only module that provides the symbols it needs
[03:32] <mjg59> But tifm_711x has no logical dependency on tifm_sd
[03:33] <Keybuk> why doesn't it?
[03:34] <mjg59> Why should it?
[03:34] <mjg59> The part doesn't necessarily have an SD reader
[03:35] <Keybuk> what's a 711x then?
[03:35] <mjg59> So from a driver point of view, it's sensible
[03:35] <mjg59> A chip that can read various different flash formats
[03:35] <Keybuk> what's tifm_sd?
[03:35] <mjg59> A driver for the SD portion of that class of chip
[03:36] <mjg59> I've got a Sony which has the same chip, but only memory stick
[03:36] <mjg59> It doesn't hurt to always load tifm_sd, but there's no reason for the driver to depend on it
[03:36] <Keybuk> so surely if you have a 711x device, you always need sd loaded?
[03:36] <mjg59> No
[03:37] <Keybuk> meh, this smells of the "it's not a dependency, it's a recommends" type of bullshit :p
[03:37] <mjg59> On the Sony, I'd want tifm_711x and tifm_ms (which isn't fully written yet, but still)
[03:37] <Keybuk> ahh
[03:37] <Keybuk> but you'd get sd anyway
[03:37] <mjg59> Right
[03:38] <mjg59> The driver author has no reason to introduce a dependency
[03:38] <Keybuk> we can add it in userspace easily enough
[03:38] <mjg59> Yeah
[03:39] <Keybuk> install tifm_711x modprobe --ignore-install tifm_711x $CMDLINE_OPTS && { modprobe -Qb tifm_sd ; : ; }
[03:39] <Keybuk> in a modprobe.d file
[03:39] <mjg59> Yeah
[03:39] <Keybuk> reminds me, TODO; bridge jcm to add "depends" syntax
[03:46] <Keybuk> the location of UDSMV interests me
[03:46] <Keybuk> I wonder which hotel the staff will be in, I don't recall one being anywhere near the Googleplex
[03:46] <Keybuk> will we need to hire cars?
[03:47] <mjg59> Yeah, I was curious about that
[03:47] <zul> i thought the public transit would be pretty good in mountain view
[03:48] <wasabi> UDSMV?
[03:49] <mjg59> Ubuntu Dev Summit Mountain View
[03:49] <wasabi> oh that's what keybuk mentioned
[03:49] <mjg59> zul: It is, but the Googleplex is pretty much on the edge of nowhere
[03:49] <zul> ah...gotcha
[03:51] <wasabi> Heh. I need a common to reinstall every package on my system.
[03:51] <wasabi> command.
[03:51] <Burgundavia> mjg59: welcome to the US. They build for cars there
[03:52] <Keybuk> wasabi: aptitude reinstall '~i'
[03:53] <wasabi> Nice.
[03:54] <wasabi> here I am trying to parse the output of dpkg -l
[03:55] <Burgundavia> zul: seems like Xen is a major pain to keep updated
[03:58] <zul> Burgundavia: yeah once it smooth out it will be ok
[04:18] <SEJeff> Keybuk: ping
[04:20] <Keybuk> SEJeff: hi
[04:22] <SEJeff> Keybuk: Evening. I would like to implement the RapidReboot spec for edgy+1 and know that it will need to play nicely with upstart seeing as how thats what ubuntu is moving towards.
[04:23] <Keybuk> RapidReboot spec?
[04:23] <Keybuk> let me bring that one up
[04:25] <SEJeff> Keybuk: https://wiki.ubuntu.com/RapidReboot
[04:25] <Keybuk> yeah, was waiting for the wiki to kick into gear
[04:26] <wasabi> Interesting.
[04:27] <Keybuk> SEJeff: sweet.  Not sure that it'd need to play much with upstart, tbh
[04:27] <SEJeff> Keybuk: It is basicly implementing kexec reboots for things like kernel upgrades, etc
[04:27] <SEJeff> Keybuk: And it would go perfectly with the idea of the Teardown spec just implemented
[04:27] <Keybuk> it seems kernel-level
[04:28] <SEJeff> Keybuk: kernel-level is above me but advanced shell scripting / python is not.
[04:28] <Keybuk> I mean that it seems that you'd be shutting down and starting the system AT the kernel level
[04:29] <Keybuk> i.e. you replace a call to reboot() with a call to kexec()
[04:29] <Keybuk> and what gets started as a result is a new kernel, that would start upstart again
[04:29] <SEJeff> Keybuk: Sort of. It isn't done at the kernel level though. It is done through the init scripts. Hence the need for upstart integration
[04:29] <wasabi> Might pass a "quick" option to upstart.
[04:29] <wasabi> As the first option to "shutdown". ;)
[04:29] <Keybuk> is there any change to the boot?
[04:29] <wasabi> Or something. ;)
[04:29] <Keybuk> or is it just to the final step in the shutdown?
[04:30] <wasabi> Kexec is a patch to the Linux kernel that allows you to boot directly to a new kernel from the currently running one. In the boot sequence described above, kexec skips the entire bootloader stage (the first part) and directly jumps into the kernel that we want to boot to. There i
[04:30] <wasabi> ^ sounds like it's done in the kernel to me.
[04:30] <Keybuk> right
[04:30] <SEJeff> Keybuk: the boot is identical but the reboot is different.
[04:30] <Keybuk> ok
[04:30] <Keybuk> how do you mean "the reboot is different"?  different how?
[04:30] <SEJeff> wasabi: kexec is already integrated into the mainline AND ubuntu kernel
[04:30] <SEJeff> Keybuk: Just a sec...
[04:30] <wasabi> Nice. So, the work is done. Just need a user space utility to read grub's config. :)
[04:30] <wasabi> ANd send the first initramfs/kernel into kexec.
[04:31] <SEJeff> wasabi: actually, install kexec-tools from universe and everything you need is in place
[04:32] <wasabi> Sure. YOu just need to make the last task in the runlevel change call it.
[04:32] <SEJeff> /etc/init.d/reboot would need to be changed
[04:32] <SEJeff> Correct
[04:32] <Keybuk> right
[04:32] <Keybuk> so the way that shutdown upstart works (downshut? :p) is:
[04:32] <wasabi> Or however we figure out how to make upstart do that right.
[04:32] <wasabi> Not sure what that is. It sends reboot() on it's own doesn't it?
[04:32] <Keybuk> upstart emits a "shutdown" event (so apps can stop, etc.)
[04:32] <jdub> "I can't understand someone like Osama bin Laden understanding the joy of Hanukkah." - GWB
[04:32] <Keybuk> once that's been handled, upstart then emits a second event, that is specified by /sbin/shutdown
[04:33] <Keybuk> power-off for -P
[04:33] <Keybuk> system-halt for -H
[04:33] <Keybuk> reboot for -r
[04:33] <Keybuk> etc.
[04:33] <wasabi> AHh.
[04:33] <Keybuk> you can specify a custom event as well, and we can add options for others
[04:33] <Keybuk> so we could have a fast-reboot for /sbin/shutdown -Q now or something
[04:33] <Keybuk> then we just make sure any scripts with "on reboot" are also run for "on fast-reboot"
[04:33] <Keybuk> except the one that calls /sbin/reboot -f, that would be replaced by one that calls kexec
[04:33] <SEJeff> Right
[04:34] <wasabi> jdub, obviously the whole "muslim" thing threw him off.
[04:34] <Keybuk> the sysvinit analogy would be adding a runlevel 7 for a fast reboot
[04:35] <SEJeff> ok
[04:35] <wasabi> yay I fixed my fss enough to save some data
[04:36] <SEJeff> I won't be able to set up an edgy box for a few weeks until I complete my move to Cali for a new Sys Admin job but want to work on what I can here and now
[04:36] <Keybuk> how much does Windows restart on a shift-restart?
[04:37] <Keybuk> it's just the shell, isn't it?  not the whole kernel?
[04:37] <jdub> windows 9x?
[04:37] <SEJeff> Keybuk: thats how I understand it
[04:37] <wasabi> shift-restart?
[04:37] <jdub> that was just 'win'
[04:37] <jdub> not reboot+dos
[04:37] <SEJeff> wasabi: it is a faster method of rebooting
[04:38] <wasabi> Thought windows just did a plain ol' power off reboot, but with a fast boot up time.
[04:38] <wasabi> (preload all previously used drivers into a big blob)
[04:38] <jdub> shift-restart still exists in nt land?
[04:38] <wasabi> Sure didn't think so.
[04:38] <jdub> (the turn off dialogue doesn't change labels for restart as it does for sleep)
[04:38] <SEJeff> wasabi: well that can't be correct because shift reboot doesn't do the bios post. That means it has to be a software only reboot
[04:39] <wasabi> How does one do "shift reboot" on NT?
[04:39] <wasabi> I see no such feature here. =/
[04:40] <SEJeff> wasabi: http://www.lawrencegoetz.com/95tips.htm
[04:40] <wasabi> The name of tha tURL does not suggest NT.
[04:41] <SEJeff> wasabi: I can't speak for nt
[04:41] <wasabi> Yeah. NT doesn't support that no more. That means XP.
[04:43] <SEJeff> Keybuk: So I'm curious to the upstart design and how difficult you think it would be to do this
[04:43] <Keybuk> SEJeff: very easy
[04:43] <Keybuk> see above
[04:43] <SEJeff> Keybuk: As in can you think of any gotchas or things I should look into
[04:44] <Keybuk> nothing obvious
[04:44] <Keybuk> we already have three different "when you're done shutting down" things
[04:44] <Keybuk> I don't see how a fourth option hurts
[04:45] <SEJeff> Keybuk: Great news to hear. You might see my post to ubuntu-devel sometime earlier today about this. Feel free to reiterate the same thing.
[04:45] <Keybuk> I haven't caught up on u-d yet
[04:49] <SEJeff> Keybuk: good luck with that one :)
[04:56] <Keybuk> SEJeff: mostly I use the "Delete" key
[04:59] <wasabi> Keybuk, all that upstart stuff makes me wonder how shutdown would be optimally implemented in that case.
[04:59] <wasabi> You'd think you'd have everything set to "stop on shutdown"
[04:59] <wasabi> Oh wait n/m. Nothing changes.
[04:59] <wasabi> Hah. 
[04:59] <wasabi> n/m. take back everything. forget I spoke.
[05:00] <Keybuk> "stop on shutdown" would be things that are teardown-exempt
[05:01] <Keybuk> "on reboot/halt/power-off" would be things like mounting the disks, etc.
[05:01] <wasabi> Yeah. For a second I had forgotten that shutdown and poweroff were seperate things.
[05:02] <Keybuk> this isn't entirely in stone yet, of course
[05:02] <wasabi> Yeah
[05:20] <jdub> Keybuk: so how many distro team people are refusing to enter the USA for UDSMV?
[05:20] <Lathiat> UDSMV?
[05:20] <jdub> ubuntu developer summit mountain view
[05:20] <Lathiat> ah i see
[05:20] <wasabi> When is that?
[05:20] <jdub> nov
[05:21] <wasabi> Oh crap.
[05:21] <wasabi> I had no idea it was that soon.
[05:21] <mjg59> See Fridge
[05:21] <wasabi> wiki???
[05:21] <Lathiat> hrm i should hurry up and finish my lca proprosal
[05:21] <jdub> yes you should
[05:21] <Lathiat> interesting mvoe to have it in MV
[05:21] <Lathiat> as you say i generally see that alot of developers have a hard time getting into the US and such
[05:22] <AlinuxOS> hello mjg59 ;)
[05:22] <wasabi> Hey, I have a lot of trouble getting OUT
[05:22] <Keybuk> jdub: none that I'm aware of
[05:22] <Keybuk> a couple of us had immigration issues, I've sorted mine out, I believe the person who forgot to return his slip last time is also sorting that out
[05:23] <mjg59> Last time I had US immigration issues, I got a passport from another country...
[05:23] <wasabi> Keybuk, have any times during the UDSMV week where upstart stuff is planned?
[05:23] <wasabi> I might not be able to make it for the entire week.
[05:24] <jdub> plans! hah!
[05:24] <wasabi> =(
[05:25] <wasabi> destination airports?
[05:25] <mjg59> wasabi: SFO or San Jose
[05:25] <wasabi> thx
[05:25] <mjg59> San Jose is slightly nearer, but not very international
[05:25] <wasabi> coming from the us. ;)
[05:25] <mjg59> Yeah, San Jose is probably a better bet
[05:26] <Keybuk> wasabi: we don't do any form of planning I'm afraid
[05:26] <Lathiat> whats SFO?
[05:26] <mjg59> I'm hoping to get there, but I won't know if it's possible for a month or so
[05:26] <mjg59> Lathiat: San Francisco
[05:26] <Keybuk> wasabi: usually the schedule is drawn up that morning, and altered throughout the day
[05:26] <Lathiat> ah right
[05:26] <Lathiat> whats the 'O' mean?
[05:26] <wasabi> heh. nice.
[05:26] <Keybuk> wasabi: generally you register when you turn up, and your interest in particular specs
[05:26] <mjg59> Lathiat: Airport codes rarely have a great deal to do with the city name
[05:26] <Lathiat> oh right
[05:26] <Lathiat> airport code
[05:26] <Keybuk> and then the scheduler ensures we talk about things of interest to you on the day that you are there
[05:27] <Keybuk> uh, I mean by that "you register when you are going to turn up, in launchpad, in advance"
[05:27] <mjg59> Keybuk: Is the scheduler automatic now, or is there still someone stuck doing that all evening?
[05:27] <Keybuk> which always struck me as a public list of people whose house was going to be empty, but still
[05:28] <Keybuk> mjg59: jamesh's schedule-o-matic
[05:28] <wasabi> I doubt I'll be able to afford this either.
[05:28] <wasabi> Might turn out to not be practical.
[05:28] <Keybuk> wasabi: it'd be a shame, you'd be great to have there
[05:28] <Keybuk> how far do you have to travel?
[05:28] <wasabi> Not far at all. Tickets aren't a problem.
[05:29] <wasabi> Hotel is. ;)
[05:29] <wasabi> Also gf would insist she be brought along.
[05:29] <wasabi> Heh.
[05:29] <jamesh> the plan is to move the schedulomatic code into Launchpad, so that it is easier for multiple people to work on the schedule
[05:29] <Keybuk> I can suggest sponsorship for you, but not the partner
[05:29] <Keybuk> jamesh: btw, can we mark our attendance as private please? :p
[05:30] <jamesh> stalking problems?
[05:30] <Keybuk> (if I can't get my own partner sponsored ... etc.)
[05:30] <wasabi> Heh. I don't know what the requirements of "sponsorship" are. =)
[05:30] <Keybuk> jamesh: house insurance problems
[05:30] <Keybuk> after the recent "google calendar" scandal, they've added a condition to the insurance
[05:30] <Lathiat> er, haha
[05:30] <wasabi> Ya know, if I could find a place to stay, a room to share, somebody to split room with... I might be able to leave her at home. ;)
[05:30] <jamesh> Keybuk: file a bug
[05:31] <wasabi> I will have to see. And buy some roses.
[05:31] <jamesh> so they don't want you advertising to the world when your house will be empty?
[05:31] <mjg59> wasabi: After what happened to Mez, that might be a bad idea
[05:31] <wasabi> ?
[05:31] <wasabi> Uh oh. Conference horror stories.
[05:31] <mjg59> Oh, UBZ in Montreal
[05:31] <Keybuk> jamesh: exactly
[05:31] <Keybuk> jamesh: in fact, if I do so on a public web page, I am uninsured
[05:32] <wasabi> What's this insurance stuff?
[05:32] <jamesh> Keybuk: that'd be a pain if you were a speaker at a conference
[05:33] <jamesh> "come to $CONFERENCE and you may or may not see a talk by Keybuk"
[05:33] <Lathiat> indeed
[05:33] <Lathiat> haha
[05:33] <jdub> Keybuk: suck!
[05:33] <Keybuk> jamesh: the wording implies publishing of one's schedule
[05:33] <jdub> insurance companies are such parasites
[05:34] <Keybuk> though I'm sure they'd wriggle out of it in any case
[05:35] <Lathiat> so i wonder when the ubuntu+google rumor mill will start spinning again :)
[05:35] <Keybuk> which rumour?
[05:35] <Lathiat> rumors liek google working on a ubuntu based googleOS and other such things :)
[05:37] <Keybuk> they are?
[05:37] <Keybuk> they use it internally
[05:37] <Keybuk> this is well known
[05:37] <Lathiat> nah nah like for public release and such
[05:37] <Keybuk> and there are many comments from senior people of both companies that confirm this
[05:37] <Lathiat> google pcs, takign over the world
[05:37] <mjg59> Keybuk: Nonono, that's not true
[05:38] <mjg59> After all, asuffield says that Ubuntu has no large deployments
[05:39] <wasabi> Yeah, not making it.
[05:39] <wasabi> Can't do the hotel. That's really the deal breaker.
[05:39] <Lathiat> mm i love this i just searched for flights from here to SFO its $1600 + taxes... of $550
[05:39] <Lathiat> which they dont give you the value of until you click through to start actually booking the flight
[05:40] <mjg59> wasabi: Well, you can always apply for sponsorship
[05:40] <Keybuk> mjg59: asuffield can blow me
[05:40] <mjg59> Keybuk: I'm sure you wouldn't actually want that
[05:40] <wasabi> I don't have much to stand on. I don't spend lots of time commiting code that helps ubuntu these days.
[05:40] <Keybuk> yes, actually, you're quite right
[05:41] <Keybuk> asuffield can blow mark ray
[05:41] <jdub> *** WARNING: THIS IS NOT A DEBIAN FANFIC CHANNEL ****
[05:42] <Keybuk> jdub: isn't it called "shipping" these days?
[05:45] <wasabi> the wiki is being annoyingly slow.
[06:01] <wasabi> Keybuk, added my name to the sponsor list thingy. Best I can do now unless I win the lottery or something. :)
[06:02] <Keybuk> wasabi: alrady done
[06:03] <wasabi> I know. I did it... ?
[06:07] <Keybuk> no, I did
[06:10] <wasabi> I did first, then.
[06:10] <wasabi> I hope we didn't just overwrite each other. ;)
[06:11] <wasabi> Unless you have a new super-secret list.
[06:12] <Keybuk> oh, there's an internal list yeah
[06:13] <wasabi> thanks. sleepy time.
[07:37] <fabbione> morning guys
[07:38] <mneptok> hey fabbione 
[07:39] <fabbione> hey kurt
[07:46] <bluefoxicy> why does my home have 48 objects in it and take up of 5 seconds to load
[07:46] <bluefoxicy> while a folder with 100 objects takes under a second
[07:47] <bluefoxicy> even ls cries about it
[07:47] <bluefoxicy> what do I need to do, defragment the damn thing?
[07:48] <mneptok> maybe get rid of that MFM drive?
[08:07] <fabbione> no wonder the initramfs in edgy is at least 3 MB bigger than dapper
[08:07] <fabbione> we are pulling in libc and libpthread
[08:07] <fabbione> and in dapper we are pulling also optimized libc.. that's kind of pointless without libc
[08:09] <bluefoxicy> ouch.
[08:09] <bluefoxicy> I need sleep
[09:03] <pitti> Good morning
[09:05] <Burgundavia_> whiprush: you around?
[09:08] <dholbach> HAPPY HUG DAY!
[09:08] <Burgundavia_> hey mvo, dholbach
[09:09] <dholbach> hey mvo, pitti, Burgundavia_
[09:11] <mvo> hey burgundavia, dholbach!
[09:13] <nixternal> hug day already?  wooopeeee
[09:14] <dholbach> yeah!
[09:14] <dholbach> let's all join #ubuntu-bugs and get the party going :-)
[09:14] <nixternal> woohooo 2:15am ;)
[09:14] <nixternal> give me 7 hours ;)
[09:18] <zyga> morning
[09:18] <zyga> mvo: answering your yesterday's question: I did remove that apt bit because I didn't quite understand how it worked and I didn't (selfish me) need it for a pure one release mirror
[09:19] <zyga> so yes - we have stuff since warty :)
[09:19] <janimo> Riddell: thanks for the xubuntu artwork ;)
[09:20] <mvo> zyga: heh, ok :) I will have a look at it today and include it again then
[09:20] <zyga> okay :)
[09:20] <zyga> I can do it in the evening though - you're quite busy
[09:22] <janimo> doko, if a python binary has .install files under debian/ do I add the python dirs manually (/usr/lib/python2.{4,5}), or is there some way to have those inferred as well by pycentral?
[09:22] <janimo> the source package is exo which builds not only the python bindings but the C library as well, so .install files are needed
[09:23] <janimo> it's recent rebuild di not make is have 2.5 support as the .install file only lists /usr/lib/python2.4
[09:48] <pitti> Mithrandir: do you use the test matrix or just IRC status messages for CD testing? I'm going to test the powerpc CDs
[09:49] <Mithrandir> pitti: the matrix.
[09:49] <doko_> janimo: it shouldn't matter, the install file should list /usr/lib/python*
[09:50] <janimo> doko_: thanks
[09:50] <pitti> 'dont-test-yet' - ok, so I'm going to test the desktop CD modes first
[09:51] <pitti> Mithrandir: want me to update the desktop timestamp to 20060913?
[09:51] <Mithrandir> pitti: yes, please.
[09:51] <Mithrandir> pitti: if you'd update the install cd timestamp to 20060912.1 too, that'd be good
[09:52] <infinity> Mithrandir: Mornin'.  Do you want me to freeze edgy officially and see if malcc's fixes work?
[09:52] <janimo> doko_: I assume the dpkg-genchanges warning unknown information field `Xb-Python-Version' is expected?
[09:52] <pitti> Mithrandir: from dont-test-yet? so we shall test them now?
[09:53] <doko_> janimo: yes
[09:53] <Mithrandir> pitti: yes
[09:54] <pitti> hi infinity 
[09:54] <Mithrandir> infinity: yes, please, since it seems I still can't do it myself.
[09:54] <infinity> Mithrandir: Done.  Let's see how this works this time. :)
[09:55] <Mithrandir> infinity: it'll be.. interesting.
[09:56] <infinity> Mithrandir: Well, Ican't wait for the first freeze exception to see how it goes.
[10:01] <Kamion> morning
[10:02] <infinity> Evenin'.
[10:02] <Fujitsu> Evening, Kamion.
[10:03] <realist> Is there an easily accessible list of orphaned packages up for adoption anywhere?
[10:04] <fabbione> Mithrandir: permission to upload lvm2 to add -fno-stack-protector to CFLAGS. It fixes a bus error on sparc and some random segfault i noticed on i386 with certain disk labels
[10:04] <fabbione> Mithrandir: it's not a blocker for knot but it would be nice to have
[10:04] <Fujitsu> Isn't Soyuz meant to notify me when my sync requests are uploaded?
[10:05] <Fujitsu> And they didn't appear on my packages page either :S
[10:05] <Mithrandir> fabbione: if nothing else, it'd be a useful way to test the FROZEN bit now.  Please upload.
[10:06] <fabbione> Mithrandir: thanks. will do in 2 minutes max
[10:06] <Fujitsu> Am I right in saying that sync requests should be done in the name of the requestor, not the approver (ie. me rather than crimsun)?
[10:07] <Fujitsu> Sorry, `syncs should be done', not `sync requests should be done'.
[10:10] <pitti> Mithrandir: ugh, it seems Kamion's ppc usplash fixes did not make it to the desktop CD?
[10:10] <fabbione> Mithrandir: uploaded
[10:16] <fabbione> Mithrandir: i just got the mail from Ubuntu Installer that it has been Accepted
[10:17] <infinity> fabbione: Grr.  Really?
[10:17] <fabbione> infinity: only that one
[10:17] <fabbione> not the one to the mailing list
[10:17] <fabbione> but yeah.. "Really!"
[10:17] <infinity> Oh, no.  It's in UNAPPROVED.  Yay.
[10:17] <infinity> The mail is just incorrect.
[10:18] <Kamion> Fujitsu: I'm a bit haphazard about that
[10:18] <Fujitsu> Kamion, what do you mean?
[10:19] <Kamion> Fujitsu: normally what I try to do is make it be in the name of the first person in the chain who knew what a sync request was; so for example if some random user files a bug saying "I'd like to have a newer version of such-and-such", then that doesn't count, but the first person who knew what they were talking about gets their name in the changelog
[10:19] <Fujitsu> Well I didn't :(
[10:19] <Kamion> Fujitsu: however this requires a fair bit of memory from me of who's who, so I'm not entirely consistent
[10:19] <Fujitsu> Yeah...
[10:19] <Kamion> Fujitsu: I beg to differ - you very often do
[10:19] <Fujitsu> It was Keybuk :)
[10:19] <Fujitsu> I often do.
[10:19] <Kamion> I've typed 'sync-source -b fujitsu' lots of times
[10:19] <Fujitsu> But not with this last batch.
[10:19] <fabbione> infinity: now i got 2 times the email about Accepted
[10:20] <Fujitsu> fabbione, fantastic!
[10:20] <Kamion> I don't think it will be entirely consistent until you're in ubuntu-dev
[10:20] <fabbione> infinity: ah feh.. it sends one to me and one to the mailing list
[10:20] <Fujitsu> Probably.
[10:20] <fabbione> infinity: weird.. before it was only one
[10:20] <Kamion> Keybuk and I have never talked about this practice; it's just what makes sense to me
[10:20] <infinity> fabbione: Yeah, that buglet's not a big deal, just annoying.
[10:21] <infinity> fabbione: At least it appears to be doing the right thing in the queues now...
[10:21] <Kamion> pitti: the livefses seem very out of date
[10:21] <fabbione> infinity: yeah.. emails.. who cares :)
[10:21] <Kamion> cdimage/www/full/daily-live/current/edgy-desktop-amd64.manifest:usplash 0.4-14
[10:21] <Kamion> cdimage/www/full/daily-live/current/edgy-desktop-i386.manifest:usplash 0.4-20
[10:21] <Kamion> cdimage/www/full/daily-live/current/edgy-desktop-powerpc.manifest:usplash 0.4-20
[10:21] <pitti> ugh
[10:22] <pitti> Kamion: this will make 'check CD' totally unusable on ppc, too -- so should we roll new live fses?
[10:22] <Kamion> definitely, none of the most recent batch of builds worked
[10:23] <Kamion> Setting up human-theme (0.1-0ubuntu2) ...
[10:23] <Kamion> update-alternatives: unable to make /usr/share/icons/default/index.theme.dpkg-tmp a symlink to /etc/alternatives/x-cursor-theme: No such file or directory
[10:23] <Kamion> dholbach: didn't you have a fix in progress for that?
[10:23] <Zdra> hi, any chance to have the package spiftacity updated with metacity 2.16 and his compositor ?
[10:23] <Kamion> oh, apparently that was a different bug
[10:23] <Kamion> I'll fix it now
[10:23] <Kamion> (human-theme that is)
[10:24] <pitti> iwj: any idea why the firefox font on the current live CD looks totally ugly? (not hinted and subpixel'ed); the desktop in general is fine, so is the ffox menu bar, it only affects the document contents
[10:24] <Fujitsu> Also, will builds that are of state `Dependency wait' automatically restart themselves when the dependency becomes available?
[10:25] <slomo_> pitti: same here... looks much better (at least in epiphany) when selecting bitstream vera as fonts instead of sans-serif/monospace
[10:26] <dholbach> Kamion: urg, I thought I had fixed that one - was that an upgrade problem?
[10:26] <infinity> Kamion: The freeze appears to be correctly shunting things to -Q unapproved, so don't forget to drive drescher to accept stuff. :)
[10:26] <Kamion> dholbach: no, fresh install. I've just committed a fix to bzr
[10:27] <infinity> Kamion: And, on the flipside, it means I can upload non-approved stuff.  YAY.
[10:27] <dholbach> Kamion: thank you very much
[10:27] <Kamion> dholbach: the guideline is that every time you use update-alternatives, you must ship the target directory in the package
[10:29] <infinity> Fujitsu: Yes, dep-wait handling is automatic.
[10:30] <Fujitsu> OK, thanks :)
[10:30] <pitti> slomo_: does the current edgy/alternate ppc work for you? 'check' complains about a wrong md5sum in the release file
[10:30] <pitti> slomo_: I try another burn, though
[10:31] <slomo_> pitti: i tried two days ago and the cd drive didn't even accept the alternate cd
[10:32] <pitti> hm, I do an alternate install from time to time, but I seldomly use 'check'...
[10:33] <dholbach> Kamion: that makes sense - thanks again.
[10:33] <dholbach> rodarvus: thanks for joining the telepathy team :-)
[10:34] <mneptok> "telepathy" is something insane people tell themselves they have to explain the voices in their heads
[10:34] <Kamion> really committed to bzr now - I didn't mean to use get rather than checkout ...
[10:43] <pitti> Kamion: against which package do I file a failing 'check' boot target?
[10:43] <pitti> I ruled out bad media by checking the md5's manually
[10:44] <pitti> Kamion: cdrom-checker?
[10:44] <Kagou> pitti: for firefox ugly fonts, have you looked also at yelp ?
[10:44] <pitti> Kagou: no, will do next time
[10:45] <Kamion> pitti: alternate or desktop?
[10:45] <pitti> Kamion: alternate
[10:45] <Kamion> pitti: cdrom-checker, yes
[10:45] <pitti> thank you
[10:45] <Kamion> I've never heard of it failing on genuinely good media and drive though
[10:46] <Kamion> it basically just does md5sum -c on each line
[10:46] <infinity> Kamion: Is this you?
[10:46] <infinity>    94377 | S- | human-theme          | 0.1-0ubuntu3         | six minutes
[10:46] <infinity>          | * human-theme/0.1-0ubuntu3 Component: main Section: x11
[10:46] <Kamion> infinity: yeah
[10:46] <Kamion> I'll approv it
[10:46] <Kamion> +e
[10:46] <infinity> Too late.
[10:46] <infinity> Now you can have the keys. :)
[10:46] <Kamion> heh, ok
[10:51] <pitti> Kamion: me neither, but see bug 60188, I think I correctly compared md5sums
[10:51] <Ubugtu> Malone bug 60188 in cdrom-checker "check does not work on ppc/alternate 20060912.1" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/60188
[10:51] <MrFaber> hi all
[10:51] <Kagou> pitti: for firefox ugly fonts -> Bug #56682
[10:51] <Ubugtu> Malone bug 56682 in Ubuntu "Raster fonts appear in Edgy" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/56682
[10:52] <pitti> Kamion: the md5sum in md5sum.txt is definitively wrong
[10:52] <Mithrandir> it's an ubuntu-cdimage bug, then.
[10:52] <Kamion> definitely wrong? you said in the bug that it was definitely right
[10:52] <thom> http://benjamin.smedbergs.us/blog/2006-09-12/deploying-the-airbag/
[10:52] <Mithrandir> Kamion: he said it matched the .iso on his disc.
[10:52] <Kamion> or did you?
[10:52] <Mithrandir> (and that the .iso matched what's on cdimage.u.c)
[10:52] <Kamion> I don't understand
[10:53] <MrFaber> Is it possible that edgy changed the shell option of users? If I add an user in edgy or just upgrade vom dapper to edgy and have another user accounts than the standard one console shows for this one only a $ . After useing usermod -s /bin/bash USER everything works fine. It is a bug or a feature?
[10:53] <pitti> Kamion: I mean the md5sum of the actual file on the CD is right
[10:53] <pitti> Kamion: but the sum that c-c checks against in md5sum.txt is wrong
[10:53] <Kamion> MrFaber: /bin/sh is dash now, but we should be giving new users bash by default, imho
[10:54] <Kamion> $ isoinfo -R -i edgy-alternate-powerpc.iso -x /dists/edgy/main/binary-powerpc/Release | md5sum
[10:54] <Kamion> 76da0c7fb89cfae9595a369f927bb1dd  -
[10:54] <Kamion> $ isoinfo -R -i edgy-alternate-powerpc.iso -x /md5sum.txt | grep /dists/edgy/main/binary-powerpc/Release
[10:54] <Kamion> 76da0c7fb89cfae9595a369f927bb1dd  ./dists/edgy/main/binary-powerpc/Release
[10:54] <Kamion> no ubuntu-cdimage bug there
[10:54] <MrFaber> Kamion, yes please since sh doesn't support tab completion of file names and so on
[10:54] <MrFaber> Kamion, but standard user still uses bash
[10:55] <Kamion> MrFaber: please file a bug on whatever you used to add the user
[10:55] <pitti> Kamion: argh, I picked the wrong file, sorry; /me updates the bug
[10:56] <MrFaber> Kamion, ok
[10:56] <ogra> is base-install failing on initramfs-tools known ? 
[10:56] <Kamion> pitti: it's still correct
[10:56] <pitti> Kamion: so, md5sums are all correct
[10:56] <Kamion> $ isoinfo -R -i edgy-alternate-powerpc.iso -x /dists/edgy/restricted/binary-powerpc/Release | md5sum
[10:56] <Kamion> b60e1e09e13135b8537380ebb1882f60  -
[10:56] <Kamion> $ isoinfo -R -i edgy-alternate-powerpc.iso -x /md5sum.txt | grep /dists/edgy/restricted/binary-powerpc/Release
[10:57] <Kamion> b60e1e09e13135b8537380ebb1882f60  ./dists/edgy/restricted/binary-powerpc/Release
[10:57] <pitti> Kamion: yeah, I updated the bug, md5sums all match, and still 'check' complains and claims it's wrong
[10:57] <Fujitsu> Wazzup with it, StevenK?
[10:57] <StevenK> Fujitsu: It hasn't been touched for a while.
[10:57] <Fujitsu> StevenK, how long's a while?
[10:58] <Fujitsu> Mine sat for 8 days.
[10:58] <Mithrandir> pitti: what does grep /dists/edgy/restricted/binary-powerpc/Release md5sum.txt | md5sum -c say?
[10:58] <StevenK> Fujitsu: Damn it, I was hoping to get away with not checking.
[10:58] <Kamion> pitti: updated bug
[10:59] <Kamion> StevenK: it's still in the queue
[10:59] <StevenK> Heh, 4 days. It seems I'm less patient than I thought.
[10:59] <StevenK> Kamion: Yeah, sorry, I honestly thought it had been longer.
[10:59] <Kamion> looks like we're a bit backed up unfortunately
[10:59] <ogra> Kamion, base-install breaks for me when installing initramfs-tools is that a known one or should i run a new install to get a log ? it says initramfs-tools is already the newest version and is not allowed to remove it ?
[10:59] <Mithrandir> (food)
[10:59] <dholbach> Mithrandir: bon apptit
[10:59] <pitti> Mithrandir: 'Ok'
[10:59] <Kamion> ogra: not known, I'll try to duplicate in a moment; a bug on base-installer with the full syslog would be appreciated
[10:59] <pitti> Mithrandir: same for 'main'
[11:00] <ogra> ok, i'll run a new install then ... (need to upload some other fixes before)
[11:07] <pygi> sivang, poke?
[11:09] <dholbach> troy_s: did you have luck with subscribing ubuntu-art to bugs of packages?
[11:10] <pitti> Kamion: (bug updated)
[11:13] <giftnudel> are there known problems with the german mirror of archive.ubuntu.com?
[11:13] <pitti> giftnudel: doesn't work for me either ATM
[11:14] <giftnudel> it has been bad for nearly a week now
[11:15] <Mithrandir> pitti: weird.  Can you get the syslog with DEBCONF_DEBUG=5 out somewhere?
[11:16] <pitti> Mithrandir: where do I set that?
[11:16] <Kamion> I'm not sure how much that will help ...
[11:16] <pitti> as boot option?
[11:16] <Kamion> kernel boot parameter
[11:16] <pitti> will do after my current test install (only server, should be fast)
[11:17] <Mithrandir> that reminds me, I should rebuild the server ISOs
[11:18] <broonie> Documents just aren't the same without the shading effects.
[11:19] <pitti> broonie: we want cube rotation when scrolling pages, of course
[11:19] <giftnudel> and the paper should behave like paper when dragged
[11:19] <Kamion> mvo: help
[11:19] <Fujitsu> Or... not.
[11:19] <Kamion> mvo: your auto-remove stuff broke d-i - 'apt-get -y --no-remove install <package that is already installed>' exits non-zero
[11:20] <Kamion> ogra: ^-- that's the initramfs-tools problem
[11:20] <ogra> ah
[11:20] <ogra> right 
[11:20] <mvo> Kamion: does it print anything? or just exists zero?
[11:21] <mvo> Kamion: aha, I see the problem
[11:21] <mvo> Kamion: give me a bit
[11:21] <Kamion> mvo: thanks; it's needed for knot-3 obviously
[11:22] <Kamion> it says "E: We are not supposed to delete stuff, can't start AutoRemover"
[11:22] <Kamion> that should be entirely silent and not error, IMHO
[11:22] <ogra> right, thats the message 
[11:22] <mvo> Kamion: I'm on it now, it was a oversight of a side-effect 
[11:24] <Kamion> Mithrandir: can I upload console-setup to switch the dependencies back to xkb-data? I'd like to get rid of the transitional xkeyboard-config from minimal
[11:25] <Mithrandir> Kamion: is there any harm in waiting until after knot 3?
[11:25] <Kamion> I guess not
[11:26] <Kamion> I'll just let it sit in unapproved then
[11:26] <Kamion> just means people have an extra junk package lying around. :)
[11:27] <Mithrandir> yeah, which is unfortunate, but that'll be fixed with the apt magic. :-)
[11:31] <Kamion> mjg59: would you mind if I changed usplash.postinst to fetch the X modes from /etc/X11/xorg.conf rather than debconf? I'd be much more comfortable with that - we shouldn't be relying on values in the debconf db
[11:31] <iwj> pitti: No, no idea, sorry.  I hadn't noticed.  I suppose it's possible that 2.0 beta 2 will fix it, but it sounds more like something's wrong with the font machinery.
[11:32] <infinity> Kamion: Basing it on the config file is saner, yes.  Just do it. :)
[11:32] <infinity> Kamion: Debconf is not a registry and such.
[11:34] <Kamion> just need to suck the sed magic out of dexconf for parsing xorg.conf reliably
[11:34] <thom> iwj: are we likely to see beta2?
[11:34] <Kamion> er, out of xserver-xorg.postinst
[11:35] <mvo> Kamion: fix is here, I'm doing a final test and will upload then. I assume it is urgent?
[11:36] <zyga> mvo: g-a-i postinst crashed during knot-2 current upgrade, interested?
[11:37] <mvo> zyga: yes, can you put it into a pastebin please?
[11:37] <zyga> mvo: not untill evening, the python script could not import pygtk though
[11:37] <Mithrandir> mvo: yes, it's urgent. :-)
[11:39] <pitti> ok, 'check' on amd64 works fine
[11:40] <iwj> thom: Yes, I'm merging it right now.  Urgh, etc.
[11:41] <thom> heh. Urgh indeed
[11:43] <Kagou> doko or iwj around ?
[11:43] <simira> who broke x?
[11:46] <infinity> Kamion,Mithrandir: We win.  Binaries are being built and accepted correctly too.  FROZEN works.
[11:47] <Mithrandir> infinity: yay!
[11:49] <Kagou> For Bug #56682 i'v proposed a solution, but i don't know if it's working, 'cause i have to simulate a clean/fresh installation with a modified version of fontconfig. So if doko  or iwj can have a look at resolving this bug :)
[11:49] <Ubugtu> Malone bug 56682 in fontconfig "Raster fonts appear in Edgy" [Medium,Confirmed]  http://launchpad.net/bugs/56682
[11:52] <iwj> Kagou: Here.  Just reading that now.
[11:53] <iwj> Kagou: Are you `VETSEL Patrice' ?
[11:53] <Kagou> iwj: yes
[11:53] <iwj> Right ...
[11:54] <pitti> hi iwj
[11:54] <iwj> pitti: Hi.
[11:54] <pitti> hm, does anyone know how to get a file from the d-i system to the hard disk or usb stick? all my mount attempts just fail with 'invalid argument' or 'no such device'
[11:56] <pitti> iwj: yup, that's the same issue I pinged you about some hours ago
[11:56] <iwj> pitti: I thought it might be.  Kagou seems to be saying it happens on a desktop install too.
[11:57] <iwj> Kagou: Or was this on a livecd ?
[11:57] <iwj> The code here in the fontconfig postinst is a bit odd.
[11:58] <Kagou> iwj: i'v this bug on since knot1, and i'v done fresh install of knot1 . And fresh install of knot2too
[11:58] <ogra> pitti, install the ssh client udeb and scp it :)
[11:58] <iwj> Kagou: Hmm.
[12:00] <Kagou> iwj: in my dapper system i'v "30-debconf-no-bitmaps.conf" not in edgy knot1 and knot2 after install
[12:01] <pitti> ogra: already tried netcat, no ethernet modules at this stage
[12:01] <ogra> hmm... thats bad
[12:02] <iwj> Where can I read about how the livecd is constructed ?
[12:02] <Kamion> infinity: w00t
[12:02] <Kagou> iwj: i must disconnect. i came back in 1/2 H 
[12:03] <Nafallo> pitti: ehrm... footage of the file in a hexeditor and then playing with a scanner? :-)
[12:03] <iwj> Kagou: OK.
[12:03] <iwj> Kamion: Does the livecd builder to dpkg --extract and then run postinsts, or something ?
[12:03] <iwj> s/ to/ do/
[12:03] <Kamion> iwj: no, just 'apt-get -y install'
[12:04] <Kamion> there's no magic that could be at all relevant here AFAICS
[12:04] <iwj> Hmm.
[12:04] <Kamion> I think looking at casper will be more fruitful
[12:04] <Kamion> although it doesn't do any fontconfig tweaking any more ...
[12:05] <Kamion> iwj: if there's anything in the fontconfig postinst (say) that's specific to the machine (like running laptop-detect), then casper will have to reconfigure it at live CD boot
[12:05] <iwj> Do we know whether this font wrongness happens on D-I installs too ?
[12:05] <iwj> I confess I haven't installed either knot from CD.
[12:06] <Kamion> similarly a desktop CD install (ubiquity) will just copy the read-only filesystem so if there's hardware-specific work in the postinst then it will have to be reconfigured in a ubiquity hook
[12:06] <Kamion> there's no use comparing the live CD to a ubiquity install really, it'd have to be a live CD install
[12:06] <Kamion> er, it'd have to be a d-i install, I mean
[12:06] <iwj> No, the postinst does a lot of silly faff but it's not hardware specific.
[12:07] <Kamion> I seem to remember that fontconfig does LCD detection internally now
[12:07] <iwj> The LCD detection is to do with subpixel rendering IIRC.
[12:07] <iwj> Or autohinting.
[12:07] <iwj> But anyway not whether bitmap fonts are enabled.
[12:08] <doko> going offline again ...
[12:09] <iwj> I'm looking inside the casper source package and it doesn't seem to have anything relevant.
[12:11] <agutierr> hi all
[12:11] <agutierr> someone can I help me with preseeds, debian-installer and partman? :-) Thanks.
[12:11] <Kamion> agutierr: maybe
[12:11] <agutierr> I am using a preseed to auto-install workstations
[12:11] <agutierr> But I have a problem: 
[12:12] <agutierr> I would like to reuse partitions, without mading a new partition talbe
[12:12] <agutierr> table
[12:13] <Kamion> is there going to be free space on the disk?
[12:15] <Kamion> or are you talking entirely about writing over existing partitions?
[12:15] <mvo> Mithrandir: its uploaded and waiting for approval (apt)
[12:15] <pygi> seb128, we'll just wait for libburn to release as it should be soon, then I'll bug you ^_^
[12:15] <pygi> sivang, ping?
[12:15] <seb128> pygi: ok
[12:16] <pygi> seb128, relase is sep. 26...can we do update in debian and sync to ubuntu before the universe freeze?
[12:16] <seb128> pygi: ivoks said he would have a look on them, no?
[12:17] <Kamion> mvo: ok - it looks a bit inefficient in that it'll still run through the cache looking for autoremovable packages, but it'll fix the error
[12:17] <pygi> seb128, before yes, but he won't be able to update in debian :) So it's easier to update it there, and just sync here, no?
[12:17] <seb128> pygi: I'm too busy to start maintain that package to Debian, so no
[12:17] <janimo> mvo: hi, is the recent recommends work in the metas described somewhere? I'd like to know what I need to modify in the xubuntu seeds, and why :)
[12:17] <pygi> seb128, ahm, ok then, no problem
[12:21] <mvo> Kamion: going over the cache is very fast, it shouldn't be a issue
[12:21] <Kamion> ok
[12:21] <kagou> iwj: back
[12:22] <mvo> Kamion: I can do some measurements on older hardware to verify that though
[12:22] <agutierr> Kamion: I would like to mount existing partitions
[12:22] <agutierr> Kamion: and format, if necesary
[12:22] <mvo> janimo: I uploaded a xuubuntu-meta that is prepared for the recommends already
[12:22] <Mithrandir> seb128: on the live cd, why don't I get the edgy gdm theme?
[12:22] <iwj> Kamion: Do you happen to know if the livecd apt-get install -y somehow suppresses apt's usual running of the .config ?
[12:22] <mvo> janimo: all you need to do now is to modify the seeds and put a () around the packages that you want as recommends
[12:23] <seb128> Mithrandir: is edgy-gdm-themes installed?
[12:23] <Mithrandir> seb128: yes
[12:23] <Mithrandir> hmm
[12:23] <Mithrandir> however, it's ancient.
[12:23] <janimo> mvo: I saw you changed the meta, thanks. I was wondering if some packages _have_ to beput in recommends to play nice with ubuntu-desktop or it
[12:23] <Mithrandir> oh
[12:23] <janimo> is entirely optional. And what besides base install size is the motivation behind this?
[12:23] <seb128> Mithrandir: what theme do you get?
[12:24] <mvo> janimo: the use of recommends is complettely optional. the only case I can think of would be xubuntu-system-tools vs gnome-system-tools
[12:24] <Mithrandir> seb128: ignore me, the amd64 live fs failed to build
[12:24] <seb128> ok ;)
[12:24] <iwj> kagou: Hi again.  I'm pretty sure your suggested change is wrong, I'm afraid, but I'm still trying to understand why it's not working the way the author intended.
[12:24] <janimo> mvo: right, and evince-gtk evince maybe
[12:24] <mvo> janimo: those recommends will be installed by default now, so its not really about size, more about choice
[12:24] <janimo> ok
[12:24] <mvo> janimo: to let user remove non-core stuff (scim, accessability etc)
[12:27] <janimo> Kamion: will there be usplash-theme-{k,x,ed}ubuntu packages as well?
[12:29] <kagou> iwj: in fact as 30-debconf-no-bitmaps.conf and 30-debconf-yes... are not at all created we can suspect that debian/fontconfig-config-postinst is wrong no ?!
[12:30] <kagou> except if $enable_bitmaps is set to TRUE
[12:31] <kagou> iwj: it's why i'v suggested modifications on fontconfig-config.config
[12:31] <Mithrandir> doko: prod
[12:31] <Mithrandir> Unpacking ia32-libs-openoffice.org (from .../ia32-libs-openoffice.org_15_amd64.deb) ...
[12:32] <Mithrandir> dpkg: error processing /var/cache/apt/archives/ia32-libs-openoffice.org_15_amd64.deb (--unpack): trying to overwrite `/etc/java/security/security.d/1000-gnu.java.security.provider.Gnu', which is also in package java-gcj-compat
[12:32] <iwj> We can definitely conclude that the debconf key is being set to true.  (Not TRUE.)
[12:32] <iwj> The question is, why ?
[12:33] <kagou> if /etc/fonts/fonts.conf exist i don't see why we have to set to true
[12:33] <Mithrandir> doko: ^^ ; This is a blocker for knot 3.
[12:34] <Nafallo> [12:08]  <doko> going offline again ...
[12:34] <kagou> iwj: may be /etc/fonts/fonts.conf is installed before fontconfig-config.postinst is called. So enable bitmaps is set to true
[12:34] <Mithrandir> Nafallo: sigh, ok, thanks.
[12:35] <Mithrandir> dholbach: Setting up human-theme (0.1-0ubuntu2) ...
[12:35] <Mithrandir> update-alternatives: unable to make /usr/share/icons/default/index.theme.dpkg-tmp a symlink to /etc/alternatives/x-cursor-theme: No such file or directory
[12:35] <dholbach> Mithrandir: Kamion already fixed it
[12:35] <Mithrandir> ok
[12:35] <Kamion> agutierr: I think you may have to write your own partman script to do that unfortunately; I don't think partman-auto can do it natively
[12:36] <agutierr> Ok
[12:36] <Kamion> iwj: I'm pretty certain it doesn't. Note however that debconf will be running in noninteractive mode (and thus db_input will return non-zero etc.).
[12:36] <Kamion> iwj: (it's not apt that runs the .config - it's debconf, via the sourcing of /usr/share/debconf/confmodule from the .postinst)
[12:37] <Kamion> janimo: usplash-theme-ubuntu is misnamed - should've been ubuntu-artwork-usplash. You already have a xubuntu-artwork-usplash.
[12:38] <Kamion> iwj: what debconf key is this?
[12:38] <Kamion> er s/key/question/
[12:41] <Mithrandir> Kamion: I guess we need new alternate CDs too?
[12:41] <Kamion> Mithrandir: after apt has built, yes
[12:41] <Mithrandir> oh, we need that too, yes.
[12:41] <Mithrandir> is that needed for the live cd?
[12:42] <Kamion> iwj: oh, I think I see what's happening
[12:42] <Mithrandir> (shouldn't be?)
[12:42] <Kamion> iwj: somebody has fundamentally misunderstood how debconf runs
[12:42] <Kamion> Mithrandir: probably, yes - ubiquity will get upset otherwise
[12:42] <Kamion> (maybe)
[12:42] <Mithrandir> Kamion: oh well, I get to exercise the buildds a bit, then.
[12:42] <Kamion> actually ubiquity uses --no-upgrade so it might not matter
[12:43] <Kamion> iwj: so, the .config script actually gets run twice. (It MUST be idempotent.)
[12:43] <Kamion> iwj: The first time is during preconfiguration, before anything has been unpacked; so /etc/fonts/fonts.conf won't be there
[12:44] <agutierr> Kamion: do you know if ubuntu-installer with preseeds uses debian-installer ?
[12:44] <Kamion> iwj: The second time is while the package is being configured; IIRC conffiles are moved into place before the postinst is run, so /etc/fonts/fonts.conf is there
[12:44] <Kamion> agutierr: "uses" does not make sense, but the Ubuntu installer is a derivative of debian-installer
[12:44] <Nafallo> maswan: ping
[12:44] <agutierr> ok
[12:44] <fabbione> Kamion: i think he means for the preseed strings where you use debian-installer
[12:45] <Kamion> fabbione: I think he should be more specific then :-)
[12:45] <Kamion> agutierr: ?
[12:45] <fabbione> Kamion: that too :)
[12:48] <carlos> pitti: hi, I'm going to move language pack generation earlier every day
[12:48] <Kamion> iwj: I'm trying to work out what the best fix would be
[12:49] <carlos> pitti: and move back edgy lang packs to be generated daily now that the process has been optimised
[12:49] <pitti> carlos: cool
[12:49] <Kamion> unfortunately (well, in this particular case), db_input does not mark a question as seen if it was never shown (i.e. you aren't running at priority low)
[12:50] <carlos> pitti: I will give you concrete times after tomorrow's export
[12:50] <Kamion> so you cannot use the seen flag as an indicator
[12:50] <iwj> Kamion: Oh.
[12:50] <zyga> carlos: nice :)
[12:51] <iwj> I didn't realise that .config was run via the postinst like that.
[12:51] <Mithrandir> infinity: can you byhand the publisher for me?  I need to roll a new ia32-libs-ooo and won't make :03
[12:51] <carlos> zyga: are you still using those tarballs?
[12:51] <maswan> Nafallo: pong
[12:51] <iwj> The whole scheme is rather crazy.  All of this copying a single bit into a complex encoding involving the existence or not of several files and symlinks.
[12:52] <zyga> carlos: not at the moment but I'll resume the old stuff after I'm finished with cnf
[12:52] <zyga> probably for edgy+1
[12:52] <zyga> (s/probably/for sure/)
[12:52] <carlos> zyga: people.ubuntu.com/~carlos/language-packs is the new location
[12:52] <zyga> I hope you don't store every single one :)
[12:52] <Kamion> .config being run from .postinst is arguably a debconf bug, but (a) would require explicit support for preconfiguration in dpkg to fix and (b) is probably impossible to change now anyway.
[12:53] <iwj> No, I think it's fine.  But it does explain the difficulty.
[12:53] <carlos> zyga: I remove them from time to time
[12:53] <pitti> carlos: hm, you use language-packs/dapper instead of dapper-updates
[12:53] <Kamion> (It has to be run from the .postinst because if you aren't using apt then preconfiguration won't happen.)
[12:53] <iwj> Yes.
[12:53] <zyga> carlos: right, thanks
[12:53] <iwj> Err, I see.
[12:53] <carlos> pitti: hmm.... then I did a mistake yesterday... anyway, I will fix it now in my side, sorry
[12:53] <Kamion> I think the basic problem is that its test for are-we-upgrading in the .config is wrong.
[12:54] <iwj> That's not readily fixable, I think.
[12:54] <Kamion> Find some other way to do that than testing for the conffile and it'll probably become idempotent.
[12:54] <iwj> Yes.
[12:54] <zyga> carlos: rosetta-edgy.tar.gz is -> current, right?
[12:54] <iwj> Err, it unconditionally sets the debconf key./
[12:55] <carlos> pitti: fixed
[12:55] <iwj> Surely it would make more sense for it just to default the debconf key ?
[12:55] <carlos> zyga: is latest snapshot, it's not the on in our official archives
[12:55] <Kamion> iwj: I don't think it can. Boolean questions don't have a "not set yet" state, and as I said above it can't use the seen flag to help.
[12:56] <Kamion> As in, I don't think it can without breaking other thing.
[12:56] <Kamion> s
[12:57] <rodarvus> dholbach, I forgot to upload my package of telepathy-inspector 0.3 to p.u.c yesterday, sorry
[12:57] <rodarvus> its there now
[12:57] <janimo> Mithrandir: exception req for xubuntu main: add newer version of xfwm4 mostly with bugfixes
[12:58] <dholbach> rodarvus: no problem - we're storing all the packaging in bzr now
[12:58] <Mithrandir> janimo: do you need it for knot?  If not, I'd recommend holding it off.
[12:58] <janimo> Mithrandir: as usual not critical, just in case spinning CD for the other releases takes time.
[12:58] <janimo> Mithrandir: ok I'll hold off then
[12:58] <Mithrandir> janimo: your call though, it's xfce-only.
[12:58] <dholbach> rodarvus: so if you get the time to do that, it should be easy for us to review/update it and upload it
[12:58] <dholbach> rodarvus: thanks for your efforts!
[12:58] <rodarvus> thank YOU
[12:59] <Kamion> Unfortunately there's nothing that the postinst does that can easily be tested for in the config. What you really want to know is "has the postinst ever run".
[12:59] <dholbach> rodarvus: http://wiki.ubuntu.com/Telepathy/InBzr
[12:59] <rodarvus> telepathy-inspector is developed by a friend of mine, and he asked for help getting it into Ubuntu ;)
[12:59] <seb128> trying the current i386 desktop CD, bbl
[12:59] <dholbach> rodarvus: NICE
[12:59] <dholbach> rodarvus: it looks pretty cool
[12:59] <dholbach> rodarvus: and the packaging looked good to me too
[12:59] <dholbach> rodarvus: I just noted there was 0.3.0 upstream already
[12:59] <Kamion> iwj: There's one hack that would work. It would probably get things wrong if anyone ever blew away their debconf db, but at worst it would have the wrong default for the question on the next upgrade (and perhaps not even that)
[12:59] <janimo> dholbach: haven;'t looked yet, is telepathy relying on gnome libs or just has parts which work with gnome?
[01:00] <rodarvus> dholbach, it uses scons, which is rather different from anything autotools related
[01:00] <pygi> rodarvus, but scons rules, compared to autotools :)
[01:00] <rodarvus> version 0.2.5.1 had a few bugs (and one bug on the packaging)
[01:00] <rodarvus> which were fixed on 0.3
[01:00] <Kamion> iwj: at the end of the postinst, 'db_fset fontconfig/enable_bitmaps processed true'
[01:00] <janimo> pygi: s/scons/anything/
[01:00] <Kamion> (processed is an arbitrary identifier)
[01:00] <pygi> jamesh, hehe :)
[01:00] <dholbach> janimo: the framework uses glib, dbus, gstreamer stuff and its own libs - some frontends might make small use of gnome stuff (python, etc)
[01:00] <pygi> janimo, hehe*
[01:01] <pygi> dholbach, there is that C#/Mono telepathy client which even works :)
[01:01] <dholbach> rodarvus: ah nice, so it might be worth to get 0.3.0 in
[01:01] <iwj> Kamion: Hmmm.
[01:01] <dholbach> pygi: which one is that?
[01:01] <zyga> mvo: what is curentArchOnly for, in practice?
[01:01] <rodarvus> pygi, scons is nice - I just wish it had more plugins available by default (so you don't have to do it all by yourself)
[01:01] <Kamion> iwj: then in the config, replace 'if [ ! -f "/etc/fonts/fonts.conf" ] ' with 'if ! db_fget fontconfig/enable_bitmaps processed || [ "$RET" != true ] '
[01:01] <zyga> mvo: since we have arch=all now
[01:01] <rodarvus> dholbach, its done already :)
[01:01] <dholbach> we should move telepathy discussion probably to #ubuntu-desktop
[01:01] <pygi> dholbach, I forgot, but I can look for it :P
[01:01] <dholbach> rodarvus: rock and roll :-)
[01:01] <pygi> rodarvus, true :-/
[01:01] <rodarvus> I'll create a product on LP for it too
[01:02] <Kamion> iwj: I believe I'm just about the only person who ever uses flags that way, but it is supported :-)
[01:02] <mvo> zyga: just sugar, if we ever want to pre-caclculate it
[01:02] <Kamion> you might want to add a check for reconfiguration or something too, but the above should fix the live CD case with minimal fallout
[01:02] <Kamion> and in fact the general first-install case
[01:02] <pygi> janimo, you coming to Hungary, right? 
[01:02] <zyga> mvo: it adds an dependency on dpkg 
[01:02] <iwj> Kamion: Is this documented somwhere ?
[01:02] <iwj> I mean, this use of a `processed' flag ?
[01:03] <mvo> zyga: how do you mean?
[01:03] <zyga> mvo: you need dpkg and python2.4 to run that
[01:03] <zyga> but I'm just picky
[01:03] <janimo> pygi: I promised, so I think I am 
[01:03] <Kamion> iwj: flags in general are documented in /usr/share/doc/debian-policy/debconf_specification.txt.gz
[01:03] <Kamion> iwj: the use of a made-up processed flag is an invention
[01:03] <pygi> janimo, nice, see you there perhaps ^_^
[01:03] <janimo> pygi: :)
[01:04] <iwj> Aha.
[01:04] <pygi> janimo, I am supposed to talk, don't know yet if I'll be able to attend :P
[01:04] <Kamion> It's just a way to transmit state around
[01:04] <Kamion> (As if fontconfig-config.* needs any more such ways ...)
[01:05] <iwj> Quite.
[01:06] <iwj> I think I'll do as you suggest but it seems to be just adding wrongness to an already very wrong thing.
[01:07] <Kamion> I agree
[01:07] <kagou> iwj: can you understand the if "/etc/fonts/fonts.conf" file exist -> # We are upgrading. No file for bitmaps means: -> enable_bitmaps="true" ?
[01:07] <Kamion> kagou: See near the end of the postinst where it doesn't create a symlink if enable_bitmaps is true because that's the fontconfig default
[01:08] <Kamion> it's convoluted, but at least consistently so
[01:08] <pygi> janimo, o yes, wanted to discuss something with you = xubuntu related, so when you get time ;)
[01:08] <janimo> pygi: -> /msg now
[01:09] <Kamion> iwj: the alternative would be to always create the symlink (uncomment the above bit in the .postinst) and use the existence of either symlink as a "we're upgrading" indicator
[01:09] <Kamion> iwj: you'd have to think very hard about upgrades, either way
[01:09] <janimo> Kamion: does usplash select the appropriate png depending on detected resoluion? So artworks should have all 4 or so variants?
[01:10] <kagou> Kamion: what is the default for edgy : to enable or not bitmaps ?!
[01:10] <Kamion> janimo: yeah, it iterates through the linked list of themes in the theme library and picks the first one that will fit in the specified resolution
[01:10] <Kamion> kagou: I neither know nor care; I'm just saying how to fix the code
[01:11] <Kamion> iwj: (There's an argument that depending on filesystem state is better than depending on debconf database state, so maybe the latter approach would be better.)
[01:12] <kagou> Kamion: you think that if we create 30-debconf-yes-bitmaps, fonts will be well rendered ? I think no. 
[01:12] <kagou> may be i'm wrong i will test this...
[01:12] <Kamion> kagou: I don't know. Don't talk to me about how fonts will actually look. I'm just being a "how to use debconf correctly" person.
[01:13] <kagou> oh :)
[01:13] <Kamion> And it's obvious that the current code is incorrect (because the config script isn't idempotent across the two runs in a fresh installation with preconfiguration), regardless of what it's intending to do
[01:14] <Kamion> If we fix it to actually work right, then it'll be a lot easier to make the default swing either way
[01:15] <zyga> mvo: hmm, should I mirror and re-scan with ia64 and hppa?
[01:16] <zyga> StevenK: not so large
[01:16] <zyga> StevenK: now it's pretty stable :)
[01:17] <rodarvus> dholbach, did you create a Telepathy project to include all telepathy-related products?
[01:18] <dholbach> rodarvus: yes - http://wiki.ubuntu.com/Telepathy
[01:18] <mvo> zyga: oh jesus ... hppa/ia64 
[01:18] <Treenaks> dholbach: I knew you were going to say that :P
[01:18] <mvo> zyga: don't forget sparc ;)
[01:19] <dholbach> rodarvus: I'd love to see all the goodness in to play with it and make upstream's life easier :-)
[01:19] <StevenK> zyga: I'm just wondering if it's less than 30Gb a month, and how large the archive is.
[01:19] <StevenK> (For i386, amd64, sparc and all)
[01:20] <StevenK> I should really be doing an assignment, but whatever. :-P
[01:20] <rodarvus> dholbach, no, I mean a 'Project' on LaunchPad
[01:20] <dholbach> ahhhhh!
[01:20] <rodarvus> so I can inform it when creating the telepathy-inspector product
[01:20] <dholbach> no, I didn't know where to do that
[01:20] <dholbach> but it'd make sense
[01:20] <rodarvus> indeed
[01:20] <Kamion> mjg59: I think I'm going to change the installer always to install usplash as part of desktop, rather than hacking it in first. That way it can rely on the result of xserver-xorg.config
[01:22] <mvirkkil> pain and suffering with edgy :(
[01:22] <dholbach> rodarvus: https://launchpad.net/projects/+new -> Sorry, you don't have permission to access this page.
[01:22] <Nafallo> hmm
[01:22] <Nafallo> something throws Generic_Recipe_Edgy_agreement.html into my ~ all the time. anyone else got that? :-P
[01:22] <mvirkkil> I installed edgy on an oldish computer, it's not booting all the way to gdm, though gdm comes up when started manually.
[01:23] <mvirkkil> After logging in I get a notice about some HAL failure.
[01:23] <mvirkkil> And nautilus and a lot of other programs don't work.
[01:23] <mvirkkil> I'm suspecting some dbus issue, but that's just a guess.
[01:23] <pitti> mvirkkil: please file a bug against hal and do the steps on the second half of https://wiki.ubuntu.com/DebuggingRemovableDevices
[01:24] <pitti> mvirkkil: it most likely is a dbus failure if it worked for you before and you didn't change hardware
[01:25] <mvirkkil> pitti: well, one piece of hardware which wasn't supported in dapper, and should be in edgy (but in reality isn't, I'm buggind u-kernel about it).
[01:25] <mvirkkil> pitti: I'll try booting without it, to see if it makes a difference. Otherwise I'll file a bug. Thanks :)
[01:26] <infinity> Mithrandir: Gah, missed your request to drive the publisher by hand, sorry.  Was having a late dinner.
[01:26] <pitti> mvirkkil: nevertheless, filing a bug would be appreciated if it is due to that hardware
[01:26] <pitti> mvirkkil: even if it isn't supported, it doesn't mean that it should make the system unusable
[01:26] <Mithrandir> infinity: no problem.
[01:26] <Mithrandir> infinity: it takes almost a full cycle to upload anyway.  Hooray for huge .tar.gz-s
[01:27] <infinity> Yeah.
[01:27] <Kamion> Mithrandir: should I accept xfwm4?
[01:27] <mvirkkil> pitti: and in this case it should be supported :) The kernel got a patch just for this piece (new zydas1211b chip)
[01:29] <Mithrandir> Kamion: I let that call be janimo's, so if he uploaded it, yes.
[01:29] <Kamion> ok, accepted
[01:29] <mvirkkil> pitti: removing the hw didn't make any difference :(
[01:29] <mvirkkil> pitti: booting without it, that is..
[01:30] <mvirkkil> oh well, I'll start looking in to what's up.
[01:30] <pitti> mvirkkil: thanks
[01:31] <mvirkkil> pitti: hmm.. Hal isn't running after bootup
[01:31] <Kamion> I do wish dpkg had a package relationship field that said "please configure me after package A if A is to be installed, but don't worry about it if it isn't"
[01:31] <pitti> mvirkkil: well, that's what the message said :)
[01:31] <StevenK> Kamion: What, Half-Depends? :-P
[01:31] <infinity> Kamion: A soft-depends?
[01:31] <Kamion> maybe Recommends or Suggests should do that
[01:31] <Kamion> yeah
[01:31] <pitti> mvirkkil: please do the debugging steps, it might point out the source of the problem
[01:32] <Kamion> because I want usplash to be configured after xserver-xorg if possible, but I don't want to make usplash require X
[01:32] <infinity> Kamion: Don't Recommends and Suggests already try to order packages in a dependency chain (barring loops)?
[01:32] <pitti> Kamion: Ladies-First: :)
[01:32] <Kamion> pitti: heh
[01:32] <Kamion> infinity: if they do, that would be ideal, but I didn't know about it
[01:32] <infinity> Kamion: If my guess is correct, having usplash suggest xserver-xorg should be good enough.
[01:32] <Mithrandir> Kamion: sounds like a sensible (and easy enough to add) thing for dpkg to do.
[01:32] <Kamion> iwj: any wisdom on this?
[01:33] <infinity> Kamion: The reason I'm guessing this isn't so much because it makes sense (though it does), but because I'd expect as much code re-use as possible in those areas.
[01:34] <infinity> Actually, wait.  dpkg doesn't do loop unrolling and reordering at all, does it?  Doesn't apt do that?
[01:34] <Kamion> At configure time, apt passes a big load of packages to dpkg --configure and says "do it", as far as I know
[01:35] <infinity> I seem to recall a few days ago doing "dpkg -i foo.deb foolib.deb" and having to do two passes, since the first complainted about foo and then worked with foolib.
[01:35] <Kamion> infinity: was Pre-Depends involved?
[01:35] <infinity> Heck if I can recall. :)
[01:36] <Kamion> if not, I'd say that's a dpkg bug. According to the comment above deferred_configure, it's meant to keep going round the packages in a loop until it's done as much as it can
[01:37] <mvirkkil> pitti: hald stops running after printing (a boatload of) messages.
[01:37] <pitti> mvirkkil: can you please attach that log to the bug report?
[01:37] <Kamion> I don't see anything that handles Recommends or Suggests the way you mention, anyway
[01:37] <janimo> Kamion, what is the development section of the seeds? It is included in ship, but ship alreday has the components in devel. Is it used elsewhere too?
[01:39] <Kamion> janimo: it's intended to become a task for use by tasksel. It shouldn't be included in ship
[01:39] <Hobbsee> mjg59: have you heard of any other regressions with the alps hack in xserver-xorg-input-synaptics?  I'm getting bug 47971, as are some other edgy people, it seems.
[01:39] <Ubugtu> Malone bug 47971 in xserver-xorg-input-synaptics "synaptics touch pad(laptop) clicks when i dont want it to" [Critical,Confirmed]  http://launchpad.net/bugs/47971
[01:39] <Kamion> janimo: err, you mean the "development" seed, right?
[01:39] <janimo> Kamion: yes
[01:40] <Kamion> the one starting with python-adns?
[01:40] <janimo> yes
[01:40] <Kamion> that's definitely not included in your ship
[01:40] <janimo> not in ship, but in the STRUCTURE file after ship:
[01:40] <janimo> oh, sorry supportted
[01:40] <Kamion> ship: boot minimal standard desktop
[01:41] <janimo> yeah, they both begin with an s
[01:41] <Kamion> supported: in STRUCTURE should always list all the other seeds
[01:41] <janimo> aha. ok that makes sense then.
[01:41] <Kamion> because everything you seed is by definition supported
[01:42] <janimo> ok, sorry for the noise
[01:44] <Kamion> fabbione: could you add lvm2 to the list of problems at the end of https://wiki.ubuntu.com/GccSsp ?
[01:44] <infinity> Kamion: Oh, hrm.  Seems a sane wishlist, then.  And what I'd expectas resonable behaviour.
[01:45] <infinity> Kamion: Actually, maybe that behaviour's backward.  Perhaps Enhances would be a better way to express that you should configure Y before X.  Or something.
[01:45] <StevenK> infinity: Hey yeah, that actually makes sense.
[01:45] <Kamion> In the absence of that, I wonder how to deal with getting xserver-xorg configured before usplash. Maybe I should hack this into tasksel somehow.
[01:46] <Kamion> infinity: logically I'd expect Recommends to have similar ordering to Depends
[01:46] <janimo> is the alternate CD installer accessible? Does it make sense to have a11y in ship (as opposed to ship-live) in xubuntu?
[01:46] <Kamion> janimo: not yet
[01:46] <infinity> Kamion: Yeah, I'd expect Recommends to be similar to Depends, I'dexpect Suggests to be backwards, perhaps.  I'm a bit fuzzy on that.
[01:46] <mvo> infinity: will you have time later to talk about the auto-dist-upgrade testing and where we could set it up?
[01:47] <infinity> mvo: Did we ever get past elmo's "I don't want that crap on the buildds" and perhaps get some extra DC hardware for it?
[01:47] <infinity> mvo: If not, we eitherneed to talk to elmo, or you and I can set it up outside the DC (I'm happy to host the PowerPC one, for instance)
[01:48] <mvo> infinity: ok. my memory is that he disliked the specific approach how it was implemented when I/we proposed it, but I may be wrong :) [I reworked it since then] 
[01:48] <iwj> Kamion: I think I'll write down all of the states that this madness might encounter and see if they're distinguishable.
[01:49] <iwj> I'll try to think of a plan and talk to you after lunch.
[01:49] <infinity> mvo: Well, yes.  We should perhaps re-discuss this, then. :)
[01:49] <mvirkkil> pitti: https://launchpad.net/distros/ubuntu/+source/hal/+bug/60219
[01:49] <Ubugtu> Malone bug 60219 in hal "Hald exits after starting" [Untriaged,Unconfirmed]  
[01:50] <mvirkkil> oh :)
[01:50] <pitti> mvirkkil: thanks
[01:50] <pitti> mvirkkil: your first contact to ubugtu? :)
[01:50] <Kamion> iwj: Thanks. I have an alternative plan anyway so it's not at all urgent.
[01:50] <mvirkkil> I'll be glad to provide any other information you might need
[01:51] <mvirkkil> pitti: no, just forgot it existed.
[01:51] <pitti> 14:35:49.721 [E]  hald_dbus.c:3258: dbus_bus_get(): Failed to connect to socket /var/run/dbus/system_bus_socket: No such file or directory
[01:51] <mvirkkil> pitti: I was just as surprised last time.
[01:51] <pitti> hmm
[01:51] <mvirkkil> so it seems it's a dbus problem
[01:51] <pitti> mvirkkil: ps aux|grep dbus-daemon?
[01:52] <Mithrandir> doko: hope you're ok with my ia32-libs-ooo upload; I just nuked /etc/java/security from the package.
[01:52] <mvirkkil>  /usr/bin/dbus-daemon --fork --print-pid 8 --print-address 6 --session
[01:52] <pitti> mvirkkil: ok, so no system daemon
[01:52] <pitti> mvirkkil: let's continue this in /msg
[01:52] <mvirkkil> pitti: note though that now I've started gdm manually and logged in etc.
[01:54] <pitti> mvirkkil: did you get my private message?
[01:55] <theine> Is tepepathy in Edgy actually usable right now (i.e. with Google Talk)?
[01:55] <iwj> Kamion: Right.
[01:56] <doko> Mithrandir: ahh, ok. missed that. Did you freshen the packages as well (don't see the install mail yet)
[01:56] <Mithrandir> doko: no, I didn't see a point in doing that.
[02:19] <pitti> seb128: mvirkkil's problem is that /etc/rc2.d is almost empty
[02:19] <pitti> seb128: you had the same problem, right?
[02:19] <pitti> seb128: is there a bug about it, so that I can make bug 60219 a dup?
[02:19] <Ubugtu> Malone bug 60219 in dbus "system dbus not started at boot" [Untriaged,Needs info]  http://launchpad.net/bugs/60219
[02:20] <seb128> I had only README left from all the /etc/rc[0-6] .d
[02:20] <seb128> pitti: the bug I had had been fixed quickly
[02:20] <mvirkkil> seb128: well, I've got almost nothing. 
[02:20] <pitti> seb128: was that a transient upstart bug?
[02:21] <seb128> pitti: I just got that "failed to initialize HAL!" message right now
[02:21] <seb128> pitti: no, it was a transition gnome-system-tools services-admin bog
[02:21] <mvirkkil> seb128: it's  because hal failes to start because dbus isn't alive
[02:21] <pitti> seb128: I guess the dbus rc2.d symlink is missing for you as well?
[02:21] <seb128> pitti: and I didn't got it this morning, so it's probably due to a recent update
[02:22] <seb128> lrwxrwxrwx 1 root root 14 2006-09-09 00:04 /etc/rc2.d/S20dbus -> ../init.d/dbus
[02:22] <jono> if I want to book a meeting in #ubuntu-meeting - do I just choose a time that doesnt conflict with another meeting and mail the fridge team about adding it to the calendar?
[02:22] <mjg59> Kamion: Sure
[02:22] <seb128> jono: seems good
[02:22] <tseng> jono: sounds pretty reasonalbe
[02:22] <jsgotangco> pretty much i guess
[02:22] <Hobbsee> jono: yep
[02:22] <pitti> mvirkkil: ok, so since this seems to have been fixed in a later gnome-system-tools, can we close that bug?
[02:22] <jono> ok cool :)
[02:22] <pitti> seb128: thank you
[02:22] <jono> gonna book a LoCo meeting :)
[02:22] <jono> thanks folks
[02:22] <seb128> pitti: np
[02:23] <Hobbsee> jono: fun :)
[02:23] <jono> Hobbsee, we should talk today about Kubuntu :)
[02:23] <seb128> pitti: bug #60219 doesn't look like that g-s-t issue
[02:23] <Ubugtu> Malone bug 60219 in dbus "system dbus not started at boot" [Untriaged,Needs info]  http://launchpad.net/bugs/60219
[02:23] <pitti> seb128: well, something erased mvirkkil's /etc/rc2.d/
[02:23] <Hobbsee> jono: perhaps.  i've actually got an assignment that i'm supposed to be doing at the moment, but clearly am not.
[02:24] <seb128> mvirkkil: did you use services-admin?
[02:24] <realist> Hobbsee: could you direct me where to look for package upload sponsors?
[02:24] <mvirkkil> seb128: I don't think so :O)
[02:24] <Hobbsee> realist: universe or main?
[02:24] <seb128> mvirkkil: did you played with the NTP option from times-admin?
[02:24] <realist> Also, is there a list of orphaned packages for adoption?
[02:24] <Hobbsee> realist: poke someone in #ubuntu-motu to upload it for universe, or subscribe-ubuntu-universe-sponsors
[02:24] <realist> Hobbsee: universe at this stage
[02:24] <jono> Hobbsee, later this week ?
[02:24] <Hobbsee> realist: you can modify almost anything you like - there arent set "maintainers"
[02:25] <Hobbsee> jono: should be fine
[02:25] <jono> :)
[02:25] <mvirkkil> seb128: Well, I installed ntp support by going through the clock applet
[02:25] <realist> Hobbsee: is that intentional ?
[02:25] <seb128> mvirkkil: did you get that issue just after that? for how long do you have the issue?
[02:25] <Hobbsee> realist: yes
[02:25] <realist> I don't want to end up reproducing other people's work
[02:25] <realist> Or touching other people's code base...
[02:26] <mvirkkil> seb128: I did a fresh install yesterday of knot2 and I've been upgrading and dist-upgrading ever since (slow net connection)
[02:26] <seb128> hum
[02:26] <fabbione> Kamion: sure
[02:26] <mvirkkil> seb128: so this is a pretty clean install.. 
[02:26] <seb128> mvirkkil: ah, knot2 might have had the boggus version
[02:27] <mvirkkil> seb128: so, what can I do to repair the situation?
[02:27] <seb128> reinstall :/
[02:27] <pitti> mvirkkil: cp -a /etc/rc3.d/* /etc/rc2.d
[02:27] <seb128> hum
[02:27] <seb128> rc3.d is ok?
[02:27] <pitti> (provided that the other runlevels are still intact)
[02:27] <mvirkkil> no
[02:27] <pitti> oh
[02:27] <seb128> that bug cleaned rc[0-6] .d on my box
[02:27] <mvirkkil> none of the runlevels have anything useful.
[02:28] <pitti> mvirkkil: then you are pretty screwed, I'm afraid
[02:28] <mvirkkil> A re-install will take me two days.
[02:28] <Kamion> aptitude reinstall '~i' (?)
[02:28] <Kamion> dunno if that will work
[02:28] <Kamion> or dpkg-reconfigure every package that has update-rc.d in its postinst
[02:28] <Kamion> you'd get quite a lot of questions, but you wouldn't have to re-download anything
[02:28] <mvirkkil> How about someone tar their /etc/rc?.d/ and send that to me?
[02:28] <Kamion> you're much safer doing the above
[02:29] <seb128> would require somebody having your exact config
[02:29] <pitti> mvirkkil: that would cover different packages, etc.
[02:29] <Kamion> oh, 'dpkg-reconfigure --default-priority' to get fewer questions
[02:29] <seb128> dpkg-reconfigure on the packages using update-rc.d should be fine
[02:29] <mvirkkil> how can I find out which packages those are?
[02:29] <seb128> I did something like that and got my box booting again correct
[02:30] <seb128> grep update-rc.d /var/lib/dpkg/info/*.postinst
[02:30] <Kamion> cd /var/lib/dpkg/info; grep -l update-rc.d *.postinst
[02:30] <mvirkkil> ok, I'll do that. Thanks for yourhelp :)
[02:30] <seb128> np, sorry for the breakage
[02:30] <pitti> mvirkkil: for i in `grep -l update-rc.d /var/lib/dpkg/info/*.postinst`; do i=`basename $i`; sudo dpkg-reconfigure ${i%.postinst}; done
[02:31] <pitti> mvirkkil: ok, you should have enough shell runes now :)
[02:31] <Mithrandir> Kamion: argh, I forgot to ask you to accept the new ia32-libs-ooo; mind doing that?
[02:32] <Kamion> Mithrandir: will do
[02:32] <mvirkkil> 56 packages...
[02:33] <Kamion> pitti: (dpkg-reconfigure can be given a list of packages)
[02:33] <pitti> right, there's certainly room for optimization
[02:33] <Kamion> probably doesn't matter for 56 packages
[02:34] <Kamion> smurf: is there a way to blacklist certain keysyms from certain keymaps in keymapper?
[02:34] <realist> Hobbsee: I don't quite understand how there can be any quality control, if anyone can change any package in the universe repository?
[02:34] <Hobbsee> realist:  see #ubuntu-motu
[02:35] <Kamion> smurf: the X keymap for gb includes latin, which lists a lot of symbols on AltGr which aren't typically engraved on UK keyboards
[02:35] <Kamion> smurf: so if you just go through the keymap detector without knowing about this, you get US, because it asks if you have an "" key and of course you say no
[02:36] <Kamion> but actually AltGr-a gives 
[02:36] <Mithrandir> Kamion: ISTR hacking keymapper to not look at level 3 or 4 maps.
[02:37] <Kamion> Mithrandir: and uploading the result? :)
[02:37] <Kamion> thing is, we'd need that to be done only for certain keymaps ...
[02:37] <zul> hi, can someone new the xen-image kernel generic stuff for me?
[02:38] <Mithrandir> Kamion: no, since it's part of the "parse X11 keymaps" bit of keymapper rather than the console-* bit of keymapper.
[02:39] <Kamion> zul: done
[02:39] <zul> Kamion: thanks..
[02:39] <Kamion> zul: shouldn't -amd64-generic be renamed to -generic, to match the mainline kernel?
[02:40] <zul> Kamion: it probably should...ill do it for the next upload
[02:40] <Kamion> thanks
[02:56] <doko> seb128: ping
[02:56] <Mithrandir> hmm, ubuntu-server's ports seems to fail to build
[02:56] <fabbione> Mithrandir: hppa is non-existant in edgy
[02:56] <fabbione> Mithrandir: what's the problem with ia64?
[02:56] <Riddell> Kamion: could you check, merge and upload my ubiquity branch, it's got a fix to enable the Install button on the last page
[02:56] <Mithrandir> fabbione: unsure.  It seems to die with tar: ./usr/lib/debootstrap/scripts/edgy: Not found in archive
[02:57] <Kamion> Riddell: knot-3 urgent I guess
[02:57] <Mithrandir> fabbione: I'll investigate later, now I want to get alternate cds built.
[02:57] <Riddell> Kamion: yes, won't install without it
[02:57] <Mithrandir> Riddell: "ouch".
[02:57] <fabbione> Mithrandir: ok
[02:57] <Mithrandir> Kamion: is there anything you know of holding the alternate cds back?
[02:57] <mvirkkil> Yay, seems to be working now (after an additional dpkg-reconfigure dbus). Thanks guys. Now if I could just get my wlan dongle working (the reason I installed edgy in the first place)
[02:58] <pitti> mvirkkil: \o/
[02:58] <jdong> is OOo to be rebuilt soon?
[02:58] <jdong> or are we shipping knot3 with unusable OOo?
[02:58] <simira> ~
[02:59] <Mithrandir> jdong: OOo seems to work fine for me in knot 3
[02:59] <jdong> Mithrandir: you can save new documents without crashing?
[02:59] <Riddell> Mithrandir, Kamion: I noticed that openssh-client isn't installed by default, seems the recommends from ubuntu-standard don't get in
[03:00] <jdong> Mithrandir: bug 58508... it's due to the dbus 0.92 change
[03:00] <Ubugtu> Malone bug 58508 in openoffice.org "cant save under edgy" [High,Confirmed]  http://launchpad.net/bugs/58508
[03:00] <Kamion> hmm, remind me to make "Device for boot loader installation:" translatable at some point
[03:00] <Mithrandir> jdong: not new ones.  Old ones seem to work.  *shrug*; too late.
[03:00] <ogra> Mithrandir, how are you able to install knot3 ? is the apt fix on the CDs ?
[03:01] <jdong> Mithrandir: I know; too late :-/
[03:01] <Mithrandir> ogra: I'm about to build alternate cds now.
[03:01] <Riddell> Kamion: will do.  did you notice that the dialogue said "blah" in it when you did the last merge? :)
[03:01] <ogra> ah, cool !
[03:01] <Kamion> Riddell: I think it flicked past my consciousness but I was past caring
[03:01] <Kamion> it had probably been a long day or something :)
[03:01] <Nafallo> Mithrandir: don't forget to write it down in the release mail as a critical bug though :-P
[03:01] <Mithrandir> Nafallo: point.
[03:02] <Mithrandir> ogra: I'll tell you when you have ISOs
[03:02] <Kamion> Mithrandir: let me look into Riddell's openssh-client issue first please?
[03:02] <Mithrandir> Kamion: 'k
[03:02] <Nafallo> s/ :-P//
[03:02] <simira> hi sladen 
[03:03] <Kamion> oh, argh
[03:03] <ogra> Mithrandir, thanks :)
[03:03] <Kamion> ok, probably easy to fix, just need to upgrade germinate on lithium I think
[03:03] <Mithrandir> Kamion: so I should just kill the current cd build?
[03:03] <Kamion> yeah
[03:03] <simira> woooah
[03:03] <simira> the rainforest just stepped into my workplace...
[03:04] <Kamion> the overrides in the archive are broken, though
[03:04] <simira> *installing edgy current*
[03:04] <ogra> you have rainforest in norway ? 
[03:04] <simira> ogra: ubuntu ships it
[03:04] <simira> anywhere
[03:04] <ogra> ah
[03:04] <Kamion> we need to upgrade germinate on drescher somehow, but that's complicated with dapper-updates closed for business at the moment
[03:04] <Kamion> it'll have to wait
[03:04] <Kamion> Mithrandir: ok, you can probably build now
[03:05] <simira> ogra: this was more a sound-experience
[03:05] <ogra> ah
[03:05] <Mithrandir> Kamion: ok, building.
[03:06] <Riddell> Kamion: actually none of the recommends are installed from any seed
[03:06] <Mithrandir> robertj: just write one?
[03:07] <robertj> Mithrandir: hehe, I've got my hands busy with other pet projects ;)
[03:07] <Kamion> Riddell: yes, same reason, fixed by upgrading germinate on lithium
[03:07] <robertj> I'm very impressed that they aren't entirely crapped out though
[03:08] <ogra> Kamion, ouch, does that mean that my next CD build can be oversized again ? 
[03:08] <Kamion> ogra: yes
[03:08] <ogra> (i dont have much time left anymore today)
[03:08] <ogra> gah
[03:08] <Kamion> the same bug rendered the CDs rather small
[03:08] <ogra> sh*t
[03:08] <Kamion> ogra: I hope you didn't go "woohoo" and stuff more things into it
[03:08] <Kamion> if you did, you'd better take them out again
[03:09] <ogra> well, not yet ... there are still 20M free
[03:09] <robertj> Mithrandir: although it might not be as easy as say GmailFS because it uses a java applet to do checksumming & such before sending
[03:09] <ogra> but i was planning to
[03:09] <Kamion> Mithrandir: could you do Edubuntu alternate early so that ogra can have something to work on?
[03:09] <ogra> that'd be nice, yes :)
[03:09] <Mithrandir> Kamion: yes, I'll do it when ubuntu finishes (since it's already building)
[03:10] <Mithrandir> robertj: there's no info on their pages about what APIs they have, but I suspect they have something nicer than a web UI as they have something that integrates into windows explorer.
[03:10] <simira> Kamion: I am stuck in the installer. When I choose "Manually edit partition table" and click "next", it just seems to be loading, but nothing happens
[03:11] <simira> Kamion: current Edgy
[03:15] <tseng> Mithrandir: ping
[03:15] <Mithrandir> tseng: hi dude
[03:15] <tseng> Mithrandir: looking at ajmitch's bug about mono on a livecd not finding a config file
[03:15] <tseng>  open("/casper/filesystem.squashfs/usr/etc/mono/config",
[03:15] <tseng>           O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory)
[03:15] <tseng> do you have an idea why something would get that path?
[03:15] <tseng> is presumably the "wrong" path
[03:16] <Mithrandir> tseng: see https://launchpad.net/distros/ubuntu/+source/launchpad-integration/+bug/60071
[03:16] <Ubugtu> Malone bug 60071 in launchpad-integration "Gets confused when using --translate --pid $pid on the live cd" [Untriaged,Unconfirmed]  
[03:16] <infinity> Kamion: php5 heading to incoming is meant for post-knot (ie: don't ask if it needs approving, just leave it be)
[03:16] <Mithrandir> looks like the same bug
[03:17] <Kamion> simira: need /var/log/syslog and /var/log/partman in a bug report against partman-base, please
[03:17] <Mithrandir> tseng: it looks like an unionfs bug, really.
[03:17] <simira> Kamion: it's skipping keyboard layout as well. Something's real buggy here
[03:17] <tseng> Mithrandir: really, you think the app should be changed?
[03:18] <Kamion> simira: could be entirely disconnected
[03:18] <Kamion> simira: since we replaced the keyboard handling recently
[03:18] <Mithrandir> tseng: unsure.  I thought it was just the /proc/$pid/exe symlink which was wrong, but now I'm wondering if realpath does something weird, or something.
[03:19] <Kamion> I'd like a separate bug on console-setup about the keyboard handling, with /var/log/syslog attached
[03:19] <Mithrandir> tseng: I don't have a live cd booted here just now, but I'll investigate it
[03:21] <simira> Kamion: I'll do
[03:22] <mvirkkil> Would anyone be interested in helping me get to the bottom of the issues mentioned at https://launchpad.net/distros/ubuntu/+source/linux-source-2.6.17/+bug/60222 ? I tried #ubuntu-kernel but no-one seems to be awake.
[03:22] <Ubugtu> Malone bug 60222 in linux-source-2.6.17 "zd1211rw tries to load firmware from wrong file" [Untriaged,Unconfirmed]  
[03:22] <tseng> seb128: you get that?
[03:27] <Mithrandir> ogra: new cds
[03:28] <ogra> thanks !
[03:28] <ogra> AAAAAARRRGGGGHHHHH !!!!!!!
[03:29] <ogra> edgy-install-i386.iso            13-Sep-2006 14:20  761M 
[03:29] <jdong> lol
[03:29] <ogra> thats not even remotely solvable
[03:29] <tseng> thanks alot Mithrandir 
[03:29] <jdong> a bit oversized, ogra?
[03:29] <ogra> so i suspect no knot3 for edubuntu 
[03:29] <Mithrandir> ouch.
[03:30] <ogra> but i wonder why ...
[03:30] <ogra> it cant be that much i didnt add any stuff over dapper 
[03:30] <Kamion> simira: thanks
[03:31] <Mithrandir> I wonder why i386 + ppc seem to be oversized for Ubuntu, though.
[03:31] <ogra> apart from small things like pessulus or student-control-panel and some kb for ltspfs it shouldnt differ from dapper
[03:31] <ogra> i even dropped the linux-headers so i should have more space than ubuntu 
[03:33] <Mithrandir> Kamion: any thoughts on what we can get rid of to save 5MB on i386 and 3MB on ppc?
[03:33] <Kamion> I can save a bit by kicking out the old console-keymaps udebs
[03:33] <tkamppeter> pitti, have you already looked at my Gutenprint package? It's my first Debian package.
[03:33] <pitti> tkamppeter: oh, not yet, will do in 10 minutes, ok:?
[03:33] <tkamppeter> ok
[03:34] <pitti> tkamppeter: wow, that merge was terribly difficult for starting to learn about Debian source packages :)
[03:34] <Mithrandir> Kamion: that needs a -meta upload, doesn't it?
[03:34] <Mithrandir> or can you just twiddle the archive?
[03:35] <Kamion> Mithrandir: Mark said we could drop the book and replace it with a link to the online content. Has that been done?
[03:35] <bluefoxicy> did I miss doko
[03:35] <Kamion> Mithrandir: no, installer seed, so it's just a seed change
[03:35] <Mithrandir> Kamion: I don't think it has happened, no.
[03:35] <dholbach> Kamion: no it wasn't - it'd need changes to example-content and to gnome-panel
[03:35] <dholbach> rodarvus: we have 'telepathy' project now
[03:36] <Kamion> dholbach: how long would that take?
[03:36] <dholbach> Kamion: I'm not sure where to find the online content
[03:36] <Kamion> eventually we'll be able to drop console-data and save some space, but that's not ready yet
[03:36] <Kamion> jono: could you point dholbach at the right place? see above
[03:36] <rodarvus> dholbach, I noticed, telepathy-inspector is there already :)
[03:36] <dholbach> Kamion: example-content: easy - seb128 did the gnome-panel change, so I need to look at it first
[03:36] <dholbach> rodarvus: oh right, yes
[03:37] <Kamion> > I'm easy on both counts. I do think though that we should have SOME
[03:37] <Kamion> > multimedia content, even if it's tighter than the old 10.6MB file. If
[03:37] <Kamion> > the book is to be removed then please ensure we have prominent linkage
[03:37] <Kamion> > to it on the landing page of System->Help->Online documentation. Thanks!
[03:37] <Kamion> As it's large, I think it makes more sense to host it primarily online and
[03:37] <Kamion> link to it, rather than bundling it on the CD for Edgy.
[03:37] <Kamion> ^-- ubuntu-devel discussion, > is Mark, main paste is mdz
[03:38] <jono> Kamion, I will check into where the book is now, I am not sure
[03:38] <pitti> carlos: do you know why apport is not translatable in Rosetta? (https://launchpad.net/distros/ubuntu/edgy/+source/apport/+translations)
[03:38] <dholbach> Kamion: I can prepare the changes and wait for jono to do it
[03:38] <pitti> carlos: it has a proper .pot file which is auto-updated on build
[03:38] <jono> dholbach, do you just need to know where the online content is?
[03:38] <Kamion> ogra: is denemo really worth including on your CD?
[03:38] <carlos> pitti: I'm still working on the new .pot files for Edgy
[03:38] <carlos> I guess that's one of the missing ones
[03:38] <bluefoxicy> One day I need to get paid.
[03:38] <pitti> carlos: ok, so not a problem on my end?
[03:38] <realist> pitti: what is the best way to submit security patches for review?
[03:38] <ogra> Kamion, its a note setting app 
[03:39] <carlos> pitti: let me check to be completely sure
[03:39] <ogra> Kamion, and its quite small
[03:39] <dholbach> jono: yes, for now it is: file:///usr/share/example-content/book-toc.html
[03:39] <Kamion> yes, I know what it is, I used to maintain it in Debian
[03:39] <pitti> realist: a ML, please see https://wiki.ubuntu.com/SecurityUpdateProcedures
[03:39] <Kamion> it's a megabyte, and while it's nice, it doesn't seem required?
[03:39] <jono> dholbach, right, and in Edgy are we just pointing to it online?
[03:39] <ogra> Kamion, right, its broken as well atm ... i'l drop it
[03:39] <realist> pitti: same procedure for debdiffs?
[03:39] <dholbach> jono: no, not yet - we're about to do that change now :-)
[03:39] <ogra> but that still doesnt explain where 80M came from 
[03:39] <pitti> realist: yes, debdiffs are the prefered form of review
[03:39] <Kamion> ogra: the only way to determine that is by looking through the list
[03:40] <ogra> i'm looking through the log atm
[03:40] <jono> dholbach, right, so we need someone online to point to where the book content will live :)
[03:40] <jono> s/someone/somewhere
[03:40] <carlos> pitti: confirmed and approved
[03:41] <dholbach> 11,8M -> 8,6M
[03:43] <realist> pitti: do we submit the tested debdiff to malone?
[03:43] <ogra>  /srv/cdimage.no-name-yet.com/scratch/edubuntu/daily/tasks-previous/edgy/lamp-server
[03:43] <ogra> whats that ? 
[03:44] <ogra> err, and why does server-ship get added as well ? 
[03:44] <pitti> realist: either that or directly to the ML (the latter is easier for review)
[03:44] <dholbach> jono: exactly
[03:45] <jono> dholbach, ok, so are you guys putting it online, or should someone else?
[03:46] <dholbach> Kamion: what do you think: is people.ubuntu.com/~dholbach/book good enough for now?
[03:46] <Kamion> ogra: huh, let me check that
[03:47] <ogra> Kamion, that would explain 80MB else i'd hav no idea how it should grow by such a huge amount
[03:48] <Kamion> dholbach: when it exists, sure :-)
[03:48] <Kamion> ogra: yes, that's definitely a bug, I'll fix that - thanks
[03:48] <dholbach> Kamion: ok, cool - I'll get working on it
[03:48] <ogra> Kamion, thanks as well ... i nearly fell off my chair, that saves my day :)
[03:49] <ogra> hmm, why doesnt my metapackage talk about recommends ... i thought mvo's upload should have fixed that ...
[03:50] <ogra> mvo_, ?
[03:50] <Kamion> ogra: hmm, are you sure? lamp-server being germinated doesn't mean it's on the CD
[03:50] <ogra> hmm
[03:51] <mvo_> ogra: did you changed your seeds already? used () around the stuff that should become recommends? if so, I can have a look to see what goes wrong
[03:51] <ogra> mvo_, the recommends are in the package, but my update script keeps quiet about them 
[03:51] <Kamion> ogra: is this your first upload of edubuntu-meta with recommends?
[03:51] <ogra> yes, i changed to brackets long ago
[03:51] <ogra> Kamion, nope
[03:51] <ogra> the third or something like that
[03:51] <mvo_> ogra: so its only  missing in the changelogs? 
[03:51] <ogra> yes
[03:52] <ogra> i thought your upload was supposed to fix that
[03:52] <Kamion> there was nothing to fix
[03:52] <Kamion> it was only ever a problem on the first upload
[03:52] <ogra> hmm, k 
[03:52] <Kamion> ogra: oh, what version of germinate do you have installed?
[03:52] <ogra> they are in, so its fine for now ...
[03:52] <mvo_> ogra: my initial version was very bad at getting the changelogs, kamion fixed it in germinate 
[03:52] <ogra> Kamion, 0.22
[03:52] <Kamion> that should work
[03:53] <Kamion> oh, no, it shouldn't
[03:53] <ogra> yes, you told me that a week ago or so
[03:53] <Kamion> there was something to fix in fact :)
[03:53] <ogra> oh
[03:53] <Kamion> ogra: upgrade to 0.23
[03:53] <ogra> ah :)
[03:53] <ogra> will do
[03:53] <Kamion>   * Fix recommends-related changelog entries from update-metapackage.
[03:53] <ogra> but i suspect that still wont fix my 80MB issue
[03:53] <Kamion> if it still doesn't work, let me know
[03:54] <Mithrandir> Riddell: new kubuntu alternate images for you, FYI.
[03:54] <Kamion> lamp-server is *not* being included on the edubuntu CDs; that was a false alarm
[03:54] <Kamion> note how e.g. php5 isn't there
[03:54] <ogra> hmm, then i suspect server-ship is neither
[03:54] <Kamion> it is not
[03:55] <ogra> but there are at least 50M more than should be on the CD ...
[03:55] <Riddell> thanks Mithrandir 
[03:55] <Kamion> for instance, zope3-sandbox isn't there
[03:55] <ogra> i'd have expected an oversizedness of 10 or 15MB 
[03:55] <Kamion> ok, please stop it
[03:55] <Kamion> we *know* it's oversized, both of us are looking into it
[03:55] <ogra> yes
[03:55] <ogra> but i dont see anything suspicious
[03:56] <Kamion> if you spend less time explaining how bad it is, you'll have more time to investigate. :)
[03:56] <ogra> :)
[03:59] <dholbach> Kamion: gnome-panel is test-building
[03:59] <Kamion> I wonder why libglib1.2 is back on the Edubuntu CD
[04:02] <Kamion> gs-esp has a new dependency on libglib1.2
[04:02] <ogra> ouch
[04:02] <Kamion> not ouch, libglib1.2 is small
[04:03] <ogra> didnt we want to drop it from main ages ago ? 
[04:03] <ogra> or was that only gtk1.2
[04:03] <infinity> That was gtk1.2
[04:04] <ogra> to be honest i dont see *anything* that would cause such a huge oversizedness ...
[04:04] <pitti> ogra: we *want* to drop both, unfortunately we can't
[04:04] <tkamppeter> The newest SVN state of ESP GhostScript auto-detects which GTK is there and uses GTK2 if possible.
[04:05] <infinity> pitti: Has either of us (okay, that's a trick question, I know I haven't) looked at ReducingDuplication for edgy yet?
[04:05] <dholbach> Kamion: ok to upload gnome-panel and example-content?
[04:05] <dholbach> Mithrandir too: ^
[04:05] <tkamppeter> For what does GhostScript need libglib1.2?
[04:05] <Kamion> no idea, I just saw the dependency
[04:06] <pitti> tkamppeter: biff
[04:06] <Mithrandir> dholbach: yes.
[04:06] <Kamion> if you can look at fixing its dependency on libglib1.2, that would be good, although it isn't urgent
[04:06] <pitti> infinity: I didn't
[04:06] <Kamion> tkamppeter: ^--
[04:06] <dholbach> Mithrandir: gracias
[04:06] <pitti> infinity: the only thing I looked at is gnutls12 -> 13
[04:06] <Mithrandir> infinity: you didn't read my mind and stopped the publisher so we can sneak dholbach's stuff in, did you?
[04:06] <infinity> pitti: We should probably run our extremely unscientific scripts again soon.
[04:06] <pitti> infinity: on !{sparc,hppa} the transition is almost complete (only gutenprint missing, which is just getting merged by tkamppeter)
[04:07] <infinity> Mithrandir: Oddly, no.
[04:07] <infinity> pitti: I'd like todo a wholesale transition to db4.4 as well.
[04:07] <pitti> infinity: on ia64 and hppa half of the world still depends on 12, no idea about that
[04:07] <pitti> infinity: (when I said 'sparc' I meant 'ia64' of course)
[04:07] <fabbione> pitti: there is no hppa in edgy.. 
[04:07] <Kamion> damn, I forgot to approve ubiquity
[04:07] <infinity> pitti: Remind me to look into the sparc gnutls thing.  hppa is broken because it's not built edgy in any way (and might not do so)
[04:07] <fabbione> pitti: so just forget about that
[04:07] <Kamion> done now
[04:07] <pitti> fabbione: oh, I see
[04:07] <infinity> pitti: s/sparc/ia64/ then. :0
[04:08] <pitti> infinity: so anastacia should not look at hppa any more either
[04:08] <infinity> pitti: Yeah, we need to make all the tools ignore hppa soon, I think.
[04:08] <pitti> fabbione: amd64 works just fine :)
[04:08] <fabbione> pitti: amd64 is hibrid
[04:08] <dholbach> Mithrandir, Kamion: done
[04:09] <jdong> why are some of the restricted modules missing from the livecd?
[04:09] <jdong> fglrx is nowhere to be found
[04:11] <infinity> jdong: casper disabled building the fglrx and nvidia modules.
[04:11] <SEJeff> jdong: Doesn't that have something to do with gpl violations like what Kororaa got into
[04:11] <jdong> infinity: is it due to licensing?
[04:12] <jdong> SEJeff: yes, but when the system is installed, fglrx magically pops back in.... which is just as illegal by the kororaa phenomenon...
[04:12] <infinity> jdong: No, to cut down on memory usage.
[04:12] <jdong> infinity: ah, ok
[04:13] <jdong> infinity: is there any way to temporarily get it back?
[04:13] <infinity> jdong: There's no licensing issue, but linking those modules in a ramdisk eats precious memory.
[04:13] <jdong> infinity: I suppose apt-get install --reinstall l-r-m?
[04:13] <infinity> jdong: edit /etc/default/linux-restricted-modules
[04:13] <jdong> once I set up networking...
[04:13] <jdong> oh
[04:13] <jdong> duh
[04:13] <jdong> :)
[04:13] <jdong> thx
[04:13] <ogra> Kamion, if i compare http://people.ubuntu.com/~ogra/removals.txt with http://people.ubuntu.com/~ogra/additions.txt there should roughly be a +/-0 
[04:14] <infinity> jdong: (and then re-run the init script)
[04:14] <jdong> yep, got it :)
[04:14] <Kamion> ogra: what's that a diff between? I'm looking at quite a different diff
[04:15] <ogra> the ubuntu and edubuntu .list files
[04:15] <tkamppeter> pitti: biff
[04:15] <Kamion> I'm diffing edubuntu dapper->edgy, which I think is probably more useful
[04:15] <ogra> removals.txt is what is removed in edubuntu vs ubuntu 
[04:15] <ogra> additions.txt is the other way around
[04:16] <fabbione> hey Keybuk 
[04:16] <Keybuk> heyhey
[04:16] <fabbione> Keybuk: i was just waiting for you
[04:16] <fabbione> :)
[04:16] <Keybuk> fabbione: did you have any luck with the --debug thing?
[04:17] <tkamppeter> I have made the last Mandriva package of ESP GS with the current SVN state and by checking the binaries with ldd I see that glib2 and gtk2 is used. So probably we need to update ESP GhostScript from upstream.
[04:17] <fabbione> Keybuk: i added the output to the bug and prepared access for you
[04:17] <fabbione> Keybuk: i was waiting to give you info on how to access the *
[04:17] <Kamion> ogra: I've committed a seed change to drop linux-generic from your i386 CD, which should save you some space
[04:17] <ogra> ok
[04:17] <Kamion> well, am committing
[04:18] <ogra> Kamion, but dapper->edgy should roughly only be ltspfs/fuse and student-control-panel and deps ... plus some bytes in ltsp
[04:18] <fabbione> Keybuk: i am taking 5 minutes off and then i will give you details on how to access..
[04:18] <ogra> that cant be such a huge difference
[04:19] <Kamion> tkamppeter: we're well past the point where we are comfortable updating from upstream by default; we'll need to consider such an update rather carefully. If the problems we have can be fixed without doing that, that might be preferable
[04:19] <Keybuk> fabbione: many thanks
[04:19] <Kamion> but it depends what else is fixed by the new upstream, and what risks it entails
[04:19] <Kamion> ogra: I'm not much bothered what it *should* be; I'm looking at what it is
[04:19] <ogra> well, i now what i added
[04:20] <ogra> *know
[04:20] <Kamion> ogra: about 30/40MB of the diff is the addition of ttf-arphic-ukai, ttf-arphic-uming, ttf-baekmuk, and ttf-dustin
[04:20] <Kamion> well, -dustin is small
[04:20] <ogra> argh ...
[04:20] <Kamion> but the others are big
[04:20] <ogra> how did they end up in there agaon ...
[04:21] <Kamion> apparently you don't know what you added :P
[04:21] <ogra> ah, i see, my seed merge for recommends added them
[04:21] <ogra> they were comented since breezy :)
[04:21] <kagou> iwj: do you have you found a solution for Bug #56682 ?
[04:21] <Ubugtu> Malone bug 56682 in fontconfig "Raster fonts appear in Edgy" [Medium,Confirmed]  http://launchpad.net/bugs/56682
[04:21] <Keybuk> fabbione: from the output, it looks like you have a duplicate of jbailey's bug
[04:22] <Keybuk> it wouldn't surprise me if init is spinning at 100% CPU at that point with waitid() returning the same (dead) pid each time
[04:22] <Kamion> ogra: together with the removal of linux-generic from ship, that should get rid of basically all your oversizedness
[04:23] <ogra> yeah
[04:23] <ogra> updating the seeds now
[04:23] <tkamppeter> if getting rif of glib1.2/gtk1.2 is essentially important, we should perhaps extract a patch from the ESP GS SVN. If the unwished libs are small, we simply leave it this way in Edgy and let it simply come by itself on the way to 7.04.
[04:23] <Keybuk> fabbione: it makes some amount of curious sense that sparc and ppc64 would exhibit the same problem -- I believe the kernel code is similar?
[04:24] <fabbione> Keybuk: re.. 
[04:24] <fabbione> dunno if the code is similar..
[04:24] <iwj> kagou: No.  Well, we have a workaround.  I was just about to look at it again in fact as my ff build seems not to be dying at the first hurdle now.
[04:24] <fabbione> but one thing is sure... the same behaviour is seen in both .15 and .17
[04:24] <Keybuk> fabbione: that would be consistent with jbailey's problem too
[04:24] <Keybuk> the same behaviour was seen with the dapper and edgy kernels
[04:25] <Kamion> tkamppeter: it's relatively unimportant in itself, so if there's nothing else that points to updating gs-esp as a good thing, we can leave it
[04:25] <Kamion> Keybuk: they're kind of similar architectures at the CPU level
[04:25] <Keybuk> Kamion: indeed
[04:26] <Kamion> ogra: finally committed my seed change, so you'll need to update again; sorry, it took longer than I expected for bzr to get its act together for some reason
[04:27] <ogra> dont worry
[04:27] <ogra> thanks for the help :)
[04:27] <Kamion> np
[04:28] <pitti> tkamppeter: answered
[04:38] <doko> Mithrandir: would you mind an python2.5 upload at this time (not on the CD's)?
[04:42] <kagou> iwj: do you know if edgy is installed with the wanted of enabled bitmaps or do you want to install edgy with no bitmaps by default ?
[04:43] <infinity> doko: It'll just end up in queue/unapproved anyway.  We're actually frozen properly this time.
[04:43] <infinity> doko: So upload away, just don't be shocked if no one bothers to approve it right away.
[04:43] <doko> infinity: so it shouldn't hurt?
[04:43] <doko> ok
[04:44] <pitti> oh, that means I can also upload the fixed pkgbinarymangler already
[04:46] <infinity> pitti: Oh, what did I break? :)
[04:47] <infinity> pitti: Oh, the backports bug.  Right.
[04:48] <pitti> infinity: yes, that and some manpage fixes
[04:48] <pitti> nothing earthshaking, I'd just like to get it off my hard disk and forget about it
[04:48] <infinity> pitti: You realise this will need to be fixed for dapper, right?
[04:48] <pitti> infinity: yes, it has a dapper task
[04:48] <pitti> infinity: but dapper-updates is frozen, too
[04:49] <infinity> Yeah, I know.
[04:49] <infinity> You can upload to it, though.  We just aren't approving currently. :)
[04:50] <iwj> kagou: I think the intent is to install it with them disabled by default.
[04:51] <infinity> pitti: How are you detecting the target pocket?  Just `dpkg-parsechangelog | grep ^Distribution`?
[04:51] <pitti> infinity: yes
[04:52] <pitti> infinity: I was going to ask you for the CurrentlyBuilding format
[04:52] <pitti> infinity: but then I saw that backports have an automatic changelog entry
[04:52] <infinity> pitti: CurrentlyBuilding doesn't have a distribution in it, but it certainly could.
[04:52] <infinity> pitti: Not much point in duplicating that info, however.
[04:53] <pitti> infinity: as long as we have the changelog entry, this seems more robust to me
[04:53] <infinity> As "robust" as anything in that package is. :)
[04:53] <infinity> I fear I made it scarier with my last upload, but it *does* seem to work, so I'll not jinx it.
[04:53] <pitti> mdz: infinity and I had a short discussion about adding a checkbox to the apport-gtk dialog and a gksudo wrapper to disable apport; do you consider that an edgy+1 feature or an usability bug fix?
[04:55] <ogra> Riddell, would it hurt to drop the menu-xdg dep from kdelibs ? 
[04:55] <iwj> kagou, Kamion: After a bit of archeaology, it is now clear more or less why that weird code is there in the fontconfig-config.config but in fact we can just get rid of it.  It serves no useful purpose for us.
[04:55] <iwj> So I will remove it.
[04:57] <iwj> As in, we can get rid of all of the efforts to look at symlinks and mess with the debconf db.
[04:57] <Kamion> Keybuk: did I remember to warn you about gtk2-engines? the binaries in NEW are bogus, I think - at least it should be checked whether they're supposed to be there before accepting
[04:57] <Kamion> iwj: ah, ok, excellent
[04:57] <Keybuk> Kamion: no, but I think I already saw them last night and postponed dealing with them
[04:57] <Kamion> ok, good
[04:57] <infinity> Kamion: They're not bogus, but the queue tool's handling of them is bogus.
[04:58] <Kamion> infinity: it doesn't display them properly because they existed in dapper, yes - but those binaries were supposed to be gone ...
[04:58] <kagou> great iwj. do that mean that by default 30-debconf-no-bitmaps will be installed ?
[04:58] <Kamion> oh, are they transitional maybe?
[04:58] <infinity> Kamion: I was leaving it there as an example of bug #59291
[04:58] <Ubugtu> Malone bug 59291 in soyuz "queue tool seems confused about NEWness" [High,Confirmed]  http://launchpad.net/bugs/59291
[04:59] <Kamion> ah, they are
[04:59] <infinity> Kamion: They're new transitional arch:all binaries, it's the arch:all business that's causing the bug.
[04:59] <Kamion> hmm, I don't want to leave an upload there indefinitely until the soyuz people fix a bug
[04:59] <infinity> I wasn't planning on it being indefinite. :)
[05:00] <Kamion> I'm a bit uncomfortable with gtk2-engines being out of sync for knot-3
[05:00] <infinity> I'll put it on my "poke cprov and malcc" TODO for tomorrow.
[05:00] <infinity> Unless you want to accept them now.
[05:00] <Kamion> but I guess if it isn't actually causing breakage ...
[05:00] <infinity> I can reproduce the bug at will.
[05:00] <Kamion> seb128,dholbach: is gtk2-engines 2.8.0 important to have?
[05:00] <kagou> iwj: i don't know if it's possible, but update fontconfig package for knot3 will be great
[05:00] <seb128> Kamion: no
[05:00] <seb128> why?
[05:01] <ogra> Keybuk, ?? "...Evolution still doesn't :-/..." ???
[05:01] <ogra> i use it daily
[05:01] <Kamion> seb128: it's stuck in NEW due to the above, that's all
[05:01] <Kamion> if it's not urgent I'll leave it there
[05:01] <seb128> ah ok
[05:01] <Keybuk> ogra: doesn't obey Mail-Followup-To
[05:01] <seb128> right, no hurry
[05:02] <ogra> Keybuk, ctrl-l ?
[05:02] <Keybuk> ogra: that is not Mail-Followup-To ... that is Reply-To-List which uses the List-* headers
[05:02] <ogra> ah, k
[05:02] <ogra> sorry
[05:02] <ogra> i'm to confused today 
[05:03] <Treenaks> mail-followup-to specifies if you want a Cc on the mail-to-a-mailinglist, afaik
[05:04] <Keybuk> right, poster preference
[05:04] <infinity> Kamion: Okay, malcc's going to attack it and either find a fix, or extract enough data from the queue item that we can take away his example.
[05:04] <Riddell> ogra: why do you want to drop menu-xdg?
[05:05] <Riddell> ogra: although it's probably fine to drop
[05:06] <iwj> kagou: If I give you a .deb under the table, can you test it for me ?
[05:06] <iwj> Whether to include this change in knot-3 is up to Tollef.
[05:06] <Mithrandir> iwj: which bug?
[05:06] <iwj> Ah, hello :-).
[05:07] <iwj> bug 56682
[05:07] <Ubugtu> Malone bug 56682 in fontconfig "Raster fonts appear in Edgy" [Medium,Confirmed]  http://launchpad.net/bugs/56682
[05:08] <iwj> I can explain more if you like.
[05:08] <Mithrandir> iwj: while I see that it is/can be annoying, I don't think it's worth holding up knot 3 for.
[05:08] <iwj> Fair enough.
[05:08] <Mithrandir> iwj: if you think otherwise, feel free to convince me. :-)
[05:09] <iwj> I'll upload after the release announcement then.
[05:09] <Mithrandir> you can upload it now, it'll just get stuck in the unapproved queue
[05:09] <iwj> I don't have an opinion really.  I'm not one to hanker after prettiness :-).
[05:09] <iwj> Oh, good.
[05:09] <iwj> There was a note in your knot-3 message saying not to.
[05:09] <Mithrandir> yes, since FROZEN was broken for knot 2 and we didn't know if it was fixed or not.
[05:09] <Mithrandir> (actually, it wasn't fixed until today, UTC)
[05:10] <iwj> Ah, but it is fixed now.  Excellent.
[05:10] <Mithrandir> yup
[05:10] <iwj> kagou: I'd still like to get this off my radar, so if you'd like to test it so I can upload then that would be very helpful.
[05:11] <Riddell> hmm, usplash broken on kubuntu alternate CD and booting without usplash X doesn't seem to start
[05:11] <ogra> Riddell, it breaks gnomes menu if the menu package isnt installed (which i surely dont want)
[05:14] <kagou> iwj: i'm, of course, ready to test it
[05:14] <Kamion> infinity: great, thanks
[05:16] <Kamion> Riddell: broken how?
[05:16] <Mithrandir> Riddell: removing splash from the command line doesn't fix it either?
[05:17] <iwj> kagou: http://www.chiark.greenend.org.uk/~ian/d/fontconfig/
[05:17] <iwj> It may not fix a system that was already broken, though.
[05:17] <iwj> Because the broken system has the wrong values in debconf.
[05:18] <iwj> So to test it you should clear out the relevant debconf data and purge the fontconfig packages, and then install these.
[05:18] <Kamion> as long as (a) dpkg-reconfigure fixes it and (b) dapper->edgy upgrades DTRT then that's probably ok
[05:18] <iwj> dpkg-reconfigure will fix it and I'm confident about the dapper upgrades.
[05:18] <iwj> kagou: You can confirm for me whether installing that and then saying dpkg-reconfigure fixes it :-).
[05:19] <crispin> pitti: are you around ?
[05:19] <pitti> crispin: yes
[05:19] <tkamppeter> pitti: biff
[05:19] <tkamppeter> pitti: New gutenprint uploaded
[05:19] <crispin> pitti: ahh, excellent, I was wondering how to best go about debugging bug #55809 ? I can repro all the time on my laptop and would dearly love to fix it (and am willing to put in the time this weekend to do so)
[05:19] <Ubugtu> Malone bug 55809 in hal "HAL changes wireless interface (net.80211) to wired (net.80203) in info.category after suspend" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/55809
[05:20] <infinity> mvo: Can I have you in ##soyuz1.0 for a few minutes?
[05:21] <HiddenWolf> dholbach: can you check if https://launchpad.net/distros/ubuntu/+source/gstreamer0.10-pitfdll/+bug/59948 contains what you need? 
[05:21] <Ubugtu> Malone bug 59948 in gstreamer0.10-pitfdll "Crash in totem-thumbnailer" [Low,Needs info]  
[05:22] <Riddell> Kamion: blank screen, nothing happens
[05:22] <Riddell> Mithrandir: removing splash from command line boots up into command line, need to start X manuall
[05:23] <janimo> tkamppeter: what was the fix for foomatic-db and duplicate printer description files for two epson stylus models? It should go in ubuntu as a patch if the whole package is not synced with upstream
[05:23] <pitti> crispin: hm, the log does not show anything useful at first sight
[05:23] <Mithrandir> Riddell: uh, so kdm doesn't start for some reason?
[05:23] <pitti> crispin: I guess this requires a proper gdb session on hald
[05:24] <Riddell> Mithrandir: seems to be the case
[05:24] <crispin> pitti: so build up a debug hald, find where it detects the network driver, then put a break point in and suspend and resume and see what goes on ?
[05:24] <simira> Kamion: I finally managed to report a bug. I guess a partman-log of > 2MB points to something being wrong...
[05:24] <Riddell> Mithrandir: it is in /etc/rc2.d/ and it does start from the live CD I installed today
[05:24] <pitti> crispin: yes, that's the broad description
[05:25] <pitti> crispin: you should stop the system hald (kill or /etc/dbus/event.d/20hal stop) before
[05:25] <crispin> pitti: nod
[05:25] <smurf> Kamion: altgr keys are penalized so it should only ask for one of those if it can't find anything better
[05:25] <pitti> crispin: then DEB_BUILD_OPTIONS=nostrip,noopt debian/rules build in the source package
[05:25] <crispin> pitti: I'll give that a go sometime over the weekend
[05:25] <pitti> crispin: I put a run-hald.sh script into debian/ for convenience
[05:26] <pitti> crispin: it runs the locally built hald with all necessary paths adapted for the stuff in the source package (rather than installed system files)
[05:26] <crispin> ahh nice
[05:26] <pitti> crispin: TIA for debugging
[05:26] <iwj> dpkg-deb: building package `firefox' in `../firefox_1.99+2.0b2+dfsg-1ubuntu1_i386.deb'.
[05:26] <iwj> Scary.
[05:26] <crispin> pitti: I'll go for a gdb or printf session and try tracking things down then :-)
[05:27] <smurf> Kamion: one quick and dirty solution would be to add "" to the aliases for "a"; but whether it would find a better alternate is anybody's guess
[05:27] <Kamion> smurf: something's wrong with the algorithm then -  is the obvious distinguisher between gb and us and it's not trying that
[05:27] <smurf>  is where on a gb keyboard?
[05:27] <Kamion> smurf: shift-3
[05:27] <Kamion> and nowhere on us
[05:27] <_ion> Where is # then?
[05:27] <Kamion> _ion: where us has \
[05:28] <_ion> Where is \ then? ;-)
[05:28] <kagou> iwj: testing your packages... they create 30-debconf-yes-bitmaps.conf. And i said that's not good.
[05:28] <Kamion> _ion: left of z
[05:28] <Keybuk> _ion: on the # key :p
[05:28] <iwj> kagou: Yes, they would do, if you just install them on an affected install.
[05:28] <_ion> kamion: Ok. :-)
[05:28] <Keybuk> Kamion: uhh, # is left of ENTER on a sensible UK keyboard
[05:28] <iwj> Can you try dpkg-reconfigure fontconfig-config ?
[05:28] <Keybuk> oh, sorry, you were talking about \
[05:28] <Kamion> Keybuk: yes, that's what I said :)
[05:28] <iwj> config config font config reconf conf conf figure
[05:29] <iwj> I didn't have any boots beforehand.
[05:29] <jdong|laptop> Keybuk: how come filelist-order in readahead-list is not installed onto the system?
[05:29] <kagou> iwj: how can i test them like a fresh install ?!
[05:30] <Keybuk> jdong|laptop: because it was silly
[05:30] <jdong|laptop> Keybuk: you almost waste 25 minutes of my time rewriting that program :P
[05:30] <jdong|laptop> Keybuk: does it not work?
[05:30] <iwj> kagou: That's tough.  dpkg-reconfigure throws away what I think is the infection and is the first step, so can you try that ?
[05:30] <Keybuk> jdong|laptop: oh, I suspect it works
[05:30] <iwj> kagou: If that fails then it's buggy.
[05:30] <Keybuk> jdong|laptop: it was just easier to make readahead-watch do that
[05:30] <jdong|laptop> Keybuk: well, if you put a filelist-order call in S01readahead, that solves the reordering problem
[05:31] <jdong|laptop> and you're right, it does happen in nanoseconds :)
[05:31] <iwj> kagou: If that works then there's a more extensive test which involves clearing the relevant debconf entries, purging the packages, and then dpkg -i'ing them.
[05:31] <smurf> Kamion:  is probably in the alias list as too-easy-to-confuse with lLiI
[05:31] <smurf> one moment
[05:31] <Keybuk> jdong|laptop: tbh, why not just put the ioctl code in readahead-list itself?
[05:31] <Keybuk> that way you save a fork/exec
[05:31] <Kamion> smurf: seriously? the only people with  on their keyboards know what it is :)
[05:31] <Keybuk> and don't need a writable filesystem :p
[05:31] <jdong|laptop> Keybuk: that works too. :)
[05:31] <Kamion> I don't think it looks all that much like L
[05:31] <Kamion> (any more)
[05:32] <_ion> Doing that in readahead-list would be nice.
[05:32] <Kamion> anyway, no, it's not - just ""
[05:33] <smurf> Kamion: so it is ... I'll have to check the code more closely then
[05:35] <smurf> Kamion: I'm on vacation for the next eight days, but I'll take the laptop and will check
[05:35] <Kamion> smurf: thanks
[05:35] <kagou> iwj: i'm sorry, i don't understand. "sudo dpkg-reconfigure fontconfig-config" ask severals questions. Saying yes or no at "do you want bitmaps" is ok.
[05:35] <kagou> is that what you want ?
[05:36] <iwj> dpkg-reconfigure --default-priority should avoid the questions.
[05:36] <Kamion> smurf: failing that I can probably hack something up :)
[05:36] <iwj> And take the default.
[05:36] <kagou> ok so iwj in this case it's ok for me
[05:37] <kagou> i mean, it's work fine
[05:37] <Kamion> smurf: I think adding /a would probably cause problems for the Norwegians or Danish or whoever it is that has 
[05:37] <kagou> s/goog/good
[05:37] <Nafallo> danish
[05:37] <iwj> kagou: Thanks.  If you want a more thorough test it's quite tough.
[05:38] <Kamion> simira: I can reproduce the keyboard problem, anyway
[05:39] <simira> Kamion: it's the partition problem that keeps me from installing, atm. I am downloading the alternate one now.
[05:39] <smurf> Kamion: They do have  or , but one of those will probably be the next candidate :-/
[05:39] <Kamion> simira: yeah, I'm just going one step at a time in case there are knock-on problems
[05:39] <Kamion> smurf: right, and both of those look like o by the exact same argument ...
[05:39] <simira> Kamion: but it's somewhat annoying, and the correct norwegian layout isn't available in the layout-list, it seems
[05:39] <iwj> I wasn't expecting this ff build to work !
[05:40] <simira> Kamion: no complaints, I know you're doing your best and then some
[05:40] <kagou> iwj: it's ok for me. do you think that you can include your version on knot3 ?
[05:40] <iwj> kagou: I asked Mithrandir and he'd rather not, so sorry.
[05:41] <Kamion> simira: current images have a broken version of ubiquity
[05:41] <Kamion> simira: because they're out of date
[05:41] <simira> Kamion: the alternate one also? So I need't bother to test it?
[05:41] <siretart> is universe frozen as well?
[05:41] <Kamion> should be fixed next time Tollef does a build - I think all the underlying issues are fixed
[05:41] <simira> *sigh*
[05:41] <simira> and he just left... well, off to dapper, then...
[05:41] <Kamion> simira: I think very current alternate will work
[05:42] <Kamion> simira: also
[05:42] <simira> Kamion: ok, I'll test current, then
[05:42] <simira> current alternate, even
[05:42] <Kamion> simira: try 'sudo apt-get update; sudo apt-get install ubiquity' on a live CD
[05:42] <Kamion> no shame in upgrading it on the fly :-)
[05:43] <kagou> iwj: thanks for your work. i mark the bug as "in progress"
[05:43] <Keybuk> fabbione: ok, disk1 works now
[05:43] <Kamion> simira: having done that, it does offer me *a* Norwegian layout. Can you tell me what's incorrect about it?
[05:44] <simira> Kamion: everything. It still seems to run with a US layout, to me... I'll try again
[05:44] <iwj> kagou: Right.  Well, I'll upload now and it'll hopefully be deployed in due course.
[05:44] <Kamion> simira: on the live CD, it's working for me after upgrading ubiquity
[05:45] <Kamion> I get   etc. anyway
[05:45] <simira> Kamion: lucky you. May I have yours, then?
[05:45] <Kamion> and I can get into manual partitioning
[05:46] <simira> Kamion: I get  -s' instead of , actually
[05:46] <roico> is the common customizations spec going to be implemented by edgy?
[05:46] <Kamion> ogra: because all uploads require manual approval and you didn't ping me about it :)
[05:47] <simira> Kamion: uhm... now my letters are _completely_ mesed up. I am not sure what layout it is, but it's neither US or NO, or qwerty at all
[05:47] <ogra> ah, ok
[05:47] <ogra> Kamion, ping :)
[05:47] <Kamion> accepted
[05:47] <ogra> heh
[05:47] <Kamion> simira: what did you select?
[05:47] <Kamion> I need to know *exactly* what you did, from the start
[05:47] <simira> Kamion: the only Norwegian alternative, Norwegian -> Dvorak
[05:47] <Kamion> simira: oh, in the alternate CD?
[05:48] <Kamion> that would have been a good thing to tell me first :)
[05:48] <simira> Kamion: no, current live, still
[05:48] <Kamion> uh
[05:48] <simira> I found them!
[05:48] <Kamion> what are you doing that presents Dvorak?
[05:48] <Kamion> what exactly are you running?
[05:48] <siretart> Kamion: so even universe uploads need approval?
[05:48] <Kamion> siretart: yeah, but I wave those through when I notice them
[05:48] <siretart> ah. I see
[05:49] <simira> Kamion: changing keyboard layout on the menu System -> Preferences -> Keyboard, it presents Norwegian -> Dvorak(messy), Eliminate dead keys (works), Samii and Samii eliminate dead keys
[05:49] <Kamion> I just approved the outstanding ones
[05:49] <Kamion> simira: oh, that menu has nothing to do with me
[05:50] <Kamion> simira: try just picking "Norwegian", not any of the sub-entries
[05:50] <tkamppeter> pitti: biff
[05:50] <Kamion> it's not obvious that you can do that
[05:50] <roico> is the common customizations spec going to be implemented by edgy?
[05:50] <Kamion> roico: probably not anything much more than has already been done, as we're past feature freeze now
[05:51] <Kamion> than whatever has already been done; I haven't been following the details of that spec
[05:51] <pitti> tkamppeter: oh, I didn't suggest to close the bugs right now, but just to collect some which will be definitively fixed (there are some which are already marked so)
[05:52] <pitti> tkamppeter: but don't worry about it too much, closing them after upload after checking with the reporter is fine
[05:52] <Kamion> simira: I still need that /var/log/partman, if you can; feel free to gzip it first
[05:53] <simira> Kamion: should be added to the bug
[05:53] <jdub> go edgy
[05:53] <jdub> it's your birthday
[05:53] <jdub> go edgy
[05:53] <jdub> it's your birthday
[05:53] <ogra> jdub, still more than a month ...
[05:54] <jdub> LIES
[05:54] <Kamion> oh, malone's attachment UI changed and there now isn't a comment for each attachment
[05:54] <ogra> :)
[05:54] <jdong|laptop> :)
[05:54] <simira> Kamion: yes, it fooled me too, for a minute
[05:54] <hunger> Keybuk: What was the trick again to make cryptdisks work with upstart again?
[05:54] <tkamppeter> janimo, was it the Epson Stylus C65/C68? I have fixed this upstream.
[05:55] <jdub> hunger: -> #nsa
[05:55] <ogra> haha
[05:55] <jdong|laptop> lol
[05:55] <jdong|laptop> hunger: I don't think it's sudo apt-get install sysvinit.... ;)
[05:55] <Kamion> hunger: last time you asked somebody said 'console owner'
[05:55] <Keybuk> hunger: fix the cryptdisks script to do </dev/console not </dev/tty :p
[05:55] <simira> Kamion: I still don't get into manual partitioning on the current live, though
[05:56] <hunger> jdong|laptop: Actually that is what I am doing right now.
[05:56] <tkamppeter> I have also fixed bug #16703 upstream in foomatic-db.
[05:56] <Ubugtu> Malone bug 16703 in foomatic-db "hpijs is recommended for hp laserjet 1100A but ljet4 prints better" [Medium,Confirmed]  http://launchpad.net/bugs/16703
[05:56] <jdong|laptop> hunger: shhhh! Keybuk is gonna go ballistic on me... ;)
[05:56] <hunger> Keybuk: I do not get any output whatsoever... do I need to add console owner somewhere?
[05:56] <Kamion> simira: yeah, known bug, I'm just looking for the number ...
[05:56] <Keybuk> hunger: if you like output
[05:56] <Keybuk> jdong|laptop: for?
[05:56] <jdong|laptop> suggesting sysvinit :)
[05:57] <hunger> Keybuk: Well, I need to see cryptsetup ask for its password;-)
[05:57] <Keybuk> hunger: so fix the cryptsetup script
[05:58] <Keybuk> it'd be a useful patch if you also added usplash support
[05:58] <Keybuk> that has an INPUT command now
[05:58] <hunger> Keybuk: I have a somewhat different cryptsetup script:-)
[05:58] <Keybuk> you could fix that too :p
[05:59] <hunger> Keybuk: See bug #563 about my init script:-) Did not get round to update it for this release round (did not make it into breezy or dapper, so my motivation is somewhat low;-)
[05:59] <Ubugtu> Malone bug 563 in cryptsetup "improved start/stop script" [Wishlist,Needs info]  http://launchpad.net/bugs/563
[05:59] <dholbach> Keybuk: there's a new release of libjingle and new packaging for it - I'll have a look at it before pestering you with it again - i just really really hope we can get it in
[06:00] <Keybuk> dholbach: I saw
[06:00] <ogra> libjingle ? 
[06:00] <Keybuk> ogra: Google Talk
[06:00] <ogra> does it play advertisements  ?=
[06:00] <ogra> ah
[06:00] <hunger> Keybuk: When will you have upstart mount the disks? At that point it would probably make sense to integrate cryptdisks into upstart instead of trying to improve the init-script.
[06:00] <Keybuk> hunger: edgy+1
[06:01] <hunger> Keybuk: Oh, OK, I'll just stick with sysvinit till then.
[06:02] <Seveas> hunger, edgy is "sysvinit as /sbin/int", edgy+1 is "replace all intiscripts kthxbye" and edgy+2 will be "DIE SYSVINIT DIE!!"
[06:02] <desrt> is the startup screen in edgy supposed to be completely screwed up?
[06:02] <mdz> pitti: I think it's appropriate to allow the user to disable it
[06:02] <Seveas> desrt, no
[06:03] <mdz> pitti: if it's simple and quick, it can go into edgy with a FF exception
[06:03] <desrt> http://desrt.mcmaster.ca/random/bad_startup.jpeg
[06:03] <desrt> ^ this is what the bootup screen looks like for me
[06:03] <desrt> the old test pattern was fine
[06:04] <jdong|laptop> desrt: your server is down
[06:04] <desrt> erm.  wfm?
[06:04] <janimo> tkamppeter: yes the C65/C86 issue. I asked what the fix was in case a new version of the package does not make it to edgy, we should pick that oartilcuar fix at least
[06:04] <desrt> jdong|laptop; your problem, fairly sure
[06:04] <Seveas> desrt, widescreen?
[06:04] <desrt> Seveas; yup.  macbook 13.3
[06:04] <Keybuk> Mithrandir: ping?
[06:05] <Seveas> ah, mac
[06:05] <desrt> Seveas; the test pattern worked fine
[06:05] <Seveas> maybe the bogl backend is borking
[06:05] <Seveas> desrt, did you upgrade usplash today?
[06:05] <desrt> well... the image looks to be correct but started at the wrong location
[06:05] <desrt> Seveas; last night, i think
[06:06] <jdong|laptop> desrt: heh, dns doesn't resolve... might be my problem :-/
[06:06] <desrt> hang on for a version tag
[06:06] <desrt> jdong|laptop; definitely your problem
[06:06] <desrt> the dns is hosted by mcmaster university... quite redundantly
[06:06] <Kamion> Seveas: macbook => i386 => svga surely
[06:07] <Kamion> Seveas: the powerpc Macs aren't called "macbook"
[06:07] <Seveas> Kamion, *headdesk* -- I keep forgetting that
[06:07] <jdong|laptop> desrt: yeah, my problem... MITNet can access it just fine
[06:07] <desrt> "Fix incorrect width/height used by usplash_put_part in BOGL backend."
[06:07] <bddebian> Howdy folks
[06:07] <desrt> i bet that's what caused the bug
[06:07] <Kamion> desrt: nothing to do with you
[06:07] <Seveas> no
[06:07] <Kamion> desrt: you're not using the bogl backend
[06:07] <desrt> oh.
[06:07] <desrt> in any case i'm running 0.4-24
[06:08] <Kamion> desrt: and that only affected the progress bar anyway, not the whole picture
[06:08] <Kamion> you're in totally the wrong video mode by the looks of things
[06:08] <Seveas> Kamion, the whole picture is put on screen by usplash_put
[06:08] <Kamion> Seveas: yes, I know
[06:08] <Kamion> desrt: you can probably work around the problem by changing the resolution in /etc/usplash.conf
[06:08] <desrt> Kamion; well.  i'm willing to help if you have questions
[06:08] <pitti> mdz: the backend is already there (/etc/default/apport), it just doesn't have a GUI
[06:08] <desrt> Kamion; dpkg-reconfigure my image?
[06:09] <Kamion> desrt: just edit it
[06:09] <desrt> how does usplash read the file before the root filesystem is mounted?
[06:09] <Kamion> I bet it's 640x480 at the moment; change it to whatever your X resolution is
[06:09] <Kamion> desrt: after you edit the file, run update-initramfs -u
[06:09] <Kamion> that copies it into the initramfs
[06:09] <desrt> Kamion; what's what i meant by dpkg-reconfigure :p
[06:09] <tseng> desrt: usplash stuff is in initramfs
[06:09] <Kamion> desrt: dpkg-reconfigure won't do what you want in this case.
[06:10] <desrt> Kamion; it rebuilds initramfs
[06:10] <Kamion> the postinst won't rewrite an existing configuration file.
[06:10] <Kamion> talking about it is totally misleading
[06:10] <mdz> pitti: ok then
[06:10] <desrt> oh.  that's pretty weird/lame
[06:10] <Kamion> it's pretty damn correct :)
[06:10] <Kamion> pissing about with existing configuration isn't good
[06:10] <mdz> well, sometimes it is
[06:11] <Kamion> yeah, but probably not in this case
[06:11] <desrt> k.  it's going
[06:11] <desrt> please stand by
[06:12] <desrt> that sort of fixed the problem
[06:13] <jdong|laptop> Seveas: regarding your blog post, I certainly hope I won't be using a distro codenamed "fuzzy ferret" :)
[06:13] <Seveas> ;)
[06:14] <Nafallo> lol
[06:15] <desrt> Kamion; http://desrt.mcmaster.ca/random/bad_boot_2.jpeg
[06:15] <desrt> fills the screen now, but the aspect ratio is awful
[06:15] <desrt> skinny-tall-buntu: linux for people who aren't fat
[06:15] <Kamion> that I don't know about, and it could just be an "adjust your monitor" thing
[06:16] <Kamion> modern monitors often being per-mode adjustable
[06:16] <desrt> uh
[06:16] <Seveas> Kamion, no, it's the "your theme has only 4:3 and 16:9 images and your monior is neither" thing that causes this
[06:16] <desrt> it's a laptop LCD screen :p
[06:16] <desrt> i guess it picked 16:9
[06:16] <desrt> then scaled it
[06:16] <desrt> better to pick 16:9 then crop it, imo
[06:17] <Seveas> it does neither
[06:17] <Seveas> the scaling is done by your monitor
[06:17] <pygi> sivang, ?!
[06:17] <desrt> Seveas; it's running in the native resolution of my monitor
[06:17] <desrt> 1280x800
[06:17] <Seveas> no
[06:17] <Seveas> it's running in the svga mode closest to it
[06:18] <desrt> oh.  this is bad.
[06:18] <Seveas> and svga as used by usplash only does 4:3
[06:18] <Nafallo> 16:10 then :-)
[06:18] <desrt> Seveas; if that were the case then it would be stretched out widthwise
[06:18] <Seveas> but there are images that are 16:9 but scaled to 4:3 in the standard themes
[06:18] <desrt> not heightwise
[06:18] <Seveas> and usplash will use those for 1280x800
[06:19] <desrt> oh.  ok.  i get it.
[06:19] <desrt> so you have intentionally tall images
[06:19] <desrt> if a 'widescreen' reso is detected
[06:19] <desrt> because you know that the laptop will stretch them back out again
[06:19] <Seveas> indeed
[06:19] <desrt> but my laptop doesn't stretch far enough
[06:19] <desrt> that's clever -- but very evil :)
[06:19] <Seveas> indeed again
[06:20] <sladen> Seveas: most of the manufactuers include extra widescreen VESA modes (if it's a laptop)
[06:20] <desrt> sladen; 915resolution says otherwise :)
[06:20] <Seveas> theoretically it could support 16:10 and 14:9 too, but that really needs someting generic in usplash 
[06:20] <Seveas> which will NOT be edgy material 
[06:20] <desrt> Seveas; just so we're clear -- you have to fix this bug :p
[06:20] <Seveas> yes
[06:20] <desrt> it's damn ugly :)
[06:21] <Seveas> it's not that bad
[06:22] <sladen> desrt: 915resolution has nothing to do with VESA modes
[06:23] <desrt> sladen; i see.
[06:23] <desrt> Seveas; i guess one workaround would be for me to tell usplash that i have 1024x768
[06:23] <desrt> then i'll get a wide image instead of a tall one
[06:23] <Seveas> indeed
[06:24] <desrt> which will probably look a bit better :p
[06:24] <sladen> desrt: if your X is running at 1280x800 or some such, then usplash should be able to pick 1024x768 and hope that it's being scaled to 1280x800 and adjust (squeeze) the video inversely
[06:25] <desrt> sladen; right.  i want to shortcircuit that intelligence and have it use 1024x768 directly (thus avoiding the squeeze)
[06:26] <desrt> ya.  looks a bit better wide instead of tall :)
[06:27] <sladen> desrt: do you /have/ an 1280x800 (or whatever) VESA mode?
[06:27] <desrt> sladen; i'd imagine not?
[06:27] <sladen> desrt: then you're not going to get it without scaling...
[06:27] <desrt> sladen; but a 3rd scaling factor could be added
[06:28] <sladen> desrt: could be.  they'd all just get scaled together
[06:28] <desrt> or better yet, the installer could do custom scaling at ramfs-image-build-time
[06:31] <jdong|laptop> infinity: I'm getting lots of reports in backports land about md5sum mismatches
[06:31] <Nafallo> jdong: from more mirrors than se.a.u.c? :-)
[06:31] <jdong|laptop> lol, no this time from archive, us.archive, ca.archive
[06:31] <jdong|laptop> which makes me more worried
[06:31] <jdong|laptop> most recent report is 14 minutes ago
[06:32] <jdong|laptop> most aged report is around 10 hours ago
[06:32] <Seveas> desrt, interesting idea
[06:32] <jdong|laptop> and I don't currently have a dapper box handy
[06:32] <Nafallo> I had edgy main from se.archive.u.c around 10:45 UTC :-)
[06:33] <desrt> Seveas; i'm just chock full of interesting ideas that i want other people to do the work for :)
[06:33] <jamadagni> hello is mithrandir herer?
[06:33] <Seveas> hehe
[06:33] <Nafallo> I wget'd, bunzip2'd and mv'd the file into place :-P
[06:39] <jbailey> fabbione: Around?
[06:40] <Kamion> is anyone here a French translator? I'd appreciate it if somebody could fix bug 60064 in Rosetta
[06:40] <Ubugtu> Malone bug 60064 in ubiquity "Translation problem in french" [Low,Confirmed]  http://launchpad.net/bugs/60064
[06:40] <jbailey> fabbione: I was just looking at the glibc on sparc packaging problem you found.  Mostly I'm amazed that this bug hasn't tripped someone up before.
[06:41] <jbailey> fabbione: If you want amusement, read debian/rules.d/debhelper.mk, line 161.  The bit that starts with: #Ugly kludge: =)
[06:41] <Seveas> Kamion, asking in #ubuntu-fr may help
[06:45] <Kamion> Seveas: thanks, I've tried
[06:45] <Kamion> Seveas: looks like a straight user channel though
[06:48] <Seveas> vuntz is membr of -l10n-fr
[06:51] <bddebian> Kamion: Are LP sync requests still OK or are we supposed to use the new sync script?
[06:52] <bddebian> jbailey: Doesn't that describe most of glibc? ;-P
[06:53] <jbailey> bddebian: Well, this is the packaging in particular.  Glibc itself is remarkably elegant in a number of ways.  The build system is "remarkable" in other ways... ;)
[06:53] <bddebian> Heh
[06:54] <Kamion> bddebian: any sync script you may have seen is probably just a convenient way to file bugs in LP
[06:54] <bddebian> Kamion: OK, fair enough, thx.  We can still request for Universe right?
[06:54] <Kamion> bddebian: filing bugs in LP is fine; just give all the information requested by Scott on ubuntu-devel-announce
[06:54] <Kamion> bddebian: yes
[06:58] <zyga> re
[07:06] <bddebian> Kamion: I don't get u-devel-announce.  Do you happen to have a link?
[07:06] <tseng> bddebian: search gmane
[07:06] <tseng> bddebian: its low traffic
[07:07] <tseng> it was sent in August
[07:08] <bddebian> gmane?
[07:08] <jdong> no... BenC... don't go... I was gonna bug you about my USB chipset!
[07:09] <dsas> bddebian: gmane.org lists available in nntp/rss/html
[07:13] <bddebian> Hmm, I see the merge policy one
[07:13] <tkamppeter> janimo, here is the C65/C68 fix in foomatic-db:
[07:13] <tkamppeter> http://bzr.linuxprinting.org/devel/foomatic-db?cmd=revision;revid=pqm%40freestandards.org-20060906191142-7a8aa326766b71c7
[07:14] <janimo> tkamppeter: thanks
[07:14] <tkamppeter> and here the fix for the recommended driver of the LJ1100:
[07:14] <tkamppeter> http://bzr.linuxprinting.org/devel/foomatic-db?cmd=revision;revid=pqm%40freestandards.org-20060913115808-f5bd6515e31cc7d7
[07:21] <Riddell> Kamion: could you set off a new kubuntu livefs build
[07:23] <janimo> tkamppeter: is the second url (LJ1100) addressed to me as well? I don;t remember anything about that :)
[07:25] <mxpxpod> desrt: ping
[07:26] <tkamppeter> The second is bug #16703, should you be the one in charge of making a new foomatic-db package, then you should fix both.
[07:26] <Ubugtu> Malone bug 16703 in foomatic-db "hpijs is recommended for hp laserjet 1100A but ljet4 prints better" [Medium,Confirmed]  http://launchpad.net/bugs/16703
[07:27] <tkamppeter> bug #1
[07:27] <Ubugtu> Malone bug 1 in ubuntu-meta "Microsoft has a majority market share" [Critical,Confirmed]  http://launchpad.net/bugs/1
[07:28] <tkamppeter> That was only a test, it seems that every time when someone writes "bug #XXX" then ubugtu posts the link and the subject/
[07:28] <dholbach> tkamppeter: it works for upstream bugs too
[07:28] <dholbach> gnome bug 325375
[07:28] <Ubugtu> Gnome bug 325375 in Mailer "critical warning crasher : bonobo_ui_component_set_prop: assertion" [Critical,Resolved: fixed]  http://bugzilla.gnome.org/show_bug.cgi?id=325375
[07:29] <tkamppeter> mandriva bug 25186
[07:29] <Ubugtu> Malone bug 25186 in norwegian "norwegian: new changes from Debian require merging" [Medium,Fix released]  http://launchpad.net/bugs/25186
[07:29] <tkamppeter> ubugtu has a bug!
[07:29] <dholbach> lalalala, Seveas is the master of Ubugtu
[07:30] <mxpxpod> bah, do we really need to keep track of mandriva? ;)
[07:30] <dholbach> mxpxpod: when we link bugs to their bug tracker, etc - sure :)
[07:30] <mxpxpod> oh, right...
[07:30] <mxpxpod> is there any way to find out if a package is scheduled for a re-build?
[07:34] <licio> why ubuntu doesn't have mailx package?
[07:36] <dsas> licio: It does, http://packages.ubuntu.com/dapper/mail/mailx
[07:40] <licio> dsas, I know, but mailx should be required by ubuntu-desktop
[07:42] <Seveas> tkamppeter, what's the mandriva BTS?
[07:42] <dsas> licio: Probably something to do with it requiring postfix and ubuntu having a "no ports by default policy", though that's a bit of a guess.
[07:42] <Kamion> licio: no, we removed that because we aren't shipping an MTA by default so mailx is pointless
[07:42] <desrt> mxpxpod; pong
[07:43] <desrt> or should i say... mxpxpong ha ha ha
[07:43] <licio> Kamion, uhm.. thanks :)
[07:45] <tkamppeter> Seveas, it is bugzilla, site is http://qa.mandriva.com/
[07:45] <Seveas> @bugtracker add mandriva bugzilla http://qa.mandriva.com/ Mandriva
[07:46] <Seveas> mandriva bug 25186
[07:46] <Ubugtu> Mandriva bug 25186 in cups "Printer Brother HL-5140 does not work anymore with this version of cups" [Normal,Needinfo]  http://qa.mandriva.com/show_bug.cgi?id=25186
[07:49] <tkamppeter> Thank you. Does it also work with the custom-made bug tracker of CUPS: http://www.cups.org/str.php?
[07:50] <mdz> Seveas: yay
[07:50] <mdz> tkamppeter: is mandriva's bug tracker registered in launchpad as well?
[07:50] <tkamppeter> I think, not, but I do not really know.
[07:50] <Seveas> mdz, I was going to say "just tell me which bugtackers should be added" but the launchpad comment made me think of something more evil.
[07:51] <mdz> tkamppeter: https://launchpad.net/malone/bugtrackers
[07:51] <janimo> xfce bug 1737
[07:51] <Ubugtu> Malone bug 1737 in malone "Can't dissociate a bug tracker from a project" [Medium,In progress]  http://launchpad.net/bugs/1737
[07:51] <tkamppeter> mdx, know, it is not there.
[07:51] <mdz> tkamppeter: please add it; then you can link from malone bugs to mandriva bugs and have status tracked
[07:52] <Seveas> tkamppeter, the cups bugtracker is weird but should be usable
[07:52] <janimo> @bugtracker add  xfce bugzilla http://bugzilla.xfce.org/ 
[07:52] <janimo> xfce bug 1737
[07:52] <Ubugtu> xfce bug 1737 in general "Keyboard Shortcuts - Window key maps to different "Keys"" [Normal,Resolved: fixed]  http://bugzilla.xfce.org/show_bug.cgi?id=1737
[07:52] <Seveas> @bugtracker add xfce bugzilla http://bugzilla.xfce.org/ XFCE
[07:52] <janimo> \o/
[07:52] <Seveas> last argument is a description
[07:52] <janimo> seems it is optional :)
[07:52] <Seveas> it is
[07:53] <janimo> so is it now permanently added to ubugtu?
[07:53] <Seveas> yes
[07:53] <Seveas> @quit
[07:54] <Seveas> (hand-hacking config)
[07:54] <janimo> I wonder what bug number in Microsofts bugtracker 'Squash Ubuntu' is
[07:57] <Keybuk> I doubt they even think we've taken a single user off them
[08:02] <zul> janimo: its number 2
[08:02] <pitti> bah, why must copying with nautilus be so sloooooowww
[08:02] <ogra> Kamion, 698M ... you rule :)
[08:03] <coyctecm> will edgy ship with xorg 7.1 ?
[08:03] <ogra> it already does
[08:04] <Kamion> ogra: rock
[08:04] <coyctecm> oh? I installed it one test box at work, and I was looking that xorg was version 7.0.x
[08:05] <Kamion> $ head -n2 /var/log/Xorg.0.log | tail -n1
[08:05] <Kamion> X Window System Version 7.1.1
[08:06] <fabbione> Keybuk: how are we doing?
[08:06] <Kamion> coyctecm: you were probably looking at the xorg package, which isn't really relevant (counterintuitively)
[08:06] <Kamion> it's just a few script
[08:06] <Kamion> s
[08:06] <coyctecm> Kamion: yes, I thinks so too :) My mistake :)
[08:06] <Keybuk> fabbione: fixed it
[08:07] <Keybuk> it was, as I thought, a kernel "issue"
[08:07] <Keybuk> compat_sys_waitid behaves differently to sys_waitid
[08:08] <fabbione> Keybuk: if you can write me a detailed description of the kernel problem, i can talk to David S Miller and see to get it fixed
[08:09] <Keybuk> fabbione: it's not sparc-specific
[08:09] <Keybuk> put simply; do_wait() (called by sys_waitid) takes care to zero the struct siginfo if WNOHANG is to be returned and there's no child waiting
[08:09] <Keybuk> this means that waitid() returns 0 and you check whether info.pid != 0
[08:10] <Keybuk> compat_sys_waitid() returns early without copying if WNOHANG ... so whatever values were passed into that struct are returned, which may be the last waitid or random data
[08:10] <fabbione> Keybuk: ok.. well it might be worth to talk to david and see.. perhaps he can push a fix upstream
[08:10] <Keybuk> the man page for waitid does say, hidden in the notes, that you must zero the struct yourself if you call with WNOHANG
[08:10] <Keybuk> so it was definitely an upstart "bug"
[08:10] <fabbione> ah ok
[08:10] <Keybuk> but I'd also say that it's a kernel bug too for having a silly inconsistency
[08:11] <Keybuk> I've tested and committed the fix (installed on that machine)
[08:11] <fabbione> perfect
[08:11] <Keybuk> and have asked Mithrandir for a freeze exception
[08:11] <fabbione> can i powerit off?
[08:11] <fabbione> it's not a blocker for knot-3
[08:11] <fabbione> sparc can wait the next..
[08:11] <fabbione> i am not too fuzzy about that
[08:11] <ogra_> is there a next ? 
[08:11] <Keybuk> sure
[08:11] <ogra_> i thought next is beta 
[08:12] <fabbione> ogra: i am happy to test the fix with tomorrow's netinstall
[08:12] <Keybuk> we don't install by 64-bit-kernel-on-32-bit-userland by default, no?
[08:12] <fabbione> i don't need a cd for that
[08:12] <ogra_> ah
[08:12] <Keybuk> other than sparc
[08:12] <fabbione> i think we do it only for sparc and hppa
[08:12] <fabbione> and ppc64
[08:12] <ogra_> there are users that use a hack to install amd64 64bit ernels in 32bit userland
[08:13] <fabbione> ogra: hacks are not supported.. 
[08:13] <ogra_> i know
[08:13] <Keybuk> is ppc64 a supported arch?
[08:13] <fabbione> these are supported configs that we ship
[08:13] <fabbione> yes ppc64 is
[08:13] <Keybuk> I thought we just did powerpc and let them change the kernel if they wanted?
[08:13] <fabbione> we ship kernel and 32 bit userland
[08:13] <fabbione> nope
[08:13] <fabbione> it's automatically detected by yaboot
[08:13] <ogra_> but thats something very typical for ltsp users that want the 64bit kernel and flash support ...
[08:13] <Keybuk> oh
[08:13] <fabbione> and it boots/installs the right kernel
[08:13] <Keybuk> then it's probably more critical then :p
[08:13] <Keybuk> this breaks ppc64 too
[08:16] <fabbione> Keybuk: can i shut down the machine in the meantime?
[08:16] <fabbione> or do you still need it?
[08:17] <Keybuk> fabbione: nope, no longer any need for it
[08:17] <Keybuk> thanks
[08:17] <Keybuk> (so you can shut it down)
[08:17] <fabbione> Keybuk: ok perfect
[08:17] <fabbione> Keybuk: i won't kill your account so we can use it again if needed
[08:18] <Riddell> anyone around able to make kubuntu livefs?  mdz? infinity?
[08:20] <Kamion> Riddell: I can
[08:20] <Kamion> Riddell: I've just started off a build
[08:21] <Riddell> thanks Kamion 
[08:21] <Kamion> Riddell: also just pushed a ubiquity change to add scrolling to the mount points table that should be duplicated in the KDE frontend
[08:22] <Riddell> Kamion: ok
[08:23] <fabbione> good night guys
[08:24] <pitti> bye fabbione 
[08:24] <zul> night fabbione 
[08:29] <bluefoxicy> Hey
[08:30] <bluefoxicy> Is Ubuntu 10.04 going to release with MP3 support?
[08:30] <ogra> 10.04 ?
[08:30] <pitti> bluefoxicy: ogg vorbis will have conquered the world by then :)
[08:30] <ogra> might be
[08:30] <ogra> who knows ... 
[08:30] <bluefoxicy> pitti:  I wish :)
[08:30] <pitti> bluefoxicy: will the patent end in 2010?
[08:30] <bluefoxicy> yes, the MP3 patents expire in 2010
[08:31] <pitti> funny that the technology expires much faster than the patent, at least in terms of quality
[08:31] <bluefoxicy> We are having a dispute over MP3/vorbis in another channel and someone brought up that certain linux distros don't ship with MP3
[08:31] <bluefoxicy> so I'm now curious as to if there's other policy in place to prevent it
[08:33] <bluefoxicy> is anyone else on toolchain besides doko?
[08:33] <pitti> bluefoxicy: not actively; infinity sometimes
[08:33] <pitti> (but it's not a normal Australian guy's time to be awake)
[08:34] <bluefoxicy> hrm
[08:35] <bluefoxicy> I am looking to find out what is going to be used for Edgy+1 because I want --hash-style=both by default for the new GNU hash functionality
[08:42] <Yagisan> pitti, so true that
[08:43] <pitti> Yagisan: :)
[08:43] <pitti> Yagisan: *normal* :)
[08:44] <Yagisan> pitti, ;) no one ever says I'm normal
[09:00] <lamont> licio: mailx requires an MTA, and is not something that grandma would have on her computer -- those of us who actually want local mail user/transport agents tend to also know how to install them.
[09:00] <lamont> that's why we dropped the MTA from ubuntu-standard, and postfix now gets to actually listen on port 25 when you install it, since it's not part of the base install any more.
[09:03] <LaserJock> lamont: I liked it when MTA was dropped, I finally got to figure out what an MTA was ;-)
[09:14] <sfllaw> Shootings in downtown Montreal.
[09:15] <o_cee> what kernel will be the final version for edgy, .18?
[09:16] <smurf> fabbione: ping
[09:16] <LaserJock> sfllaw: ?
[09:16] <sfllaw> http://www.cbc.ca/story/canada/national/2006/09/13/shots-dawson.html
[09:19] <LaserJock> oh man
[09:20] <jdong> o_cee: 17. what you see now
[09:20] <o_cee> jdong: m'kay. so no ivtv in the kernel then
[09:21] <jdong> o_cee: why do you say that? can't it be backported to 2.6.17?
[09:21] <o_cee> jdong: don't think it's ready yet
[09:21] <jdong> ah, I see
[09:22] <jdong> that's a shame :-/
[09:24] <o_cee> yeah, would make things alot easier 
[09:24] <zul> sfllaw: not as bad as the time we had shootings inside the rieau center
[09:24] <sfllaw> zul: Terrible.
[09:25] <zul> sfllaw: yep
[09:41] <carlos> pitti: better here...
[09:41] <pitti> carlos: well, sort of
[09:41] <carlos> pitti: what's the difference between libfreetype and libfreetype1 source packages?
[09:42] <carlos> pitti: in edgy and translation contexts
[09:42] <pitti> carlos: not sure, I thought freetype1 was an older version
[09:42] <carlos> pitti: we got translations for libfreetype from dapper, and now, in Edgy, I got a new .pot file for libfreetype1
[09:43] <carlos> pitti: should I import that one from libfreetype1?
[09:43] <pitti> I think not
[09:43] <pitti> lemme check something
[09:43] <carlos> ok
[09:44] <pitti> carlos: only graphviz still needs freetype1, the rest uses freetype
[09:44] <pitti> so freetype1 is an excellent candidate for duplicate elimination
[09:44] <carlos> ok, then I will not import it
[09:45] <carlos> because I guess it will be moved to universe...
[09:45] <pitti> carlos: better not
[09:45] <carlos> pitti: ok, thanks
[10:01] <ogra> Keybuk, Kamion poke ... could someone unleash edubuntu-artwork -35 ?
[10:01] <Keybuk> "unleash" ?
[10:02] <bddebian> "Unleash the hounds"
[10:02] <Keybuk> I only have one "hound", and he's not very lethal ... unless you're allergic to doggy-licks
[10:02] <bddebian> heh
[10:03] <ogra> well, it seems it sits somewhere in the queue 
[10:04] <Keybuk> I'm doing NEW at the moment, if it's in there, it'll get out soon enough
[10:04] <LaserJock> \o/
[10:04] <ogra> right... it was supposed to be in the last CD build, but indeed i forgot to ask to let it through
[10:05] <Keybuk> actually, that's a good point
[10:05] <Keybuk> maybe I should hold off NEWing anything for main?
[10:05] <Keybuk> not sure whether or not that counts as freeze breaking or not
[10:06] <Nafallo> Keybuk: should be :-)
[10:07] <Nafallo> or rather, it should need Mithrandir's approval :-)
[10:07] <Keybuk> indeed
[10:08] <bddebian> Thanks Keybuk
[10:16] <ogra> Keybuk, e-a has a general freeze exception, i got it yesterday from mith...
[10:16] <Keybuk> ogra: show me
[10:16] <Keybuk> with gpg signature
[10:16] <ogra> but that doesnt really matter anyway, since i wont be able to do any further CD tests ...
[10:17] <ogra> so i need to live with the broken one
[10:17] <ogra> well, Nafallo saw it :)
[10:17] <Keybuk> I'm so not believing you after the ltsp stunt :p
[10:17] <Nafallo> Keybuk: I saw it :-)
[10:17] <ogra> what was with the ltsp stunt ... the fix was in the same day you complained about it ...
[10:18] <ogra> and would havebeen in one day ater if you hadnt complained
[10:18] <Nafallo> Keybuk: but that was for if he decided to work until 5 that night, so I'm not sure if it's still active ;-)
[10:19] <Seveas> urgh
[10:19] <Seveas> if a program crashes once, I get a zillion apport dialogs :/
[10:19] <Keybuk> ogra: claiming pitti had approved an MIR when he'd just said he'd reviewed it :p
[10:20] <Nafallo> Seveas: there is a bug for that somewhere :-)
[10:20] <ogra> Keybuk, he said he's confident that we fix it soon and thus i should ask you to let it through to unbreak the CDs
[10:21] <Keybuk> hmm?
[10:21] <Seveas> Nafallo, good 
[10:25] <tseng> whoa Seveas is on fire
[10:26] <ogra> usplash gets the config completely wrong ...
[10:27] <ogra> it created a usplash.conf for 640x480 on my knot3 testsystem ... where the installed and upgraded edgy here gets 1024x786 (which is right)
[10:28] <Kamion> ogra: right, I know about that and have a plan for fixing it
[10:28] <ogra> apparently it doesnt scale the pic right ... so its ~1/3 moved to the left and ~1/3 moved to the bottom
[10:28] <Kamion> at present, yes, it'll default to 640x480 on fresh installs, which is unfortunate
[10:28] <ogra> Kamion, ah, cool
[10:28] <ogra> fine with me ... it just looks a bit odd since the background isnt sitting at the right position
[10:28] <Kamion> bluefoxicy: not an official statement, but I doubt there's any other policy preventing it; in particular there's the free Fluendo implementation that could be used once the patent expires
[10:29] <ogra> the progressbar sits at the right place
[10:29] <bluefoxicy> Kamion:  re mp3?
[10:29] <Kamion> bluefoxicy: yes
[10:29] <Kamion> ogra: that sounds like a theme bug
[10:29] <Kamion> 21:50 < ogra> Mithrandir, can i have a general exception for edubuntu-artwork ?
[10:29] <Kamion> 21:51 < Mithrandir> ogra: sure.  It's edubuntu and nothing else which breaks, which makes me slightly less worried. :-)
[10:29] <Kamion> with regard to e-a
[10:29] <Kamion> (last night)
[10:29] <ogra> Kamion, apart from that, i386 edubuntu is fine for knot 3 ... since thats the only arch i can test right now, i'm not sure what to do about the other arches
[10:30] <Kamion> I'll check/approve it now
[10:30] <ogra> right
[10:30] <ogra> that was the conversation
[10:30] <Seveas> tseng, ?
[10:30] <tseng> Seveas: reading your blog
[10:30] <Seveas> tseng, heh
[10:31] <Seveas> ogra: ubiquity doesn't write an usplash config and the default themes don't support 640x480
[10:31] <Seveas> it's a bug in usplash that it will try to use them
[10:32] <ogra> yep
[10:32] <Seveas> fixing now
[10:33] <Kamion> huh? I fixed that already, just not uploaded
[10:33] <Seveas> ah
[10:33] <Kamion> if it's the usplash bug I think you mean
[10:33] <Kamion> that it tries to use the theme anyway and lets bogl/svgalib fail?
[10:33] <Seveas> bzr is still pulling - so can't look
[10:33] <Kamion> the ubiquity bug is fixed in bzr, but there's a more fundamental and difficult installation ordering bug in d-i
[10:34] <Kamion> need to fiddle with tasksel a certain amount to make that work smoothly
[10:34] <Kamion> fortunately said fiddling is useful to have anyway - it'll make stuff like media changes during the tasksel preinst work properly in Debian
[10:37] <Seveas> hmm, my workdir is fubar
[10:38] <Seveas> Kamion, :
[10:38] <Seveas>   * Exit immediately if no theme is found that fits in the specified
[10:38] <Seveas>     resolution.
[10:38] <Seveas> why not fallback to testcard theme?
[10:41] <Kamion> Seveas: I guess we could do that too, if it fits, but I can't help feeling that at that point text output might be better (once we fix everything to be able to *get* text output)
[10:41] <Seveas> ack
[10:42] <Seveas> testcard theme is 640x400, so should fit virtually anything worth supporting
[10:43] <Kamion> yeah, it just seems to kind of scream "hi, we broke, d'oh"
[10:43] <Kamion> in a way that text output doesn't so much
[10:44] <Kamion> I mean, yeah, it's visibly not pretty, but at least it can be made useful
[10:46] <sfllaw> http://www.flickr.com/photos/vogoa/242601540/
[10:47] <zul> sfllaw: ouch...thats kind of bloody
[10:53] <Keybuk> UDSMV is too wordy, I propose FIUH
[10:53] <Keybuk> or UFH would suffice
[10:55] <Seveas> FIUH? 
[10:56] <Keybuk> Flowers in Ubuntu's Hair :p
[10:57] <Keybuk> a URL to a famous pic of dholbach may be appropriate
[10:57] <ivoks> :)
[10:57] <Seveas> HAHAHA
[10:58] <Seveas> I know that pic dhippyolbach
[11:00] <LaserJock> hahaha
[11:01] <seb128> Keybuk: my desktop looked like edgy was hanging on boot (black screen with only a cursor in the top left corner), happens without the splash and the quiet boot options too. I figured after some reboot that it was doing a fsck (probably after n mount). Is that upstart to blame to not mention what is going on?
[11:01] <Keybuk> the spec that says "ve vill not have ze boot messages" is to blame
[11:02] <seb128> have the box stucked on a black screen for 10 mins is a bug though
[11:02] <Keybuk> 10 mins?
[11:02] <Keybuk> probably fsck mount count or something?
[11:02] <seb128> yeah, the partition is not small
[11:02] <Keybuk> you'd get the same effect with usplash
[11:03] <Keybuk> it'd just sit there at a static graphic
[11:03] <seb128> "figured after some reboot that it was doing a fsck (probably after n mount)"
[11:03] <mjg59> Keybuk: No, usplash would drop out to the fsck text
[11:03] <seb128> usplash dropped after some secs
[11:03] <mjg59> 15 second timeout
[11:03] <seb128> and I got the black screen with the cursor
[11:03] <ivoks> seb128: but you could login, even if fscking
[11:03] <seb128> no
[11:03] <seb128> I got a black screen with a cursor to the top left corner
[11:03] <AlinuxOS> Hello, I need to test new OpenOffice 2.0.4 tranlations + 2.16 GNOME translations for Georgian language... but I still use Dapper, is dist-upgrade quite painless for this moment ?
[11:03] <seb128> no prompt
[11:04] <seb128> no gdm starting
[11:04] <ivoks> no? i was able to login (at alt+f2)
[11:04] <ivoks> of course not gdm
[11:04] <ivoks> it's read only
[11:04] <Keybuk> mjg59: not it wouldn't
[11:04] <Keybuk> mjg59: the fsck text would have been lost under usplash while tty8 was the active, and svgalib was owning it
[11:04] <Keybuk> and usplash would have left them at an empty tty1
[11:04] <seb128> ivoks: that's not a solution, do you really expect users to do that?
[11:05] <ivoks> seb128: of course not, i didn't say it isn't a bug :)
[11:05] <mjg59> Keybuk: In dapper
[11:05] <Keybuk> mjg59: in edgy, right now
[11:05] <Keybuk> I tested it not 8 hours ag
[11:05] <Keybuk> +o
[11:05] <mjg59> Keybuk: Uh. I'm pointing out that it's a regression, not disputing that it's the current situation.
[11:05] <seb128> my issue is that with splash I had no indication of what was going on too
[11:05] <Keybuk> ah
[11:05] <Keybuk> sorry
[11:06] <seb128> "without splash"
[11:06] <Keybuk> mjg59: it seems there's no way for have usplash on an active tty, and not lose /dev/console output
[11:06] <Burgwork> Kamion, are you still around?
[11:06] <Keybuk> I had a bit of a fiddle
[11:06] <SlicerDicer> lol Burgwork I was just about to ask
[11:06] <SlicerDicer> haha
[11:06] <Keybuk> the best solution I came up with was to use logd to output it to tty1 by force
[11:06] <seb128> Keybuk: so if I file a bug on upstart you will close it because that's a design decision? ;)
[11:06] <Kamion> Burgwork: technically yes, but I'm just off to bed, so not any more - sorry
[11:06] <Burgwork> Kamion, I will catch you tomorrow. I wanted to speak to you about your SoC student
[11:06] <Kamion> SlicerDicer: feel free to send mail if you can't get hold of me otherwise
[11:07] <SlicerDicer> Kamion: thank you I will bother you tomorrow if I can otherwise I will mail you
[11:08] <Keybuk> seb128: right
[11:08] <Keybuk> you should have turned up at the boot-message-logging BOF in Paris to object <g>
[11:09] <seb128> nah :p
[11:10] <seb128> k, I'll bug upstart anyway ;)
[11:10] <seb128> if it blocks the box for longer than some seconds it should give some hint of what is going on
[11:11] <Keybuk> that will not change before edgy+1
[11:11] <Keybuk> so I'll reject it anyway
[11:11] <seb128> grumpf
[11:11] <seb128> that is broken
[11:11] <seb128> what do you think a guy having his computer blocking on a black screen for 10 min will do?
[11:12] <Keybuk> arguably it's a bug in sysvinit
[11:12] <Keybuk> as the checkfs scripts should ensure that what they are doing is notified to /dev/console and usplash
[11:12] <seb128> I started by trying to boot without usplash and without quiet then
[11:12] <seb128> then I tried to boot previous kernel
[11:12] <seb128> then I booted dapper
[11:13] <seb128> k
[11:13] <jdong> Keybuk: I find it hard to blame sysvinit for something that worked with sysvinit but no longer works with upstart :-/
[11:13] <seb128> I'm fine with opening a bug on sysvinit
[11:13] <seb128> but that's a bug somewhere
[11:13] <jdong> get it fixed somewhere... especialyl since we still have a periodic fsck set
[11:14] <Keybuk> seb128: would you reject a bug that asked for the old GNOME logout dialog?
[11:14] <seb128> yep
[11:14] <jdong> it'll only take 35 bootups before we get our first rush of bug reports regarding a blank screen and hung bootup
[11:14] <Keybuk> why?  because it was specced to be changed
[11:14] <Keybuk> this is the same
[11:14] <Keybuk> it was discussed, specced, reviewed and approved
[11:14] <mdz> tkamppeter: http://www.ubuntu.com/ubuntu/components
[11:14] <seb128> Keybuk: I would not reject a valid objection on the new one though
[11:14] <Keybuk> the change is completely irrelevant to upstart, it's just a setting for the rc scripts
[11:14] <mdz> tkamppeter: and the table on https://wiki.ubuntu.com/UbuntuDemystification
[11:15] <seb128> Keybuk: I'm fine filling it somewhere else as said
[11:15] <seb128> but that's a bug *somewhere*
[11:15] <Keybuk> I have no problem with deciding to drop the spec entirely
[11:15] <Keybuk> and revert both console and usplash to dapper behaviour
[11:15] <seb128> I expect most user will think the distro is broken if it's staying there for 10min and doesn't boot
[11:15] <seb128> without any indication of what is happening
[11:16] <Keybuk> I don't disagree
[11:17] <jdong> Keybuk: so is it not possible for upstart to "revert" to console output only if usplash times out?
[11:17] <Keybuk> jdong: no, you can't change the file descriptors of a running process
[11:17] <jdong> I see
[11:18] <jdong> but what if it was logged to a logfile, then upstart spawns a "tail -f" on a tty when usplash fails?
 the best solution I came up with was to use logd to output it to tty1 by force
[11:18] <Keybuk> . o O { ah, I believe we have reached the middle of this conversation }
[11:18] <jdong> :)
[11:18] <jdong> excuuuuse me for not reading the scrollback :P
[11:23] <Keybuk> the cleanest solution is probably just to make checkfs and checkroot do things properly
[11:23] <Keybuk> and make sure the message is printed
[11:26] <seb128> Keybuk: basically that's https://launchpad.net/distros/ubuntu/+source/sysvinit/+bug/58609 then, right?
[11:26] <Ubugtu> Malone bug 58609 in sysvinit "upstart doesn't display checkrootfs.sh messages" [Untriaged,Confirmed]  
[11:28] <Keybuk> yes
[11:29] <seb128> ok, thank you
[11:30] <seb128> maybe you can milestone it for edgy? ;)
[11:30] <Keybuk> I'm unsure who by and how the milestone field is supposed to be used?
[11:31] <seb128> I use it to note the bugs I planned to try getting fixed for a distro for some time
[11:32] <seb128> and nobody complained ;)
[11:32] <seb128> "I plan to fix for next distro" rather
[11:32] <seb128> would be a question for mdz maybe
[11:33] <mdz> Keybuk: the first rule of the milestone field is, we don't talk about the milestone field
[11:33] <Keybuk> mdz: that's what I thought
[11:33] <Keybuk> I've been ignoring it
[11:34] <mdz> it's meant to be used for, well, milestone
[11:34] <mdz> s
[11:34] <mdz> which I've said we don't really have a use case for
[11:34] <Keybuk> my experience so far has that people just use it for an "OMG! SO CRITICAL IT HURTS!WJKREJQUU!*("$(*!_("*$(_!*"$(_I129- _!(*U"*(U*IP" type thing
[11:34] <mdz> and then there's to be a different mechanism for targeting bugs to releases
[11:34] <mdz> which, it's my understanding, isn't implemented yet
[11:34] <mdz> Keybuk: it's supposed to be restricted to the same people who can set importance
[11:35] <seb128> I milestone "edgy" something when I think it should be worked for edgy
[11:35] <seb128> it works fine usually
[11:35] <seb128> desktop guys look at that list
[11:35] <seb128> and try to fix issues listed there in priority
[11:36] <mdz> seb128: yes, but Mark has said that it isn't meant for that, and there is work in progress to have a mechanism to mark bugs for edgy
[11:36] <mdz> milestone is meant to be for knot 1, knot 2, etc.
[11:36] <seb128> ok
[11:36] <Keybuk> what does Mark thing it's meant for?
[11:36] <Keybuk> ah
[11:37] <seb128> I do with what we have :p
[11:37] <mdz> I know
[11:37] <mdz> this is why we don't talk about it
[11:37] <seb128> I see ;)
[11:37] <mdz> so if you are happy then please continue quietly ;-)
[11:37] <seb128> will do :)
[11:37] <Keybuk> we need to take a decison about boot messages at some point
[11:37] <Nafallo> ,
[11:38] <Nafallo> hehe even
[11:38] <Keybuk> I was going to propose it for the TB meeting
[11:38] <Keybuk> but then I realised it was this week, when I was off watching Jean-Luc Prospero
[11:38] <infinity> They all seem to be easily definable as milestones/snapshots/releases/points-in-time, whatveer.
[11:40] <Nafallo> I was just thinking the same myself
[11:42] <sistpoty> infinity: do you have some time to help with bootstrapping fpc?
[11:43] <bluefoxicy> Actually I'll discuss this one with Casey
[11:43] <bluefoxicy> he's pretty bright, he might have some ideas
[11:44] <mdz> Keybuk: what about them?
[11:44] <mdz> Keybuk: on or off?
[11:44] <infinity> sistpoty: I will in a couple of hours.  I'm just doing the "wake up, pack up stuff, wander over to another house" dance.
[11:44] <Keybuk> mdz: on, or off; yes
[11:44] <mdz> Keybuk: sounds like an ubuntu-devel thing rather than tech board
[11:44] <Keybuk> you're probably right
[11:44] <mdz> or a strawman launchpad vote!
[11:44] <infinity> sistpoty: If you're around in ~2 hours, bug me about it.
[11:44] <Keybuk> I'll draft a mail
[11:44] <Seveas> Keybuk, how about 'on only for fsck'
[11:44] <bluefoxicy> (I'm pondering the potential to use OLPC-type hardware in developed countries to enhance primary school education; as a side bonus, you could potentially persuade developed countries' school systems to pay extra as a 'donation' so that the profits can go to free machines for the undeveloped schools in third world countries)
[11:45] <Keybuk> Seveas: that's only implementable for edgy+1
[11:45] <sistpoty> infinity: ok, I'll stay up till then ;)
[11:45] <bluefoxicy> later
[11:45] <sistpoty> and thx. btw. ;)
[11:45] <Seveas> Keybuk, due to FF or technical reasons?
[11:45] <Keybuk> technical reasons
[11:45] <Seveas> usplash already can do "don't do text things unless needed"
[11:45] <Keybuk> obviously "make the checkfs/checkroot scripts write over /dev/console and send usplash commands" should be possible
[11:45] <mdz> speaking of fsck, did anyone else find that email odd, about flaky power causing fsck to report unexpected inconsistencies?
[11:45] <Keybuk> define "needed" ?
[11:45] <mdz> I've rudely killed ext3 boxen an awful lot without seeing that
[11:46] <Seveas> a trivially implemented URGENTTEXT 
[11:46] <Seveas> and chkfs/chkroot using it
[11:46] <Keybuk> Seveas: right, I was thinking along those lines ... have check* send something to usplash and the console saying "I'M DOING A FSCK!"
[11:46] <infinity> mdz: It's happened to me a few times, but I can't rule out sketchy hardware as s possibility, so never bothered to distill images and hurl them at Ted.
[11:46] <mjg59> Keybuk: We still run into problems if the fsck hits anything unexpected and needs user input
[11:46] <Keybuk> mjg59: < /dev/console
[11:46] <mdz> Keybuk: it might be possible to run fsck in a "tell me if you're going to take a while" mode
[11:47] <mjg59> Keybuk: For edgy?
[11:47] <Keybuk> we could wimp out and do exec </dev/console >/dev/console 2>&1 at the top of check* too
[11:47] <Keybuk> though we'd have to kill usplash or something
[11:47] <mdz> Keybuk: what's on stdin for boot scripts now?
[11:47] <Keybuk> which is tricky :p
[11:47] <Keybuk> mdz: /dev/null
[11:47] <Seveas> s/the top/if it really goes to check things/
[11:48] <mdz> Keybuk: I like that in a sadistic sort of way, to foil packages which do stupid things during boot
[11:48] <Keybuk> mdz: right
[11:48] <mdz> like that package which was in the laptop task in...woody? potato?
[11:48] <Keybuk> it saves the boot hanging on some silly package trying to ask questions in the boot process
[11:48] <Keybuk> or blocking on tty
[11:48] <Keybuk> etc.
[11:50] <mdz> the only other use case I'm aware of is interrupting dhclient
[11:50] <Seveas> cryptsetup
[11:50] <mdz> which shouldn't block anymore in a default config, right?
[11:50] <mdz> Seveas: hmm, good point
[11:50] <Keybuk> right, that is supposed to get called from a udev rule
[11:50] <mdz> I think cryptsetup is pretty hosed in the new world order anyway though
[11:50] <Seveas> indeed
[11:51] <Seveas> although now that usplash supports INPUT, cryptsetup may recover pretty nice
[11:52] <infinity> apache with passphrases on SSL certs (something I always recommend against, mind you) is hosed in the new world order too.
[11:53] <infinity> That's something I'm happy about, mind you.
[11:53] <infinity> One more reason to point and laugh at people who think they're doing something reasonable.
[11:53] <Nafallo> that seems like more paranoia than me current firewall :-P
[11:53] <Keybuk> cryptsetup is equally hosed as checkfs and checkroot
[11:53] <Nafallo> s/me/my/
[11:54] <Keybuk> those are the three things that I'm aware of that can halt the boot process, ask for input, and probably should continue to do so
[11:55] <Nafallo> I'm the only one who misses all that text? :-)
[11:55] <Seveas> hehe
[11:56] <infinity> I miss it desperately when booting without splash.  The complete lack of visual feedback on a black screen for a minute or twois very disconcerting.
[11:56] <slomo_> Nafallo: nope ;)
[11:56] <Seveas> boot without 'quiet'
[11:57] <infinity> Seveas: That will get me kernel messages, won't get me anything from init anymore.
[11:57] <Nafallo> Seveas: will that get my usplash to display text again? :-)
[11:57] <Seveas> Nafallo, yes
[11:57] <Nafallo> what infinity said though :-P
[11:57] <Keybuk> infinity: that's actually a bug of the "keybuk can't spell quiet" variety
[11:57] <Seveas> infinity, hmm, then I misunderstood Keybuk 
[11:58] <infinity> Keybuk: Oh, hah.
[11:58] <Seveas> I thought upstart looked at 'quiet'
[11:58] <Seveas> ah
[11:58] <Keybuk> however the Ice Queen has frozen ubuntu yet again ;p
[11:58] <Keybuk> Seveas: "quite" is an entirely valid word
[11:58] <Seveas> true
[11:58] <Nafallo> lol
[11:58] <Seveas> trigraphs to the rescue
[11:59] <Nafallo> that's totally worthy of an upload after knot-3 I hope? :-)
[11:59] <Keybuk> yes
[11:59] <infinity> Still, not sure I want to get 12 screens of kernel vomit just to get my 30 lines of init script output back.
[11:59] <Keybuk> the ice queen didn't event want me to upload a fix for ppc or sparc :p
[11:59] <Keybuk> let alone something that trivial
[12:00] <Keybuk> infinity: quiet, verbose ... pick one :p
[12:00] <infinity> Makes sense for desktops, I'm still unconvinced for systems without a splash.
[12:00] <Nafallo> Keybuk: maybe you could implement a third word? :-)
[12:00] <Keybuk> infinity: *shrug* like I said earlier, most grown up UNIXes don't bother with the 30 lines of init script output either
[12:00] <Keybuk> the most annoying thing about those 30 lines is that an error message in the first 6 is lost
[12:00] <Nafallo> something like "messages" :-)
[12:01] <infinity> Keybuk: There you go, you just defined 3 states, problem solved.  :)  quiet: no output, verbose: Oh god, too much, neither: give me init and no kernel. :)
[12:01] <Keybuk> infinity: that involves a kernel patch, no? :p
[12:01] <infinity> Keybuk: Yes.  We're hardly against that. :)
[12:01] <Keybuk> the boot-message-logging spec actually defines several levels
[12:01] <Keybuk> the problem is that it's unimplementable for edgy
[12:01] <Seveas> crackful
[12:01] <Keybuk> for edgy, we can implement two levels.  "on" and "off"
[12:02] <Keybuk> actually, that isn't entirely true
[12:02] <Keybuk> we can implement three if we change the kernel
[12:02] <Keybuk> though colin wants a verbose option to mean "ignore the earlier quiet option"
[12:03] <Keybuk> tbh though, I don't see the win of three levels
[12:03] <Nafallo> Keybuk: so "messages" then :-)
[12:03] <Seveas> we could name the 3rd level infinity 
[12:03] <Keybuk> Debian always traditionally spat the kernel output over the init script messages
[12:03] <Keybuk> and never had a problem with that
[12:04] <Keybuk> a minimal/silent grown-up boot is the logical opposite to that
[12:04] <Nafallo> well, people with Debian knows that, we have already diverged from that :-)
[12:05] <Keybuk> mdz: what happened to getting rid of the grub messages?
[12:05] <mdz> Keybuk: it's done, but we don't do grub-install on upgrade
[12:05] <infinity> Keybuk: I never used to mind the kernel and init both spewing together, but modern kernels have become MUCH more verbose.
[12:05] <mdz> Keybuk: zul did it
[12:06] <mdz> I think we should change the message though
[12:06] <Keybuk> "the message" ?
[12:06] <mdz> it prints "Booting the operating system now"
[12:06] <mdz> so that something is on the screen while the kernel loads
[12:06] <mdz> I think it should just say "Starting up" or so
[12:07] <mdz> too much technical jargon
[12:07] <Nafallo> maybe we should ask mpt what he would like? :-)
[12:07] <Keybuk> infinity: I don't think that's true ... I think that modern kernels just load modules at different times
[12:07] <mdz> I will ask his opinion