[12:08] <seb128> does that makes sense to you?
[12:08] <Keybuk> try changing the exec line to
[12:08] <Keybuk> exec >/tmp/rc2.log 2>&1
[12:09] <ogra> does anyone know if port 1337 is used by anything ? 
[12:09] <jdong_> ogra: oh boy... why?
[12:10] <seb128> Keybuk: and then? start rc2?
[12:10] <pygi> ogra, I don't think so
[12:10] <jdong_> ogra: I've used it on occasion for the coolness factor... but nothing that I know actually uses it
[12:10] <ogra> for fun... i'm looking for a funny portnumber (only for internal usage)
[12:10] <seb128> Keybuk: same log
[12:10] <jdong_> ogra: it is said to be bad luck to run your daemon on 1337 :)
[12:10] <ogra> well, its likely that many people will occasionally use it i guess
[12:11] <ogra> pfft, bad luck ...
[12:11] <Keybuk> seb128: ok
[12:11] <Keybuk> seb128: change the exec /etc/init.d/rc 2 to exec sh -x /etc/init.d/rc 2
[12:13] <seb128> Keybuk: http://paste.ubuntu-nl.org/22680
[12:15] <Keybuk> seb128: errrrrrr
[12:15] <Keybuk> could you ls /etc/rc2.d for me
[12:16] <seb128> $ ls /etc/rc2.d
[12:16] <seb128> README
[12:16] <elmo> LOL
[12:16] <Keybuk> *blink*
[12:16] <seb128> hum
[12:16] <ogra> heh
[12:16] <_ion> :-)
[12:16] <Keybuk> I THINK WE MAY HAVE FOUND YOUR PROBLEM
[12:16] <_ion> You think so?
[12:17] <seb128> something nuked /etc/rcn.d
[12:17] <jdong_> whoa
[12:17] <jdong_> lol
[12:17] <seb128> *great*
[12:17] <Keybuk> seb128: all of them?
[12:17] <jdong_> quick! blame it on slomo's dbus upload!
[12:17] <seb128> rc0.d:
[12:17] <seb128> README
[12:17] <seb128> rc1.d:
[12:17] <seb128> README
[12:17] <seb128> rc2.d:
[12:17] <seb128> README
[12:17] <seb128> rc3.d:
[12:17] <seb128> README
[12:17] <seb128> rc4.d:
[12:17] <seb128> README
[12:17] <seb128> rc5.d:
[12:17] <seb128> README
[12:17] <seb128> rc6.d:
[12:17] <jdong_> oh boy
[12:18] <seb128> README
[12:18] <seb128> rcS.d is still populated
[12:18] <jdong_> I see someone reaching for a install cd soon.... :)
[12:18] <Keybuk> no postinst/rm that I can find does that
[12:18] <_ion> Does the README say "ha ha, got you"?
[12:19] <Kamion> seb128: did you have initscripts uninstalled at any point?
[12:19] <Kamion> seb128: if so, did you perchance purge it?
[12:19] <Keybuk> Kamion: why would initscripts uninstall everything?
[12:19] <Keybuk> even gdm
[12:19] <Kamion> oh, true, it wouldn't nuke the lot ...
[12:19] <seb128> Kamion: I don't think so
[12:19] <elmo> /var/log/dpkg.log should tell you?
[12:20] <seb128> nothing today
[12:21] <zyg1> eh, my primary disk is starting to die according to smartctl :/
[12:21] <zyg1> crap
[12:21] <ajmitch> zyg1: that's why you need RAID
[12:22] <ajmitch> a chance to frantically scramble to replace the disk
[12:22] <Keybuk> seb128: what have you installed since you last booted?
[12:22] <zyg1> ajmitch: not untill ubuntu installer allows me to make one without thinking (for /)
[12:22] <seb128> tonight, libssl0.9.8 liboobs-1-1 powernowd
[12:22] <ajmitch> zyg1: worked ok for me with the alternate cd (separate /boot)
[12:23] <seb128> and I was upgrading gnome-system-tools
[12:23] <zyg1> ajmitch: what I actually want is a flash disk of about 10GB that's comparable in price with 250GB of spinning metal
[12:23] <seb128> none of them looks like they are doing some weird thing
[12:24] <zyg1> ajmitch: I will try though, I'm getting sick of failing disks, that's the third one this month
[12:24] <ogra> grmbl
[12:24] <ogra> are the keymaps settings in xorg broken currently ? 
[12:24] <Keybuk> seb128: when did you install upstart?
[12:25] <jdong_> zyg1: have you had any seagates fail on you yet?
[12:25] <jdong_> zyg1: it's about the only brand that's been nice to me
[12:25] <seb128> Keybuk: after that my system had issues to reboot, I figured I might have been in a middle of an upstart upgrade because I didn't have ubuntu-minimal installed
[12:25] <seb128> so I would not blame it
[12:25] <Keybuk> no, I'm just trying to ascertain what you managed to uninstall by accident :p
[12:25] <ogra> pitti, are you able to use setxkbmap on an up to date edgy ? 
[12:26] <zyg1> jdong_: the one failing just now is a segate 
[12:26] <Keybuk> I know I definitely didn't put any "if user == seb, delete all his rc symlinks" code in there
[12:26] <zyg1> no kidding
[12:26] <ogra> i only get "couldn't interpret _XKB_RULES_NAMES property"
[12:26] <jdong_> zyg1: wow... never mind then... I was gonna tell you to run away from maxtor.... ;)
[12:26] <Keybuk> (but not the README, which is quite special ... that takes extra-perverse coding)
[12:27] <zyg1> jdong_: why is flash so mp3like and not server like, eh :/
[12:27] <jdong_> zyg1: because it's so expensive-like? :)
[12:27] <jdong_> zyg1: and I honestly question if it's any more stable
[12:27] <jdong_> zyg1: I've seen CF cards die on me before
[12:28] <zyg1> jdong_: but if hard disks were shiny and came with earphones I'm sure people would buy them... oh wait!
[12:28] <jdong_> we kind of abuse them on our robotics team... IO-intensive workloads off a sandisk 128MB CF card
[12:28] <zyg1> jdong_: yeah but they don't fail being read (as my server actually does)
[12:28] <jdong_> so flash isn't the answer :-/
[12:28] <elmo> Keybuk: it's notably the only file in there (as opposed to a symlink) so not that perverse
[12:28] <gnomefreak> is there a chance of seeing mozilla-thunderbird 2.0 to match ff in edgy?
[12:29] <jdong_> I find it interesting that rcS.d was unaffected
[12:29] <zyg1> heh, I heard a story today: one of our customers was apparently sent a development device that was .... continously updating the flash (for testing purposes) for the past ..... YEAR
[12:29] <Keybuk> elmo: I could understand rm /etc/rc[0-6.d/*
[12:29] <Keybuk> or even rm -rf /etc/rc[0-6] .d
[12:29] <Keybuk> but deliberately iterating symlinks and removing just those ... perverse
[12:29] <Keybuk> seb128: you have things in /etc/init.d still, yes
[12:29] <zyg1> jdong_: IMHO flash is more reliable than spinning disk, especially for rare writes as in this case
[12:29] <Keybuk> and you didn't run any of the Debian "re-order my rc.d" crap? :p
[12:30] <zyg1> Keybuk: lol
[12:30] <seb128> Keybuk: /etc/init.d looks complete
[12:30] <Keybuk> PIMP MY RC.D
[12:30] <_ion> rm /etc/rc[0-6] .d/*(@) (zsh)
[12:30] <_ion> :-)
[12:30] <seb128> Keybuk: ARG
[12:30] <seb128> Keybuk: thank you
[12:30] <Keybuk> seb128: ARG? :p
[12:30] <seb128> Keybuk: I tried the new services-admin from g-s-t update
[12:30] <Keybuk> I. See.
[12:30] <seb128> but didn't do anything with it
[12:30] <seb128> just opened and closed
[12:30] <jdong_> seb128: that could've done it...
[12:31] <Keybuk> clearly it did things to you in revenge :p
[12:31] <jdong_> :)
[12:31] <Keybuk> did you ship it in edgy?
[12:31] <pitti> ogra: yes, works fine
[12:31] <ogra> weird
[12:32] <elmo> aww man
[12:32] <elmo> we can't blame this on Scott?
[12:32] <elmo> damn it
[12:32] <Hobbsee> elmo: blame him anyway :D
[12:32] <Keybuk> I thought we were blaming slomo for everything this week?
[12:32] <Hobbsee> elmo: dont get caught up in minor details on whether it's his fault or not.  blame him anyway.
[12:32] <Hobbsee> Keybuk: no, that was last week.
[12:33] <Hobbsee> and the weeks suddenly started on wednesdays.
[12:33] <jdong_> if upstart knew how to start the system without sysv scripts, seb128 would still be booting fine. Thus, scott is to blame.
[12:33] <jdong_> done
[12:33] <_ion> pitti just went offline, so why not simply blame him?
[12:33] <seb128> Keybuk: sorry for making you loose time on that, thank you for the help figuring we should really not ship that services-admin :p
[12:33] <ajmitch> Keybuk: might as well, he broke f-spot for me with dbus :)
[12:33] <jdong_> hmm, should I try services-admin for fun?
[12:34] <Keybuk> three times
[12:34] <Keybuk> and don't use sbackup :p
[12:34] <Keybuk> or whatever that thing was
[12:35] <ogra> its broken by upstart i heard :)
[12:35] <ogra> wies your system etc :P
[12:35] <_ion> I started services-admin and quit it, rc*.d are fine.
[12:35] <ogra> *wipes
[12:35] <Keybuk> anyway
[12:35] <Keybuk> seeing as I'm not to blame
[12:35] <Keybuk> I'm going :)
[12:35] <Keybuk> byeeeee
[12:36] <jdong_> likewise, it didnt nuke my rc*.d
[12:38] <seb128> great
[12:38] <seb128> that's services-admin to blame
[12:39] <ogra> its probably a bit risky to ship such tools in edgy with upstart ...
[12:40] <jdong_> well, not having a services-admin type of tool kind of sucks, too
[12:40] <jdong_> has anyone managed to reproduce this?
[12:40] <ogra> right, but breaking your system because it cant cope with the new initsystem is more evil
[12:41] <jdong_> true
[12:41] <seb128> jdong_: I've not uploaded the new version yet
[12:41] <zyg1> hmm, do you guys think that this is a bug in python: "".split() != "".split(',')
[12:41] <jdong_> seb128: ah, I see
[12:41] <seb128> and I'll not 
[12:42] <jdong_> weirdly the Kubuntu services program deals with upstart just fine
[12:42] <ogra> so should services-admin
[12:45] <Kamion> ogra: what's the state of ltsp-dhcpd-autogeneration?
[12:45] <Kamion> it's the only Essential spec, and isn't done
[12:46] <wasabi> Has any consideration (maybe a bit odd) been given to using the ideas pioneered by upstart for the desktop itself?
[12:46] <ogra> Kamion, right, working on it ... let me finish login and session handling 
[12:46] <wasabi> Maybe dbus covers it. But, imagine upstart starting per-user when you log in, and replacing gnome-session (or parts of it)
[12:46] <zyg1> wasabi: huh?
[12:46] <shackan> argh, dbus
[12:47] <_ion> wasabi: Well, upstart is going to support per-user services, but not in edgy.
[12:47] <wasabi> Ahh.
[12:47] <_ion> s/services/jobs/
[12:48] <wasabi> jsut cron-like stuff.
[12:48] <wasabi> not attached to a particular X session
[12:49] <_ion> Well, i don't see anything that prevents one from implementing something like that. It hasn't really been planned AFAIK.
[12:49] <ogra> Kamion, ltsp-dhcpd-autogeneration will be basically done tomorrow evening ... its the last one i have left and i have the full day tomorrow for it ...
[12:50] <_ion> But X session management seems to be a bit out of scope for upstart.
[12:51] <Kamion> ogra: ok, please make sure that gets done; I am available for code review tomorrow, and indeed I'd like to see its state
[12:51] <Kamion> ogra: since it's Started, how much of the code is in place?
[12:56] <ogra> Kamion, i have script pieces that determine a possible free NIC and echo an entry for it to /etc/network/interfaces ... the rest is debconf template and value stuff i'll have to write tomorrow i havent looked into it this week yet since i'm nearly done with the login/session spec and want to finish it first
[12:56] <Kamion> ogra: ok, expect mdz to be asking about the status when he gets back tomorrow
[12:57] <Kamion> just by way of incentive :)
[12:57] <ogra> i do :)
[12:57] <ogra> but i have the full day for it so i'm not really worried ...
[12:58] <Kamion> I'm worried about everything that's not completed yet
[12:58] <ogra> and its really some trivial shell scripting only ... the bigger part for me is the debconf stuff ... i do this kind of things to rarely
[12:58] <Burgwork> Kamion, I think that is part of your job
[01:02] <Kamion> ogra: hence my offer of code review, since I do debconf hacking all the time
[01:02] <ogra> i know
[01:02] <ogra> but i'm not on that part yet 
[01:03] <ogra> i'll get up very early i hope i can show you something around noon ...
[01:04] <ogra> <- short break
[01:05] <Kamion> ogra: thank you
[01:06] <Kamion> Mithrandir: should live-cd-write-as-you-go be deferred now?
[01:07] <Kamion> Riddell: is kubuntu-accessibility just blocking on me doing the cdimage/gfxboot glue, or is there more to it?
[01:08] <Riddell> Kamion: it's blocking on me sending the casper changes to Mithrandir, which I plan to do in an hour or so
[01:10] <Kamion> mvo_: is cdrom-based-dist-upgrades "beta available" (or more) now?
[01:11] <mvo_> Kamion: it is beta-available now, yes. I will update it
[01:11] <Kamion> thanks
[01:12] <Kamion> zul: what's the state of xen-edgy?
[01:13] <Kamion> zul: you wrote in the status whiteboard "zulcss 2006-08-16: Testing has begun on x86 amd64 still blocked on the builds" but I don't understand what blockage exactly you mean
[01:14] <Burgwork> welshbyte, you need to talk to sivang about python-based backup software
[01:15] <welshbyte> Burgwork: yeah i'm planning to look into hubackup, it looks cool
[01:15] <Burgwork> welshbyte, given google has now funded 3 seperate backup projects, all written in python
[01:20] <welshbyte> Burgwork: yeah i guess it does duplicate work and spread the skill thin
[01:22] <Burgwork> welshbyte, you should talk with the author of system-config-backup (this years effort, being done by Fedora)
[01:22] <Burgwork> backpack was last years Fedora and sbackup was last years Ubuntu
[01:23] <welshbyte> Burgwork: yeah i took over last year's fedora one
[01:23] <welshbyte> not that i use fedora much...
[01:23] <Burgwork> I am forced to use it at work
[01:25] <welshbyte> my computer society uses it
[01:25] <welshbyte> despite my best efforts to get them to switch the desktops to ubuntu 
[01:28] <Burgwork> you need to provide them with something that is clearly better than Fedora
[01:29] <pygi> Burgwork, crack fedora boxes :P
[01:30] <welshbyte> well i'll keep pushing for it while i still hold sway on the exec committee
[02:46] <BenC> anyone know why a new debhelper would be breaking kernel compiles?
[02:46] <BenC> dh_strip debug symbol extraction: using obsolete debhelper compat mode 1
[02:46] <BenC> Directory debian/tmp-headers does not exist, aborting
[02:46] <BenC> that's how things are all of the sudden failing
[03:03] <bddebian> Heya
[03:25] <Kamion> Mithrandir: ok, I've got as far as getting "These keymaps appear to be indistinguishable" errors out of gen_keymap, so now it's just a matter of pruning keymaps from the tree until the result is sensible.
[03:41] <bluefoxicy> You know what would be cool?
[03:41] <bluefoxicy> Human as a dynamic theme.  Draw the buttons using a little random data to introduce slight, organic-looking variation.  :)
[03:42] <bluefoxicy> (I have been waiting for this since the slashdot on wobbly windows and other such experimental cruft, but no dice)
[04:06] <theCore> does Ctrl-Shift+[Unicode]  has been removed from GNOME?
[04:20] <grexk> theCore: dapper? nope
[05:40] <desrt> "If you hope some day to look back on your career and feel that it has contributed to the growth of a good and free society, you need to make your software free." -- FSF
[05:42] <infinity> Yeah, I take warm fuzzies in exchange for pay.
[05:42] <infinity> Sure, it doesn't pay the rent, but boy do I feel good about myself.
[05:42] <infinity> Also, I'm so very, very hungry.
[05:43] <desrt> programs that display the GPL when you install them and force you to "agree" are really dumb
[05:43] <infinity> Very, very dumb, since it's not an EULA, it's a distribution license.
[05:43] <LaserJock> I've noticed a lot of those on OS X apps
[05:44] <infinity> Windows stuff, too.
[05:44] <infinity> People who write free software for proprietary operating systems don't seem to "get" the license.
[05:44] <desrt> EULAs are really suspicious to me
[05:44] <infinity> To be fair, people who write free software for Linux/BSD probably don't "get it" a lot of the time too, but we have enough people willing to hunt them down and explain it to them.
[05:45] <desrt> if someone has sold you the software on the CD then you own that CD
[05:45] <desrt> you own a copy of their work.  period.
[05:45] <Fujitsu_> No.
[05:45] <desrt> you can do with it as you like as long as you don't violate their copyright
[05:45] <Fujitsu_> Said copy of their work is /licensed/ to you.
[05:45] <Fujitsu_> Not owned by you.
[05:45] <desrt> (or any other applicable laws)
[05:45] <desrt> Fujitsu_; that's BS.
[05:45] <Fujitsu_> That's what the licenses say.
[05:45] <desrt> they might want you to think that
[05:45] <desrt> but it's BS
[05:46] <desrt> that's like somone adding a note to the back of a cd saying "the music on this cd is licensed to you.  the license says we can take the cd away whenever we want."
[05:46] <desrt> it's clearly batty
[05:48] <Fujitsu_> That's life.
[05:49] <desrt> yes.  companies thinking that they have rights which they actually don't is life indeed
[06:24] <Amaranth> mjg59: Adding "nvidia" to MODULES in /etc/default/acpi-support seems to make hibernate work on my HP in edgy, thought you might like to know.
[08:19] <dholbach> good morning
[08:58] <Burgundavia> hunger: it is now installed by default
[08:59] <hunger> Burgundavia: I noticed:-(
[08:59] <hunger> Burgundavia: Does it actually work for somebody?
[08:59] <Burgundavia> yep
[08:59] <Burgundavia> you must jsut have interesting software
[09:00] <hunger> Burgundavia: It does not bring up the consoles to log in which is pretty annoying.
[09:01] <hunger> Burgundavia: That it fails to mount up encrypted devices is much more annoying though.
[09:01] <Burgundavia> file bugs
[09:01] <hunger> Burgundavia: I did.
[09:01] <Burgundavia> it is early days yet
[09:03] <hunger> Burgundavia: The first version I tested worked better than the 0.2.x :-|
[09:03] <Burgundavia> I am most definitely not hte person to talk to about that
[09:03] <hunger> Burgundavia: You are right, I'll stop bitching.
[09:04] <pitti> Good morning
[09:04] <dholbach> heya pitti!
[09:04] <hunger> Burgundavia: Sorry.
[09:04] <hunger> pitti: Hi.
[09:04] <jdub> hunger: this is edgy.
[09:05] <Burgundavia> hunger: no worries
[09:05] <jdub> hunger: we have to do *something* edgy.
[09:05] <Burgundavia> I also understand that Keybuk accepts patches
[09:05] <Burgundavia> jdub: we should have done smart
[09:05] <hunger> jdub: Yeap. Do it!
[09:05] <Burgundavia> and why does Mark have gustavo squirrlled away right a freaking management client?
[09:05] <hunger> jdub: But of course you are evil if you do something that breaks my system:-)
[09:07] <dholbach> Burgundavia: ah! you volunteer to port all the stuff to smart and hack apt features into it! :)
[09:08] <Burgundavia> dholbach: no, but isn't that what mark is paying all those smart dpkg people todo?;)
[09:08] <jdub> Burgundavia: you know, that whole "making money to ensure that canonical can continue with this whole ubuntu racket"
[09:08] <Burgundavia> yes, but anybody can write a freaking managment server. AFter all, the clowns at my company did it
[09:09] <dholbach> Burgundavia: what do the dpkg people have to do with smart?
[09:09] <Burgundavia> they write package management software
[09:10] <Hobbsee> hi pitti 
[09:10] <pitti> hey Hobbsee 
[09:10] <Hobbsee> pitti: my laptop is dying :(  i'm not happy
[09:11] <jdub> Burgundavia: who is the best person to put on a money earning project - "just anybody" or "fucking brilliant"?
[09:12] <lifeless> fucking awesome
[09:13] <jdub> lifeless: THAT WAS NOT IN THE LIST
[09:13] <lifeless> jdub: I'm a radical man
[09:13] <jdub> lifeless: THINK INSIDE THE BOXES I GIVE TO YOU
[09:14] <lifeless> She canna take it captain!
[10:01] <zyga> hey
[10:01] <ajmitch> hello
[10:08] <jdub> elmo: ping
[10:30] <carlos> Kamion: you owe me some .po files for debian-installer ;-)
[10:40] <Kamion> carlos: did you do the dapper import?
[10:40] <Kamion> IIRC I was leaving the files there for you deliberately
[10:40] <carlos> Kamion: no, are the files in place?
[10:40] <Kamion> carlos: yes, they were last time we talked as well
[10:40] <carlos> sorry, I thought you were going to ping me...
[10:40] <carlos> ok
[10:40] <carlos> let me do it
[10:40] <Kamion> carlos: ok, I've archived them to http://people.ubuntu.com/~cjwatson/installer-po-dapper/
[10:41] <Kamion> please use that, and I'll point /installer-po/ at edgy
[10:41] <carlos> ok
[10:41] <carlos> thanks
[10:43] <ogra> hmm, is there a way to excude my package from a dbgsym build ? 
[10:43] <ogra> i surely dont need that for ltsp
[10:43] <pitti> ogra: yes, there is
[10:43] <ogra> pitti, how ? 
[10:43] <pitti> ogra: NO_PKG_MANGLE=1
[10:44] <ogra> in rules ?
[10:44] <pitti> ogra: does it hurt to built with debug symbols?
[10:44] <ogra> well, its wasted energy 
[10:44] <pitti> ogra: yes, in debian/rules
[10:44] <pitti> ogra: pah, the buildds can cope :)
[10:44] <ogra> i dont have anythin in it that could have debug symbols
[10:45] <pitti> ogra: oh, if there are no ELF files at all, then there won't be dbgsym packages anyway
[10:45] <ogra> (its only a bunch of scripts)
[10:45] <pitti> if there are, that's a bug I need to fix
[10:45] <ogra> i just saw the attempt in the log
[10:45] <ogra> dpkg-deb: building package `ltsp-client-dbgsym' in `../ltsp-client-dbgsym_0.99_amd64.ddeb'.
[10:45] <pitti> oh
[10:46] <ogra> ah
[10:46] <ogra> there is this 10 line c proggy that maps printers to port 9200 mdz imported from ltsp.org
[10:46] <ogra> so there is an ELF file :)
[10:46] <pitti> ogra: getltscfg and lp_server perhaps/
[10:46] <pitti> ?
[10:46] <pitti> ah
[10:46] <ogra> i'll add the rules option
[10:47] <pitti> as you wish
[10:49] <jdub> "Elmo loves wasabi. That's why Elmo has no eyelids."
[10:53] <ogra> is there something wrong with the i386 buildd ? 
[10:53] <ogra> i wonder why my ltsp package from 7h ago is still pending
[10:53] <ogra> infinity ^^^ ?
[10:56] <Fujitsu> ogra, KDE stuff is taking its time.
[10:56] <Fujitsu> I've got 3 builds waiting, about to be 4.
[10:56] <ogra> well ...
[10:57] <Fujitsu> It's off KDE stuff now, but it'll take a while to get rid of the backlog.
[10:57] <ogra> yeah
[10:57] <ogra> 7h is quite some delay
[10:57] <Fujitsu> It is.
[10:59] <Fujitsu> We need a seperate lot of buildds for KDE! They block everything else by many hours...
[10:59] <ogra> heh
[11:01] <infinity> ogra: Before KDE, both i386 buildds were stuck on OpenOffice builds for more than 10 hours.  Blame doko.
[11:02] <ogra> infinity, thanks ... i would have blamed him anyway for fun ... :)
[11:02] <Fujitsu> Why's i386 always doing so much?
[11:02] <pitti> Fujitsu: it additionally has to build all the arch:all packages
[11:02] <ogra> because it builds all these arch all packages probably ...
[11:02] <Fujitsu> Ah.
[11:02] <Fujitsu> That's very silly.
[11:02] <Fujitsu> Has somebody considered fixing that behaviour?
[11:03] <infinity> Fujitsu: Two reasons.  1) i386 builds the "arch:all" packages.  This isn't normally a big deal, except for openoffice.org-l10n, and 2) More stuff builds successfully on i386 than anywhere else.
[11:03] <ogra> we could also drop all arch:all packages :P
[11:03] <Fujitsu> True, infinity.
[11:03] <pitti> infinity: 1) langpacks :)
[11:03] <doko> infinity: just the dapper builds, the edgy builds are coming ...
[11:03] <infinity> pitti: langpacks aren't a big deal, since they're individually quite small, so they don't cause the 10+ hour delay that ooo-l10n does.
[11:04] <ogra> as long as you let some packages through inbetween thats fine doko 
[11:04] <doko> the buildd's are idling around now ...
[11:04] <ogra> yay, it built ...
[11:05] <infinity> doko: i386 isn't.  It's still catching up. :)
[11:05] <doko> ohh
[11:06] <infinity> Anyhow, this will sort of resolve itself if and when -security builds move to soyuz, since we'll have 3 buildds of each arch again, and we rarely have a 3-package blockage.
[11:06] <infinity> Well, except on doko's uploading day.  But still.
[11:06] <sivang> morning
[11:06] <Fujitsu> Ah, I thought there were 3 each at some point...
[11:17] <Fujitsu> Oh great. vernadsky's back onto kdeutils...
[11:18] <Fujitsu> And rothera's still on kdemultimedia... KDE is TOO big!
[11:19] <ogra> seb128, would you mind to take a look at http://people.ubuntu.com/~ogra/scp-integration.patch ? its written by cbx33 and afaik ok with vuntz to add to our package
[11:20] <ogra> (its for pessulus)
[11:20] <seb128> ogra: yeah, cbx33 already talked with me about it yesterday evening
[11:20] <ogra> ah, k
[11:20] <seb128> ogra: I'll do that when I'm done with GNOME 2.16.0, the g-s-t update with some patches and the a11y patches I said I would look at next
[11:21] <seb128> ogra: ie: probably this afternoon
[11:21] <ogra> i'm just trying to sort his packages 
[11:28] <vuntz> seb128, ogra: fwiw, I didn't review the patch, but what cbx33 told me looked reasonable
[11:28] <carlos> pitti: btw, I removed the .pot exports from language packs
[11:28] <seb128> vuntz: ok, thank you
[11:28] <pitti> carlos: ok; I don't need them any more
[11:28] <seb128> vuntz: don't forget about the SoC review for peter
[11:28] <carlos> pitti: I think you told me that you don't need them anymore now that we only use Rosetta for language packs
[11:28] <carlos> ok
[11:28] <vuntz> seb128: I'm not forgetting it :-)
[11:29] <ogra> wasnt the deadline yesterday ? 
[11:29] <seb128> vuntz: gut ;)
[11:29] <seb128> ogra: no, tomorrow
[11:29] <ogra> oh
[11:29] <ogra> damned and i already reviewed him ... 
[11:29] <ogra> now he will slack the last two days ...
[11:29] <ogra> :)
[11:30] <pygi> ogra, I also thought it was yesterday :P
[11:32] <seb128> "September 8, 2006:
[11:32] <seb128> - All mentor and student evaluations due by 08:00 Pacific Daylight Time"
[11:32] <seb128> according to http://code.google.com/soc/mentorfaq.html#timeline
[11:37] <seb128> Mithrandir: around?
[11:38] <seb128> Mithrandir: have you planned to look at xkeyboard-config from Debian?
[11:39] <seb128> Mithrandir: Denis Barbier did almost 10 uploads in a month, probably a bunch of good changes that could be useful to have
[11:39] <Mithrandir> seb128: hiya.
[11:39] <Mithrandir> seb128: I keep meaning to merge stuff from there, yes.  ENOTIME so far..
[11:39] <seb128> k
[11:39] <seb128> just pointing it in case you didn't notice
[11:39] <Mithrandir> seb128: but yeah, I should probably look at their changes and merge in some of it.
[11:39] <seb128> cool, thank you
[11:43] <janimo> doko: hi, should all python packages have an __init__.py? It's not clear to me when reading the policy. The package I am making is a single importable .so w/o .py files upstream, is it worth using python-central?
[11:44] <doko> janimo: no, it's not required, but the extension should be built for all supported python versions.
[11:45] <cbx33> ping seb128 
[11:45] <seb128> cbx33: pong
[11:46] <cbx33> seb128: you interesting in seeing that pessulus update?
 ogra: I'll do that when I'm done with GNOME 2.16.0, the g-s-t update with some patches and the a11y patches I said I would look at next
[11:46] <seb128>  ogra: ie: probably this afternoon
[11:46] <cbx33> http://www.progbox.co.uk/main-additions
[11:46] <seb128> cbx33: that's the reply to ogra who already asked about like half an hour ago
[11:46] <cbx33> heheh
[11:46] <cbx33> ok
[11:46] <cbx33> thanks ogra 
[11:47] <seb128> cbx33: please open a bug on pessulus with the patch
[11:47] <cbx33> will do
[11:47] <seb128> thank you
[11:47] <seb128> hey Keybuk
[11:47] <cbx33> seb128: I have repackaged 
[11:47] <cbx33> and have the diff.tar.gz
[11:48] <ogra> mjg59, the glowing of my consoles is gone since the last upgrade, but now i have stripes and no chars there ...
[11:49] <janimo> doko: can I just assume that only current is supported or should I make 2.3 and 2.4 packages too?
[11:50] <cbx33> anyone know why this http://progbox.co.uk/scp/non-root-scp-client won't run from a launcher....if I run on terminal it's fine...and also run from launcher "in terminal"
[11:50] <cbx33> can anyone help me out...what have I done wrong
[11:50] <StevenK> cbx33: It won't get a terminal by default.
[11:50] <doko> janimo: pyversions --supported gives yo the answer
[11:50] <cbx33> I don;t want it to have a terminal
[11:50] <cbx33> I just want it to run in the background
[11:50] <cbx33> as a process
[11:51] <janimo> doko: thanks
[11:51] <ogra> cbx33, you should mention that you want it started from xdg/autostart
[11:51] <ogra> ;)
[11:51] <cbx33> ogra: it doesn't work from a normal launcher either
[11:52] <ogra> btw, why dont you start it from /etc/X11/Xsession.d ?
[11:52] <ogra> (i think the spec says that but i'm not sure)
[11:53] <cbx33> ok I'll check
[11:54] <ogra> with a number above 75 ...
[11:54] <ogra> (since you need dbus)
[11:54] <cbx33> hehe
[11:54] <ogra> hmm, no rodarvus ...
[11:54] <cbx33> no
[11:55] <cbx33> I was hoping to speak to him too
[11:55] <ogra> does anybody have an idea why i have stuff like 90x11-common_ssh-agent, 90xorg-common_ssh-agent, 99xorg-common_start and 99x11-common_start in that dir ? 
[11:55] <ogra> they are identical scripts
[11:56] <gnomefreak> mjg59: are you around?
[11:56] <ogra> one from xinit and the other from x11-common
[12:01] <Kamion> old conffiles that haven't been purged, maybe?
[12:01] <ogra> well, dpkg -S shows them to belong to the packages ..
[12:01] <ogra> do we really have conffiles in .d dirs ? 
[12:01] <Mithrandir> ogra: because x11-common hasn't cleaned them up.  It should, really.
[12:01] <ogra> yeah
[12:01] <Mithrandir> ogra: they don't do any harm, though
[12:02] <Mithrandir> (apart from wasting 0.1s of your precious login time)
[12:02] <ogra> right, i was just wondering since i had to poke around in that dir a lot yesterday
[12:02] <ogra> and it looked a bit weird
[12:14] <cbx33> ogra, just fyi, the X11 didn't work either.  :S
[12:15] <ogra> but its the right place for it ...
[12:15] <cbx33> ok
[12:15] <cbx33> I'll move it to there
[12:15] <cbx33> and figure out why it's not working
[12:15] <ogra> if you put it in autostart only desktops that use xdg will be able to use it
[12:16] <cbx33> pygi can't see a problem with it
[12:16] <cbx33> ahhh. ok thanks ogra 
[12:17] <janimo> when adding an ubuntu chaneg to a native debian package do I just append ubuntu1 to the current one?
[12:18] <cbx33> ogra: I found the problem
[12:18] <cbx33> working on a fix now
[12:18] <ogra> janimo, yes
[12:24] <janimo> ogra: thanks
[12:33] <cbx33> ogra: FIXED !
[12:33] <cbx33> rebuilding package now
[12:33] <cbx33> will have to post it when I get home
[12:33] <cbx33> will finish up MIR :D
[12:33] <ogra> dont forget to update the changelog entry
[12:33] <cbx33> so pleased that it's fixed, that was doing my nut - of course not
[12:33] <ogra> best is to pull the package from universe now
[12:33] <cbx33> ogra: is this now 0.5 ?
[12:33] <pygi> cbx33, it's still 0.4 I would say :)
[12:34] <ogra> will be, yes
[12:34] <cbx33> ok
[12:34] <pygi> oh, 0.5 :)
[12:34] <ogra> since its a native package
[12:34] <cbx33> 0.4ubuntu1
[12:34] <ogra> no
[12:34] <cbx33> oh right
[12:34] <cbx33> ok 0.5
[12:34] <ogra> no need for the ubuntuX 
[12:34] <cbx33> sorry ogra thought you were agreeing with pygi
[12:34] <ogra> but base on the package from universe
[12:34] <cbx33> of course
[12:34] <cbx33> I always do
[12:34] <ogra> i made two very minor changes to yours
[12:34] <cbx33> ok. I'll check them out
[12:34] <cbx33> problematic?
[12:35] <ogra> (your package still had the edgy distro, and didnt have the updated grep string for the userlist set)
[12:35] <ogra> s/set/yet/
[12:35] <cbx33> ah 
[12:35] <cbx33> thanks ogra 
[12:35] <cbx33> you're a start...really !
[12:35] <cbx33> s/start/star
[12:35] <ogra> me ? 
[12:35] <cbx33> yes
[12:35] <ogra> who did all the wokr ? 
[12:35] <ogra> *work 
[12:35] <cbx33> heh
[12:36] <ogra> that wasnt me :)
[12:36] <ogra> youre the awesome guy here ... dont hide your credits ;)
[12:36] <cbx33> Keybuk: too many, too little time?
[12:36] <cbx33> ogra: it's funny in other communities, I've tried to show people what I've done and been told no one cares.....but here I get told to show off more :p
[12:36] <cbx33> heheh
[12:36] <cbx33> thanks for the encouragement ogra 
[12:37] <ogra> thaks for all the hard work :)
[12:37] <ogra> *thanks too
[12:37] <Keybuk> cbx33: trying to decide whether to try and complete anything else before FF
[12:37] <cbx33> do you want me to use 0.4 or 0.41
[12:37] <cbx33> Keybuk: I know that feeling...on a much smaller scale
[12:37] <Fade> is anybody else experience a serious crash in xemacs, and a similar non-crashing bug in emacs?
[12:37] <Keybuk> I could try and port some of initscripts to upstart, but that runs the risk of dropping support for md/lvm/evms/etc.
[12:37] <Fade> on edgy
[12:37] <cbx33> 0.5 or 0.41 as it's just a tiny change?
[12:37] <Keybuk> that may be considered bad
[12:37] <ogra> cbx33, as you like :)
[12:38] <cbx33> ok
[12:38] <ogra> i consider it more your package than mine already ;)
[12:38] <Mithrandir> Fade: can you see if you still see that problem in a couple of hours?  I just uploaded a fix for Xlib which I believe might be the cause of some emacs problems.
[12:38] <ogra> so feel free to make the decidions
[12:38] <ajmitch> Keybuk: yes, I like being able to boot
[12:38] <cbx33> heh, ogra you da man !
[12:38] <Fade> Mithrandir: that would be awesome.
[12:38] <Fade> sure
[12:39] <Keybuk> ajmitch: don't suppose you could send me your /var/log/udev ?
[12:39] <ajmitch> Keybuk: sure, one for LVM on RAID, or just LVM?
[12:40] <Fade> the c stack trace seems to indicate xlib
[12:40] <Mithrandir> Fade: once you have libx11-6 2:1.0.3-0ubuntu3 installed, try to reproduce and if you can, it was something else, if not, yay. :-)
[12:40] <Fade> this is my report with traces on the xemacs package: https://launchpad.net/distros/ubuntu/+source/xemacs21/+bug/58856
[12:40] <Ubugtu> Malone bug 58856 in xemacs21 "xemacs segfaults on edgy powerpc system" [Untriaged,Unconfirmed]  
[12:41] <gnomefreak> Keybuk: you gonna be around for a little while?
[12:41] <Kamion> Mithrandir: should live-cd-write-as-you-go be deferred?
[12:41] <ajmitch> Keybuk: http://ajmitch.dyndns.org/~ajmitch/udev.lvm has output from my laptop using LVM
[12:41] <Keybuk> ajmitch: either, both
[12:41] <Keybuk> gnomefreak: umm, why?
[12:41] <gnomefreak> does upstart affect tty's?
[12:41] <gnomefreak> or can
[12:41] <Keybuk> "affect" ?
[12:42] <gnomefreak> blinking curser on all tty's
[12:42] <Mithrandir> Kamion: yes, I'll mark it so now.
[12:42] <Fujitsu> My tty1 went missing, but my others are fine.
[12:42] <Keybuk> gnomefreak: I've had a report of that
[12:42] <gnomefreak> ok cool
[12:42] <Keybuk> Fujitsu: is it still missing?
[12:42] <Mithrandir> Kamion: can we have the health check reports say how oversized any images are when they are?
[12:42] <Fujitsu> Keybuk, yes, I last upgraded yesterday...
[12:42] <Keybuk> Fujitsu: ok, what happens if you reboot without "splash" on the kernel command-line
[12:42] <ajmitch> Keybuk: http://ajmitch.dyndns.org/~ajmitch/udev.lvm_raid has output from my desktop box
[12:43] <Fujitsu> I'll try that in a couple of minutes, Keybuk.
[12:43] <Keybuk> ajmitch: what does your laptop mount as root= ?
[12:43] <ajmitch> root=/dev/mapper/acer--vg-ubuntu
[12:44] <Keybuk> right
[12:44] <ajmitch> works fine at the moment with upstart
[12:44] <Keybuk> that's because upstart isn't do anything special at the moment
[12:44] <Keybuk> if we changed the filesystem mounting to be event-based, yours would fail, because LVM is still in the dark ages and doesn't use udev
[12:44] <ajmitch> great :)
[12:45] <Keybuk> what does your desktop mount as root=?
[12:47] <cbx33> ogra: https://wiki.ubuntu.com/MainInclusionStudentControlPanel does that look better now?
[12:47] <ajmitch> it's been awhile since it was rebooted
[12:47] <ajmitch> ah, it's using UUIDs
[12:48] <ajmitch> however it may not have been rebooted since that change was made :)
[12:48] <fabbione> ajmitch: you can check with /proc/cmdline or something like that
[12:48] <fabbione> it will tell you the options passed to the kernel
[12:48] <fabbione> including root=
[12:49] <ajmitch> thanks
[12:49] <ajmitch> root=/dev/mapper/ubuntu-root
[12:51] <Keybuk> yeah, would have the same problem then
[12:51] <Keybuk> I think the right thing to do here is to make an edgy+1 goal of "ReplacementInitscripts" and have more time to spend on it
[12:51] <Keybuk> and include in that spec that dev-mapper and mdadm must be made to work with udev
[12:52] <Keybuk> given a couple of months, that's definitely possible, which it isn't in a couple of days
[01:01] <Kamion> Mithrandir: good idea; done for the next check
[01:02] <Keybuk> Kamion: do you think that's sane?  defer that work to edgy+1 so we can spend more time on it?
[01:03] <peterz> are you guys aware current edgy fails to build uml kernels?
[01:03] <Kamion> Keybuk: boot-message-logging isn't possible without replacement-initscripts?
[01:03] <Keybuk> Kamion: not sure yet, evaluating that now
[01:03] <Kamion> Keybuk: I agree though, there was always going to be some part that got deferred and AIUI it was always going to be serious initscript conversion
[01:04] <Kamion> it's a shame that we can't fix fstab handling in edgy, but there you go
[01:04] <Keybuk> Kamion: we could fix it for the common/sensible case of people just using partitions and ordinary filesystems ... but at the cost of dropping support for the "exciting" ones
[01:05] <Keybuk> I'm sure somebody would suggest that edgy is a good time to do that temporarily, while dapper is still out
[01:05] <elmo> err, what?
[01:05] <Keybuk> but I think that it'd be better to work it out so we have a very easy, step-by-step plan for edgy+1 instead
[01:05] <elmo> keybuk: are you seriously suggesting we drop support for root sw-raid?
[01:06] <Keybuk> elmo: no, I'm suggesting seriously that we be more conservative and *don't* drop support for it, at the cost of deferring part of a spec
[01:07] <elmo> Keybuk: ok, but please don't refer to root sw-raid as either uncommon, not sensible or "exciting", as I really don't think it's any of those
[01:07] <Keybuk> elmo: ok, "crufty" ? :p
[01:07] <Keybuk> "poorly maintained" ?
[01:08] <Kamion> Mithrandir: believe it or not I'm *still* running gen_keymap
[01:09] <madduck> Keybuk: poorly maintained?
[01:09] <elmo> Keybuk: poorly maintained in ubuntu maybe, but that just reinforces my fears about ubuntu on servers :-p
[01:09] <Mithrandir> Kamion: sheesh, that's a bit excessive.
[01:10] <Keybuk> and poorly maintained upstream ... given none of these systems have been updated to expose their block devices correctly, despite this changing a couple of years ago in the kernel
[01:11] <elmo> yeah, because udev is such a shining beacon of good maintenance *snort*
[01:11] <madduck> Keybuk: i am trying... http://lists.alioth.debian.org/pipermail/pkg-mdadm-devel/2006-August/000859.html
[01:11] <madduck> Keybuk: as of right now, mdadm and udev/lvm work fine together, at least in Debian.
[01:11] <Kamion> hmm, gen_keymap wants me to pick one of Bosnia/Herzegovina, Croatia, and Slovenia unicode keymaps as the default for that set of branches in the decision tree. Which is least likely to get me assassinated?
[01:12] <pygi> Kamion, Croatia
[01:12] <pygi> I can help you there :)
[01:12] <madduck> ubuntu should really upgrade mdadm and *not* ship 2.4.1 i think.
[01:12] <Keybuk> elmo: *shrug* they're a bit prone to dropping backwards compatibility
[01:12] <madduck> Keybuk: apart, what's poor about upstream?
[01:12] <Mithrandir> Kamion: iirc, they're the same so it doesn't really matter.  We might want to default to the first, alphabetically.
[01:13] <Kamion> Mithrandir: they're sufficiently close that gen_keymap can't tell them apart, anyway
[01:13] <Keybuk> madduck: "apart" ?  that's all I've mentioned
[01:13] <Kamion> hmm, yes, they're all "include cs(latinunicode)"
[01:13] <pygi> Kamion, /me really suggests Croatian :P
[01:13] <madduck> Keybuk: apart from my suggestion to upgrade
[01:14] <Keybuk> madduck: you make no sense
[01:14] <Kamion> I'll pick cs:latinunicode if it lets me, since that's the structure of the X keymaps
[01:14] <Kamion> pygi: biased opinions don't count :)
[01:14] <pygi> Kamion, ergh! :P
[01:14] <madduck> Keybuk: let's not get hung up on english, okay?
[01:14] <madduck> Keybuk: what's poor about mdadm upstream?
[01:14] <elmo> madduck: we're way past UVF - what would upgrading mdadm now give us?  it certainly wouldn't fix the problem we're discussing (upstart and root sw-raid/dm)
[01:15] <Keybuk> madduck: the fact that even after two years, they've not updated their code to work properly with udev/driver-core
[01:15] <Keybuk> (ah, I believe we've reached the beginning of this conversation)
[01:15] <pitti> rodarvus: is there a gitweb of x.org?
[01:16] <madduck> elmo: ok. tbh, the changelog is a couple of hundred lines.
[01:16] <madduck> elmo: i would not touch 2.4.1 with a stick anymore.
[01:17] <madduck> Keybuk: as said, i am working on it.
[01:17] <pitti> rodarvus: nevermind, got it
[01:17] <Keybuk> madduck: indeed
[01:17] <Keybuk> I never said you weren't :p
[01:18] <Keybuk> and I even said I'd work on the other things to make sure they're all fluffy
[01:18] <Keybuk> just not in 48 hours :p
[01:18] <Keybuk> PUT YOUR KNEES AWAY! :p
[01:18] <rodarvus> pitti, yes, http://gitweb.freedesktop.org/
[01:18] <rodarvus> oh
[01:18] <rodarvus> (/me should read all backlog before typing stuff :) )
[01:22] <ans-tor> Is there any packages in debian that doesn't exist in ubuntu?
[01:23] <Hobbsee> plenty
[01:23] <Keybuk> "plenty" ?
[01:23] <Keybuk> it's about 15-20 iirc
[01:23] <Kamion> yes, for three reasons: (1) we're past upstream version freeze and new packages have been introduced in Debian since then, (2) we didn't quite get round to syncing in all new packages pre-UVF, (3) there are a few packages we've intentionally removed because they don't work for us
[01:24] <ans-tor> ubuntu/universe doesn't contain all debian packages?
[01:24] <Kamion> most, not all
[01:24] <Kamion> the ones we've deliberately removed wouldn't work anyway
[01:24] <Kamion> the rest are just too new, and will come into universe for edgy+1 at the latest
[01:25] <segfault> morning. :-)
[01:25] <Kamion> wouldn't work> for example, there's no point in linux-kernel-di-i386-2.6 (kernel packaging machinery for the installer), since our own linux-source-2.6.17 package does all that work itself
[01:26] <ans-tor> not work in ubuntu but work in debian?  that's so weird.
[01:26] <Kamion> the Ubuntu kernel is packaged quite differently from the Debian kernel, for maintenance workflow reasons
[01:26] <ans-tor> opps, i see.
[01:26] <Hobbsee> Keybuk: oh, i thought there were more than that.
[01:26] <Hobbsee> Keybuk: dont mind me
[01:26] <Kamion> although they do share a fair amount of common code even in the packaging, of course
[01:27] <Kamion> Keybuk: that sounds like the count of deliberately-removed-and-blacklisted packages
[01:31] <Keybuk> Kamion: err, yes
[01:31] <Keybuk> lp_archive@drescher:~$ new-source | wc -l
[01:31] <Keybuk> 283
[01:31] <Keybuk> lp_archive@drescher:~$ new-source contrib | wc -l
[01:31] <Keybuk> 34
[01:31] <Keybuk> lp_archive@drescher:~$ new-source non-free | wc -l
[01:31] <Keybuk> 64
[01:33] <janimo> pitti, do you know if the "printer id" line at the start of foomatioc xml files should be exactly the same as the file name?
[01:34] <pitti> janimo: no, I don't know, sorry
[01:34] <janimo> there's one exception to this in the db now and I don't know if it's a bug or it is valid (to state that the devices are compatible or something like that)
[01:34] <pitti> janimo: once till is online, you should ask him
[01:34] <janimo> pitti: ok. Ia mhaving a look again at the fedore printing tools
[01:35] <janimo> turns out that the one we looked at in paris was old, they were writeing a new one for FC6
[01:35] <pitti> janimo: btw, till had the idea to port printerdrake to ubuntu
[01:37] <janimo> pitti: with those nastly mandrake/perl libs?
[01:37] <pitti> I hope not :)
[01:37] <pitti> let's see
[01:38] <janimo> pitti: is it ok with you if I upload python-cups (the fedora bindins library) to universe?
[01:39] <pitti> janimo: of course, if it is free software
[01:39] <pitti> sounds useful in any case
[01:39] <janimo> pitti, ok, I was not sure if ivoks worked on something similar or not
[01:46] <Kamion> use dpkg-divert
[01:47] <Kamion> or, more sensibly, edit the conffile that calls updatedb
[01:47] <Kamion> it's a conffile for a reason
[01:49] <bluefoxicy> Kamion:  there's a conffile?
[01:49] <bluefoxicy> it looks to be a crontab
[01:50] <Kamion> /etc/cron.daily/*
[01:51] <bluefoxicy> won't that get replaced on upgrade?
[01:51] <Kamion> no
[01:51] <bluefoxicy> (granted so will the binary)
[01:51] <Kamion> not if you've modified it
[01:51] <Kamion> you'll get a dpkg conffile prompt
[01:51] <Kamion> how long have you been using Debian/Ubuntu systems exactly? ;)
[01:51] <bluefoxicy> I've never touched cron
[01:51] <Kamion> or anything in /etc, apparently
[01:52] <Kamion> well, not quite anything, but the bulk of files there are conffiles
[01:52] <bluefoxicy> on gentoo certain things in /etc/ weren't treated as config files
[01:52] <bluefoxicy> I'd assumed crontabs were more system scripts than configuration
[01:52] <Mithrandir> well, this isn't gentoo
[01:52] <ogra> Kamion, just to give you an impression of the triviallity of the ltsp-dhcpd stuff : http://people.ubuntu.com/~ogra/netscript
[01:53] <Kamion> bluefoxicy: if they were, they would not be in /etc
[01:53] <ogra> i need one debconf selection and one message now
[01:53] <ogra> adding it to the ltsp-client-builder and im done ...
[01:53] <bluefoxicy> Kamion:  they'd be in /etc or /var
[01:53] <Kamion> ogra: you can't use awk in d-i
[01:53] <ogra> hmm, it worked on on the d-i console
[01:54] <bluefoxicy> anyway interview time
[01:54] <ogra> i tested it 
[01:54] <Kamion> busybox-1.1.3/debian/config-udeb:# CONFIG_AWK is not set
[01:54] <Kamion> ogra: oh, is this code running inside a chroot /target?
[01:54] <ogra> nope
[01:54] <ogra> thast for the server
[01:54] <Kamion> I'm not sure how it worked for you, then :)
[01:55] <ogra> but i ran it on tty2 on the CD
[01:55] <ogra> hmm
[01:55] <Kamion> if [ ! $iface = $DEFAULTROUTE_IFACE ] ; then 
[01:55] <Kamion> that should be:
[01:55] <Kamion> if [ "$iface" != "$DEFAULTROUTE_IFACE" ] ; then
[01:55] <ogra> right
[01:55] <Kamion> if [ $(echo ${IFACE_LIST}|wc -w) -gt 1 ] ; then 
[01:55] <Kamion> quotes around the $(...), same elsewhere
[01:55] <ogra> the condition was the other way around before ...
[01:56] <Kamion> it's more the quoting I'm commenting on - you need to be especially careful with quoting in [ arguments because it will fall over messily if you don't
[01:56] <ogra> ok
[01:56] <Kamion> I assume you'll be replacing that use of select with a debconf select template
[01:56] <ogra> yep
[01:56] <ogra> thats why it still is /bin/bash
[01:57] <Kamion> maybe put an extra blank line after the cat <<EOF
[01:57] <ogra> ok
[01:57] <Kamion> will look neater in interfaces
[01:57] <Kamion> elif [ $(echo ${IFACE_LIST}|wc -w) -lt 1 ] ;then
[01:58] <Kamion> you can't do [ -z "$IFACE_LIST" ]  ? :-)
[01:58] <Kamion> I guess it might contain just whitespace, dunno
[01:58] <ogra> it will ...
[01:58] <ogra> IFACE_TMP="$IFACE_TMP $iface" adds whitespace 
[01:59] <Kamion> oh, that's easily fixed
[01:59] <Kamion> IFACE_TMP="${IFACE_TMP:+$IFACE_TMP }$iface"
[01:59] <ogra> scary quoting :)
[02:00] <Kamion> ${var:+string} does "if var is set, then expand string, else nothing"
[02:00] <Kamion> set and non-empty actually
[02:00] <Kamion> ok, otherwise looks plausible. you don't need to write to /etc/ltsp/dhcpd.conf from that script, just /etc/n/i?
[02:01] <ogra> yep
[02:01] <ogra> we keep /etc/ltsp/dhcpd.conf as is ...
[02:01] <ogra> i'll add a grep 192.168.0. /etc/network/interfaces at the top as well ...
[02:01] <ogra> and let it exit if it finds something ...
[02:02] <ogra> but that should be about it 
[02:02] <Kamion> or pick a different subnet
[02:02] <ogra> then i'd need to touch dhcpd.conf
[02:02] <ogra> which we dont want ...
[02:02] <Kamion> ok
[02:02] <ogra> i'll just fall back to the echo "error: no interfaces found please configure /etc/ltsp/dhcpd.conf manually after install"
[02:02] <Kamion> what happens if there's only one interface?
[02:02] <ogra> the same 
[02:02] <Kamion> I'd be inclined to pick something that's less commonly used than 192.168.0.*, if possible
[02:02] <ogra> it shows that message
[02:03] <ogra> right
[02:03] <Kamion> pick a random one from one of the bigger private-use ranges
[02:03] <ogra> i thought about 10.0.0 or something
[02:03] <Kamion> it's 10.0.0.0/24, so you wouldn't need to start with 10.0.0., even
[02:03] <Kamion> er, /8
[02:04] <ogra> well, something from the 10. range ...
[02:04] <Fujitsu> Or 172.16.0.0/16.
[02:04] <Kamion> Fujitsu: /12
[02:04] <Kamion> (RFC1918 s.3)
[02:04] <Keybuk> it's amazing how many people get that one wrong :p
[02:05] <Fujitsu> :$
[02:05] <Keybuk> 10 = Class A, 172.16 = Class B, 192.168 = Class C
[02:05] <Fujitsu> B = /16!
[02:05] <Keybuk> of course, the Internet has "grown up" and the Class System is merely a memory
[02:05] <Keybuk> no, B = /12
[02:06] <Keybuk> (top-level allocations, not sub allocations)
[02:06] <ans-tor> ubuntulog: hi
[02:06] <ans-tor> ubuntulog: ?
[02:06] <ans-tor> ubuntulog: help
[02:06] <Treenaks> Keybuk: 'The correct modern representation for what would have been referred to as a "Class B" prior to 1993 would be "a set of /16 addresses"'
[02:06] <Keybuk> err, maybe I'm confused there then
[02:07] <Fujitsu> I thought so.
[02:07] <Fujitsu> Cisco teaches B = /16.
[02:07] <Treenaks> They still TEACH it, even though the internet has been classless for 13 years now?
[02:07] <Keybuk> Treenaks: most Cisco kit still IMPLEMENTS it
[02:08] <thom> Keybuk: yeah; 172.16 is 16 /16s (16 class Bs)
[02:08] <Treenaks> *shudder*
[02:08] <mdeslaur> rfc 791 defines a class B to be /16
[02:08] <Keybuk> thom: that's what I thought
[02:09] <Keybuk> and 192.168 is 256 class Cs
[02:09] <thom> so you're right to say it's a /12, but wrong to say it's a B
[02:09] <Keybuk> ahh, yes, I got myself confused along the way :p
[02:09] <thom> yup
[02:09] <Keybuk> 172.16/12 is the "set of private Class B networks", and isn't itself class B
[02:09] <Keybuk> ok
[02:10] <thom> Keybuk: when you stop being confused, can you make sysvinit non-Essential so apt'll stop trying to remove upstart? :_)
[02:11] <Fujitsu> Yes, that'd be nice...'
[02:11] <Keybuk> thom: sysvinit is non-Essential
[02:11] <Keybuk> upgrade sysvinit ;)
[02:11] <Keybuk> and then APT will be happy
[02:11] <thom> urgh
[02:11] <Keybuk> then you can file a bug on APT for failing to take into account changes in the Essential field
[02:11] <thom> oh what a pain
[02:11] <Fujitsu> Is ubuntu-minimal able to be installed with upstart yet?
[02:12] <thom> Removing upstart ...
[02:12] <thom> dpkg (subprocess): unable to execute post-removal script: Exec format error
[02:12] <thom> dpkg: error processing upstart (--remove):
[02:12] <Kamion> there's an argument that it shouldn't take them into account until it's upgraded them anyway ...
[02:12] <Keybuk> thom: stick a #!/bin/sh on the front of the postrm, sorry
[02:12] <thom> slacker :-)
[02:13] <Kamion> Keybuk: ISTR that there's special handling in dpkg for one Essential package C/Ring another, or maybe C/R/P
[02:13] <Kamion> for exactly this reason
[02:13] <Kamion> however upstart does not appear to be Essential
[02:13] <Kamion> I suspect if you make it so, it'll work better ...
[02:14] <Keybuk> Kamion: yeah, that will happen in my next upload
[02:27] <hunger> Do ttys work whith upstart again?
[02:27] <Keybuk> hunger: it appears to be a usplash bug
[02:27] <Keybuk> or, at least, a bad interaction
[02:28] <Keybuk> people report that if they downgrade usplash, their ttys come back
[02:28] <hunger> Keybuk: I have no usplash and still no ttys.
[02:28] <Keybuk> hunger: then perhaps you'd like to reply to my message on the bug report ?
[02:28] <Keybuk> hunger: can you do "sudo status tty1 tty2 tty3 tty4 tty5 tty6" for me
[02:28] <hunger> Keybuk: Damn... my mailserver must be down again:-(
[02:28] <Keybuk> hunger: no worries, we can debug here :p
[02:29] <hunger> Keybuk: Not really:-) I do not have upstart installed anymore.
[02:29] <Keybuk> hunger: then I'm afraid I'll have to Reject your bug
[02:29] <hunger> Keybuk: One sec... I'll go and break my system again by installing upstart.
[02:29] <Keybuk> :)
[02:29] <Keybuk> thanks
[02:30] <hunger> Keybuk: No IRC while I am in an upstarted system though.
[02:30] <Keybuk> why no IRC?
[02:31] <thom> Keybuk: in answer to your question about cryptdisks and upslash, i've long since turned usplash off for precisely that reason
[02:31] <hunger> Keybuk: Because my homedirs are gone, so is /usr and all the other stuff that I have encrypted.
[02:31] <Keybuk> thom: <g>  I did have a momentary cunning hack thought that one could just < /dev/console the script
[02:31] <Keybuk> hunger: ok, there's a temporary workaround for that
[02:31] <Keybuk> install upstart, upstart-compat-sysv (and make sure sysv-rc and initscripts are also installed)
[02:32] <Kamion> pitti: apt-get-debug-symbols presumably isn't happening for edgy
[02:32] <Kamion> ?
[02:32] <thom> Keybuk: eh, evil person :-)
[02:32] <Keybuk> then edit /etc/event.d/rcS and /etc/event.d/rc2 and add "console output" to both files above the "script" line
[02:32] <Keybuk> thom: evil?
[02:33] <thom> well, hacky anyway. but yes, that'd be a short term fix for the problem
[02:33] <hunger> Keybuk: All of those were installed.
[02:33] <Keybuk> hunger: ok
[02:34] <hunger> Keybuk: One sec... I'll reboot soonish.
[02:34] <hunger> Keybuk: Just need to backup some data first.
[02:35] <Keybuk> (also on a more general note, I've noticed that you almost consistently file bugs and then revert whatever is causing your problems ... it's useful for us, if you run the development release, to keep it broken while we debug it; otherwise we simply can't obtain any information about the bug, and assuming nobody else has the same problem, have no choice but to assume it's not a bug)
[02:36] <Kamion> Riddell: I noticed a couple of things about your accessibility patch to caper
[02:36] <Kamion> casper
[02:36] <hunger> Keybuk: I only have this one computer. I can not keep it broken till somebody stops by to fix it.
[02:36] <Kamion> Riddell: one is that "impairment" is consistently misspelled "imparement". Is it possible to fix that?
[02:37] <Riddell> Kamion: certainly
[02:37] <hunger> Keybuk: I most of the time need to revert the problem to be even able to report it in the first place.
[02:37] <Keybuk> hunger: we can't fix it unless you keep it broken
[02:37] <Keybuk> it's a not unreasonable assumption that it must be working for most other people, after all
[02:37] <Kamion> Riddell: the other is that you've got an entry for access=m2, whereas that doesn't seem to be mentioned in bug 58836. Do you want me to leave access=m2 switched on for Kubuntu?
[02:37] <Ubugtu> Malone bug 58836 in casper "F5 options need to be linked to the right casper options" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/58836
[02:38] <hunger> Keybuk: I am willing to rebreak if somebody is working on the fix. But I will not keep things broken till somebody finds the time to look into the bug I had.
[02:39] <jdub> Kamion: what's richard weideman's nick?
[02:39] <ogra> jdub, RichEd
[02:39] <jdub> thanks
[02:39] <ogra> (he's in -meeting atm)
[02:39] <Keybuk> hunger: ok, ready to reboot ?
[02:39] <Mithrandir> Riddell: I've fixed casper to use impairment now.
[02:39] <Kamion> Riddell: (access=m2 being on-screen keyboard)
[02:40] <hunger> keybuk: 10min max.
[02:40] <mojo> isn't it me or my computer stuffed up? the gtkfile chooser dialog of "Save as.." of any programs missing Home and Desktop icons? am i gone mad?
[02:40] <Keybuk> hunger: make sure you modify /etc/event.d/rcS and rc2, that'll give you the usual boot messages, etc.
[02:40] <hunger> Keybuk: I did.
[02:40] <hunger> Keybuk: rebooting....
[02:41] <Riddell> Kamion: the kubunut m2 has the mouse tool but no on screen keyboard
[02:41] <Kamion> I can't change the text at the moment thoughh
[02:41] <jono> David: By the way Jono pass on my "oh my god this rocks" message to the ubuntu team would you Edgy is awesome way better than dapper was this far into it's development.
[02:41] <jono> :)
[02:42] <Kamion> heno: perhaps I could revert "On-Screen Keyboard" to more generic text? any suggestions?
[02:42] <Riddell> Kamion: or I can just have the mouse tool on in m1 and not have an m2
[02:42] <heno> Kamion: Virtual keyboard, alternative input ?
[02:42] <mojo> heno: On-screen input?
[02:43] <Kamion> what exactly is the mouse tool?
[02:43] <Riddell> Kamion: it clicks the mouse if you hover over a button for some time
[02:44] <heno> Useful if you cannot click. Some people with RSI use it as well
[02:44] <heno> Riddell: does mousetool have other features than that?
[02:44] <Riddell> heno: nope
[02:45] <Riddell> I don't think it's worth having a confusing m2 text just so it fits both keyboard and mouse tools
[02:45] <heno> I seem to remember that the Windows version did, but I could be wrong
[02:46] <Riddell> Kamion: so I'd say don't have m2 for kubuntu and move the sed line from m2 to m1 in casper
[02:46] <Kamion> Riddell: ok
[02:49] <hunger_suse> Keybuk: What should I run again?
[02:50] <Keybuk> hunger_suse: how far did you/it get?
[02:50] <hunger_suse> Keybuk: The tty scripts all give a comment "expected 'on' or 'script'".
[02:50] <hunger_suse> Keybuk: upstart dropped me into a root shell (without needing a password!).
[02:50] <Keybuk> hmm, can you mount that filesystem from where you are now?
[02:51] <pitti> Kamion: it's installed on the buildds for quite a while
[02:51] <hunger_suse> Keybuk: I am online from another box.
[02:51] <pitti> Kamion: infinity and I have a plan how to get them away from the buildds to people
[02:51] <Keybuk> hunger_suse: ok, it sounds like you have a very old version of those config files installed
[02:51] <hunger_suse> Keybuk: I got the upstart-ubuntu box running.
[02:51] <Keybuk> ahh
[02:51] <pitti> Kamion: until soyuz has proper support for them (!edgy)
[02:51] <Keybuk> cat /etc/event.d/tty1
[02:51] <Kamion> pitti: perhaps the spec should be updated to talk about the workaround rather than soyuz work
[02:51] <pitti> Kamion: however, the cowboying should happen for edgy
[02:51] <Keybuk> make sure it has "# tty1 - getty" at the top 
[02:51] <pitti> Kamion: right, I'll update the whiteboard
[02:52] <Kamion> thanks; is it beta-available now then, or not?
[02:52] <hunger_suse> Keybuk: start when rc2 is stopping stop on shutdown stop on reboot respawn /sbin/...
[02:52] <pitti> Kamion: not sure about the appropriate status; ddebs are built and thrown away, so ATM it's not even 'beta'
[02:52] <Keybuk> hunger_suse: right, that's an old version of the script
[02:52] <Keybuk> no wonder it fails :p
[02:52] <Keybuk> did you modify them after installing originally, and then say "N" on upgrade?
[02:52] <pitti> Kamion: as soon as we put them on people, it's implemented (well, in an ugly way, but still working)
[02:52] <hunger_suse> I have then newest versions of everything installed.
[02:52] <Kamion> pitti: oh. how much can happen before FF?
[02:53] <Keybuk> hunger_suse: dpkg-query -W upstart upstart-compat-sysv
[02:53] <hunger_suse> Keybuk: I have not modified anything and never was asked to upgrade the files.
[02:53] <Keybuk> hunger_suse: which versions do you have?
[02:53] <pitti> Kamion: the buildd scripting will happen afterwards, infinity is too busy ATM
[02:53] <pitti> Kamion: but it doesn't affect the archive nor release in any way
[02:53] <Kamion> pitti: ok, can we talk with mdz about that? the important factor is more going to be scheduling people's time after FF, I think
[02:53] <pitti> Kamion: i. e. it's infrastructure and can happen independently from the release cycle
[02:53] <hunger_suse> Keybuk: No such command.
[02:53] <Kamion> to ensure you have enough time to fix bugs
[02:53] <pitti> Kamion: sure, he will join tomorrow's distro team meeting I presume
[02:54] <Kamion> yes, I believe so
[02:54] <kagou> jenda: like zul, and also on www.debian.org/doc/manuals/maint-guide/
[02:54] <Kamion> he said he'd be around this evening having crashed for a while
[02:54] <pitti> Kamion: I guess we will use that to discuss spec statuses?
[02:54] <Kamion> yes
[02:54] <hunger_suse> Keybuk: dpkg info is not available anyway.
[02:54] <giftnudel> hunger_suse: can't you mount your partitions?
[02:54] <Keybuk> hunger_suse: why is dpkg info not available?
[02:54] <hunger_suse> Because the partition that is on /var is not available.
[02:55] <Keybuk> why is it not available?
[02:55] <hunger_suse> Because upstart can not decrypt it.
[02:55] <Keybuk> why not?
[02:55] <Kamion> you can't mount it by hand?
[02:56] <Keybuk> hunger_suse: ok, can you please mount your /usr and /var (and any other FHS partitions) by hand at the root shell it gave you
[02:56] <Keybuk> we'll need those for debugging
[02:56] <thom> hunger_suse: /etc/init.d/cryptdisks start
[02:56] <pitti> Kamion: I set it to 'deployment' and updated the whiteboard
[02:56] <thom> if they're crypted disks
[02:56] <Kamion> pitti: righto, sounds good
[02:56] <Kamion> thanks
[02:56] <pitti> thanks for the reminder
[02:56] <Kamion>  7596 cjwatson  25   0  142m 121m 1884 R 99.9 12.1  31:09.77 gen_keymap
[02:56] <Kamion> this really doesn't look good
[02:57] <thom> um, is that for a keyboard with more keys than firefox has bugs?
[02:58] <hunger_suse> Keybuk: Both upsatrt and the compat package are at 0.2.1-5
[02:58] <Keybuk> hunger_suse: ok, thanks ... could you also cat /etc/event.d/rc2 for me
[02:58] <Keybuk> it should begin "# rc2 - runlevel 2 compatibility", does it?
[02:58] <Kamion> thom: that's ALL keymaps
[02:59] <hunger_suse> Keybuk: That one has a script section.
[02:59] <Keybuk> hunger_suse: does it have the first line I asked about?
[02:59] <hunger_suse> Keybuk: It does.
[03:00] <Keybuk> hunger_suse: ls /etc/event.d ... what files are in there?
[03:00] <hunger_suse> rcS starts with the same line but with rcS in place of rc2.
[03:00] <thom> Kamion: cripes
[03:01] <hunger_suse> Keybuk: tty[1-6]  rc-default rc0-halt rc0-poweroff rc[1-6]  rcS
[03:01] <Kamion> thom: (for the magic keymap detector in the installer)
[03:01] <Keybuk> hunger_suse: ok
[03:02] <Keybuk> hunger_suse: /etc/init.d/udev start ... with any luck that should bring the networking up ... then can you copy the /etc/event.d/tty1 script somewhere so you can paste it here
[03:02] <hunger_suse> Keybuk: You won't get through the firewall or I would just give you an account on that box.
[03:02] <Keybuk> I don't need to get through the firewall, I just need you to paste me a copy of that file :p
[03:02] <hunger_suse> Keybuk: /etc/init.d/udev start fails.
[03:02] <Keybuk> "fails" ?
[03:03] <Keybuk> try "/etc/init.d/mountkernfs.sh start" then try again
[03:03] <hunger_suse> Keybuk: Sorry, the ubuntu init scripts suck at error reporting... it just gives "[fail] ".
[03:03] <Keybuk> hunger_suse: which message fails?
[03:03] <hunger_suse> That works, udev start still fails.
[03:03] <Keybuk> hunger_suse: which message fails?
[03:04] <hunger_suse> Keybuk: "Starting kernel event manager... [fail] "
[03:04] <Keybuk> ok, that means it's probably already running then
[03:04] <Keybuk> "ifconfig -a" ... does that show an up interface?
[03:04] <Keybuk> or, in fact, any interface?
[03:04] <hunger_suse> Keybuk: eth0 is there but unconfigured.
[03:04] <Keybuk> hunger_suse: ok, ifup eth0
[03:06] <hunger_suse> Keybuk: ifup does not work (thanks to networkmanager), but I configured the interface manually.
[03:06] <Keybuk> ok, cool, can you copy that /etc/event.d/tty1 file somewhere for me to see
[03:06] <hunger_suse> Keybuk: Which files do you need?
[03:06] <Keybuk> hunger_suse: all of them would be perfect, if not, just tty1
[03:12] <bddebian> Howdy folks
[03:13] <giftnudel> hunger_suse: didn't forget the route and the dns server?
[03:14] <Kamion> Mithrandir: so at this point I'm having so much trouble getting gen_keymap to be useful that I'm close to giving up on sane-installer-keyboard
[03:14] <hunger_suse> giftnudel: I am on another computer.
[03:14] <hunger_suse> giftnudel: It is just the net being *sllooooowwww* again:-(
[03:14] <Kamion> Mithrandir: I've got one more fallback plan, which is to run gen_keymap with a small list of keymaps borrowed from console-data, rather than everything-minus-exclusions
[03:14] <giftnudel> hunger_suse: oh, ok, because these are the things I always foget, when I configure a network manually
[03:15] <hunger_suse> giftnudel: So did I... but then we have onle IP adresses here anyway:-)
[03:15] <hunger_suse> Keybuk: http://pastebin.ca/162350
[03:16] <hunger_suse> That is tty1.
[03:16] <hunger_suse> Keybuk: You want the rest as well?
[03:16] <Kamion> but that will take a little hacking
[03:16] <Keybuk> hunger_suse: no, that's fine
[03:16] <Keybuk> hunger_suse: if you haven't done already, do mount -o rw,remount /
[03:16] <Keybuk> hunger_suse: then dpkg --force-confnew --force-confmiss -i /var/cache/apt/archives/upstart_0.2.1-5_$ARCH.deb  (replace $ARCH)
[03:18] <hunger_suse> Keybuk: That fails due to /usr being ro, but that is normal;-)
[03:19] <hunger_suse> Keybuk: It installed sulogin.
[03:19] <Keybuk> hunger_suse: could you make /usr be rw please
[03:19] <hunger_suse> Keybuk: tty* scripts are unchanged.
[03:19] <Keybuk> you need that to be installed
[03:19] <hunger_suse> Keybuk: I did:-)
[03:19] <Keybuk> not half-corrupted
[03:19] <Keybuk> hmm
[03:19] <Keybuk> unchanged ?!
[03:20] <Keybuk> ok, brute force time
[03:20] <Keybuk> rm /etc/event.d/tty*
[03:20] <_ion> hunger: The files in your /etc/event.d are from upstart 0.1 i think.
[03:20] <Keybuk> then run that dpkg command again
[03:20] <hunger_suse> Keybuk: Yes.
[03:20] <Keybuk> _ion: right; it's somewhat wacky that they haven't been replaced by the ones from 0.2
[03:20] <hunger_suse> Reinstalling the sysv-compat does not change them either.
[03:21] <Keybuk> did rm'ing them and installing again work?
[03:21] <hunger_suse> Keybuk: Deleting the tty* and reinstalling them changed the files (some comments were added).
[03:22] <Keybuk> hunger_suse: more pointedly, was "start when" removed ?
[03:22] <Keybuk> should be just
[03:22] <hunger_suse> Keybuk: Yes, that is gone, too.
[03:22] <Keybuk> start on startup
[03:22] <Keybuk> stop on shutdown
[03:22] <Keybuk> ok
[03:22] <Keybuk> dpkg is being freaky
[03:22] <hunger_suse> spooky:-)
[03:22] <Keybuk> edit /etc/event.d/rcS again, put "console output" back before the script line
[03:22] <Keybuk> likewise for /etv/event.d/rc2
[03:22] <Keybuk> (but spelled right :p)
[03:23] <hunger_suse> Keybuk: I deleted everything in /etc/event.d and reinstalled the upstart and upstart compat debs.
[03:23] <Keybuk> hunger_suse: ok
[03:24] <Keybuk> you should have (sanity check) rc-default rc0-halt rc0-poweroff rc1 rc2 rc3 rc4 rc5 rc6 rcS sulogin tty1 tty2 tty3 tty4 tty5 tty6
[03:24] <Keybuk> ok
[03:24] <hunger_suse> Got all of those.
[03:24] <hunger_suse> Reboot?
[03:24] <Keybuk> you've added console output to rcS and rc2 ?
[03:24] <Keybuk> sync
[03:24] <Keybuk> then "reboot -f"
[03:25] <Kamion> Mithrandir: do you know if console-setup already has any notion of a mapping from old-style console keymap names to its names?
[03:26] <Kamion> maybe I have to borrow xserver-xorg.config to do the mapping
[03:26] <fschoep> dholbach: ping?
[03:26] <hunger_suse> Keybuk: I am back in the root shell.
[03:26] <Keybuk> hunger_suse: *blink*
[03:26] <Keybuk> ok
[03:27] <hunger_suse> Keybuk: Filesystem checks did not work out...
[03:27] <hunger_suse> Keybuk: They are encrypted, so that is no surprise.
[03:27] <Keybuk> right, this is just a normal failure mode now, yes?
[03:27] <Keybuk> the rc scripts were being run normally?
[03:27] <hunger_suse> Nope.
[03:27] <Keybuk> how un-normally were they being run?
[03:28] <hunger_suse> I got a login prompt and am asked for a password now.
[03:28] <Keybuk> ok
[03:28] <Keybuk> so it's booting
[03:28] <hunger_suse> Both programs accept input from the same place...
[03:28] <Keybuk> heh
[03:28] <hunger_suse> REALLY confusing...
[03:28] <Keybuk> alt+f2, login there, "sudo stop sulogin"
[03:29] <Keybuk> that's a known bug (sulogin is, in fact, just /bin/sh at the moment)
[03:29] <hunger_suse> Keybuk: Does not accept my root passwd:-(
[03:30] <Keybuk> hunger_suse: what doesn't?
[03:30] <Keybuk> that's between you and getty
[03:30] <fschoep> Kamion: do you have a minute to troubleshoot a packaging problem?
[03:30] <hunger_suse> Keybuk: Great.
[03:30] <bddebian> OMFG, why can't I win the frickin' lottery or something so I can just work on Ubuntu the rest of my life instead of this Godforsaken job... aaarrrggghh
[03:31] <Keybuk> it sounds like upstart is now running things the same way that sysvinit did
[03:31] <Keybuk> I don't know why your conffiles weren't updated, you may want to speak to iwj about that
[03:31] <hunger_suse> Keybuk: It does not run things the same way sysvinit did.
[03:31] <Keybuk> what is it not running the same way?
[03:32] <hunger_suse> Keybuk: It does NOT run cryptdisks before trying to mount partitions.
[03:32] <Keybuk> hunger_suse: it just runs /etc/init.d/rc
[03:32] <Keybuk> if it's mounting partitions at all, then that's working right
[03:32] <hunger_suse> Keybuk: It brings up all ttys and then tries to be clever, bringing up filesystems and I have not seen a single init script being run.
[03:32] <Keybuk> "tries to be clever" ?
[03:33] <Keybuk> we haven't put any of the clever in
[03:33] <hunger_suse> It definitly is not using the normal sysvinit boot sequence.
[03:33] <Keybuk> yes, it definitely is
[03:34] <Keybuk> could you edit /etc/event.d/tty* and change "start on startup" to "start on rc2"
[03:34] <Keybuk> that might be masking the cryptsetup prompt
[03:34] <hunger_suse> Keybuk: No I can not. I can not log into my system.
[03:35] <Keybuk> reboot, use init=/bin/bash then :p
[03:35] <Mithrandir> Kamion: I don't think it has a mapping, no.
[03:36] <Mithrandir> Kamion: I'd actually be more inclined to drop the cdebconf-keystep widget and go with console-setup, but I see the point why not to do it.
[03:40] <hunger_suse> Keybuk: It is not. It never starts cryptsetup
[03:41] <Keybuk> hunger_suse: tell you what, you fix it then :p  seeing as you're so sure of how it all works
[03:41] <Keybuk> personally, I know that I wrote no code to do any clever filesystem mounting
[03:41] <Keybuk> and I know that it just runs "/etc/init.d/rcS"
[03:41] <Keybuk> but if you think differently, go right ahead <g>
[03:43] <cbx33> mjg59: ping
[03:43] <hunger_suse> Keybuk: cryptdisks is not started. I do not know what you are doing, but this is not the normal boot sequence.
[03:43] <mjg59> cbx33: Hi
[03:44] <cbx33> what's the deal with usplash artwork
[03:44] <mjg59> In what way?
[03:44] <cbx33> heh
[03:44] <cbx33> colours
[03:44] <Spads> TEST PATTERN
[03:44] <Keybuk> hunger_suse: did you try my suggestion yet?
[03:44] <mjg59> You've got 256 of them now
[03:44] <hunger_suse> Keybuk: Which suggestion?
[03:44] <cbx33> is there a page about it
[03:44] <mjg59> Dennis has been working on various aspects of the theming
[03:44] <mjg59> Best bet is to talk to him
[03:45] <Keybuk> hunger_suse: modify the tty event.d files to say "start on rc2" instead of "start on startup"
[03:45] <hunger_suse> Keybuk: Sorry, I am concentrating on the other screen.
[03:45] <hunger_suse> Keybuk: Yes, I did that.
[03:45] <hunger_suse> Keybuk: Fixes the screen uglyness.
[03:45] <cbx33> mjg59: what is his nick
[03:45] <mjg59> seveas
[03:46] <cbx33> ahhh
[03:46] <Keybuk> hunger_suse: but you still don't get a cryptsetup prompt?
[03:48] <hunger_suse> Keybuk: That script is never run.
[03:48] <Keybuk> hunger_suse: can you ls /etc/rcS.d and make sure there's a symlink to it
[03:48] <hunger_suse> Keybuk: I added lots of echos right after the script starts. No output.
[03:49] <Keybuk> you see the normal boot output, with ok/fail?
[03:50] <hunger_suse> Keybuk: Yes, mostly.
[03:50] <Keybuk> "mostly" ?
[03:50] <hunger_suse> Keybuk: But that damn thing is missing
[03:50] <Keybuk> just that one script?
[03:51] <hunger_suse> Keybuk: That is the only one I noticed missing so far:-(
[03:51] <Keybuk> right
[03:51] <Keybuk> ls /etc/rc".d
[03:51] <Keybuk> uh
[03:51] <Keybuk> ls /etc/rcS.d
[03:51] <Keybuk> you see S26cryptsetup-early and S28cryptsetup ?
[03:51] <hunger_suse> Keybuk: It is listed there as S28cryptdisk.
[03:51] <Keybuk> "disk" ?
[03:52] <thom> Keybuk: yeah, they're called cryptdisk{,-early}
[03:52] <thom> disks, even
[03:52] <Keybuk> right
[03:52] <Keybuk> hunger_suse: confirm that it's "DISKS" please :p  not "disk"
[03:53] <hunger_suse> It is disks. The script works fine in sysvinit:-(
[03:53] <Keybuk> ok
[03:53] <cbx33> ping Seveas 
[03:54] <Keybuk> cat /etc/default/cryptdisks
[03:54] <hunger_suse> Keybuk: I'll need to do some work... I'll try again later.
[03:54] <Keybuk> make sure it's enabled
[03:54] <Keybuk> thom: you use this stuff, it's working for you?
[03:54] <hunger_suse> Keybuk: You are right, the upstart seems to do the right thing... it is /etc/init.d/rc S that does not what I want from it.
[03:55] <thom> Keybuk: i've not had chance to reboot and check recently; will try and do so soon
[03:56] <Keybuk> cryptdisks is a bit odd
[03:56] <Keybuk> I don't see how you can use it to encrypt your /usr ... given it uses openssl :p
[03:56] <Keybuk> which is in /usr/bin
[03:56] <hunger_suse> Keybuk: I have to admit that I am using a custom cryptdisks:-)
[03:56] <_ion> Please define "custom".
[03:56] <hunger_suse> Anyway, Have to get some work done. I'll be back later.
[03:57] <Keybuk> _ion: "broken" :p
[03:57] <_ion> :-)
[03:57] <Keybuk> thom: when you do, looking at this, you _may_ need to put "console owner" in /etc/event.d/rcS not "console output"
[03:57] <Keybuk> it looks like cryptdisks reads from /dev/tty, which won't be set
[03:58] <thom> ok
[03:58] <Keybuk> I'd be happy if someone not totally freakishly insane can prove this works <g>
[03:58] <Kamion> fschoep: sure
[03:58] <thom> *smirk*
[03:59] <Spads> zul: ping
[03:59] <zul> Spads: pong
[03:59] <Spads> zul: mind a pm?
[03:59] <thom> Keybuk: oh well, rebooting
[03:59] <zul> Spads: sure
[04:00] <Keybuk> _ion: I have a vague memory of him once complaining that he always says "N" to all conffile prompts and has it set to never update them
[04:00] <Keybuk> and got very broken by the hotplug->udev transition as a result
[04:00] <_ion> keybuk: ...
[04:01] <fschoep> Kamion: do you see my PM?
[04:01] <StevenK> Keybuk: Because asking any questions during an upgrade is just asking for trouble, surely.
[04:01] <Kamion> yes
[04:01] <fschoep> OK, just checking since I use a different IRC client.
[04:02] <_ion> dpkg could use a better UI for merging new changes to conffiles. An example of a good implementation: dispatch-conf from gentoo.
[04:04] <thom> Keybuk: console owner screws the pooch quite impressively
[04:05] <thom> Keybuk: it gets horribly confused between the input for cryptdisks and the input for console login
[04:05] <_ion> thom: You'll probably need to modify tty* to "start on rcS" instead of "startup".
[04:06] <thom> ah, right
[04:09] <thom> right, that works just peachily 
[04:10] <thom> and ion has fonts again, for a double win
[04:10] <ogra> iwj, http://pastebin.ca/162384 <- kills my firefox 
[04:11] <iwj> Hmm.  It's done something to mine too.
[04:12] <ogra> thats only a Xor.0.log ...
[04:12] <iwj> Stuck spinning the cpu.
[04:12] <ogra> *Xorg
[04:12] <iwj> Ah, here it is.
[04:12] <ogra> nothing spectacular big or so
[04:12] <iwj> It just took a minute or two to render.
[04:12] <_ion> thom: What fixed ion's fonts? They went broken, and i was too lazy/tired to trace the problem, so i just switched to metacity temporarily.
[04:12] <gnomefreak> iwj: you have a minute?
[04:12] <ogra> hmm, i left it open more than a minute before killing ff
[04:13] <iwj> Every line in that is a separate <li><div ... !
[04:13] <thom> _ion: wish i knew
[04:13] <thom> dist-upgraded this am, and that reboot just cured it
[04:14] <Keybuk> thom: did you change the tty thing as well?
[04:14] <iwj> And much of the time individual fragments are <span class=...
[04:14] <iwj> gnomefreak: Possibly.  What can I do for you ?
[04:14] <gnomefreak> iwj: ff b2 fixes i would say 80% or more of the crashing issues that people are having. i compiled b2 and havent had a problem yet. not sure if you updated it
[04:14] <iwj> No, I haven't.
[04:15] <thom> Keybuk: i changed tty1 to be start on rcS ; and rcS to be console owner
[04:15] <Keybuk> ok
[04:15] <iwj> I was going to wait with a new ff until after feature freeze, since there's stuff I want to get in first.
[04:15] <Keybuk> thom: and did that work better?
[04:15] <thom> Keybuk: worked perfectly
[04:15] <gnomefreak> and im not sure if you are maintainer of thunderbird too but is that gonna be 2.0 also?
[04:16] <iwj> ogra: So I conclude it's a stupid web page.  Unfortunately firefox isn't clever enough to replace it with a box saying `this web page is too stupid, please shoot the designer'.
[04:16] <iwj> gnomefreak: thunderbird> No, sorry.
[04:16] <gnomefreak> k
[04:17] <ogra> iwj, right ... looks like you have a way faster machine there ;) after 5 mins it shows up for me as well :)
[04:18] <Hobbsee> iwj: hehe, perhaps you need to patch it?
[04:18] <iwj> It's not a hugely new machine.  Athlon 2G.
[04:18] <iwj> Hobbsee: In my CFT.
[04:18] <thom> Keybuk: now to work out why ath0 is brought up but doesn't get configured
[04:18] <Hobbsee> iwj: CFT, sorry?
[04:18] <iwj> Copious Free Time.  :-).
[04:18] <Keybuk> thom: /etc/iftab, make sure you have "arp 1" ?
[04:19] <Hobbsee> iwj: ahhh...yes
[04:19] <thom> Keybuk: don't have an entry for ath0 at all
[04:19] <Keybuk> hunger: changing tty1 to "start on rcS" and rcS to "console owner" (not output) seems to fix it for thom
[04:19] <Keybuk> thom: ok, so it's not a rename bug
[04:19] <Keybuk> thom: don't suppose you have a bootchart there?
[04:20] <thom> Keybuk: remind me later and i'll generate you one
[04:20] <Keybuk> ok
[04:22] <doko> pitti, Kamion, ogra: I would need some testers for OOo on powerpc (just starting ...)
[04:22] <ogra> doko, already built ? 
[04:22] <ogra> my ppc is heavily outdated
[04:22] <doko> deb http://people.ubuntu.com/~doko/ubuntu/ edgy/$(ARCH)/
[04:22] <doko> deb http://people.ubuntu.com/~doko/ubuntu/ edgy/all/
[04:22] <ogra> Kamion, i dont have awk... before i make a mistake again, do i have sed in d-i ?
[04:22] <Kamion> ogra: yes
[04:23] <ogra> thanks
[04:23] <ogra> just wanted to be sure :)
[04:25] <hunger> Keybuk: I think I might know what is going wrong.
[04:25] <hunger> Keybuk: The next reboot will show.
[04:27] <dholbach> fschoep: pong
[04:29] <jdong> stop me before doing something stupid, but is it safe to rm -rf /usr/share/doc?
[04:29] <jdong> on a system extremely limited in free space :)
[04:30] <_ion> That depends on your definition of "safe". ;-)
[04:30] <jdong> the only thing I think I'll be losing is the ubuntu/kubuntu start pages
[04:30] <jdong> right?
[04:31] <jdong> it's 200MB of space I wouldn't mind having back
[04:35] <iwj> jdong: Delete only the files and not the directories.
[04:36] <iwj> Otherwise you'll get complaints from update-alternatives and the like.
[04:36] <jdong> iwj: k, will do
[04:36] <iwj> Also there are possibly other problems (eg, programs that automatically read bits of docs).
[04:36] <jdong> hmm
[04:36] <iwj> And perhaps the odd maintscript will fail, so good luck.
[04:36] <jdong> :-/
[04:36] <iwj> Worth a try though and do let us know how you get on :-).
[04:37] <jdong> at least there's a partial chance in data preservation.... :)
[04:37] <jdong> anyone know of any reliable fuse compressed filesystems?
[04:37] <jdong> there's so many to choose from
[04:38] <thom> i find reiserfs an excellent way to save a lot of disk space
[04:38] <thom> generally the whole partition after a few days
[04:38] <Mithrandir> mmm, eraserfs.
[04:38] <zul> w/in 19
[04:38] <zul> doh..
[04:38] <jdong> thom: lol
[04:39] <jdong> thom: it's not that bad.... unless you run reiserfsck... then it's bad :P
[04:39] <thom> jdong: no, it _is_ that bad
[04:39] <jdong> novell/suse has been defaulting to reiserfs for years now...
[04:39] <jdong> and personally, I've only damaged one reiserfs partition... by shrinking it
[04:40] <mjg59> The fact that reiser3 is unmaintained is a damn good reason not to default to it
[04:41] <mjg59> Suse can get away with it to some extent because they have developers who understand it
[04:41] <iwj> Can any other filesystem yet do both shrink (at all) and online grow ?
[04:42] <jdong> no, that's ext2/3 and reiserfs3
[04:42] <iwj> ext3 can do online grow ?
[04:42] <Mithrandir> there are patches for it, at least, yes.
[04:42] <jdong> yeah, at least in fedora/rhel4 it can
[04:42] <mjg59> With patches, I believe
[04:42] <iwj> With patches.  Riiiight.
[04:43] <jdong> does ubuntu not have the online grow patches for ext3?
[04:43] <jdong> I could've sworn I've lvm-grown ext3 online before
[04:43] <jdong> maybe that was in fedora
[04:43] <jdong> nvm
[04:43] <iwj> There have been patches to do this for ages.  I don't know any of the details but I have to ask: why aren't they in mainline ?
[04:44] <maswan> iwj: xfs possibly jfs (neither does online shrink though)
[04:44] <_ion> AFAIK any new features will not be added to ext3, instead they're developing ext3dev (to be ext4).
[04:44] <ajmitch> iwj: iirc some at least went into mainline in 2.6.10
[04:44] <maswan> at least I think that was the case last time I looked at it, I've never had reason to shrink a filesystem myself
[04:45] <jdong> I'm not impressed with ext3 or reiserfs's shrinking capabilities
[04:45] <jdong> I've trashed both before by shrinking
[04:45] <jdong> grow's safe enough
[04:45] <jdong> but shrink.... no
[04:46] <lamont> pitti: ping
[04:53] <pitti> lamont: pong, but busy; will read scrollback
[04:54] <lamont> pitti: tossed it in the other window for you to read later
[05:13] <janimo> heno: is gok staying the keyboard tool for edgy?
[05:14] <heno> janimo: no, we are moving to onBoard: https://wiki.ubuntu.com/Accessibility/Projects/onBoard
[05:15] <cbx33> LaserJock: !!!!!!!!!!!!!!!!!!!!!!
[05:15] <janimo> heno: ok, I asked since I did not see the latter in the archives and in the scrollback someone  mentioned gok
[05:15] <jsgotangco> Mr. Mantha!
[05:16] <LaserJock> heh
[05:16] <heno> janimo: it's just uploaded to universe, still in NEW I guess
[05:17] <Kamion> that can be rectified
[05:17] <Kamion> heno: er, it's not in NEW
[05:18] <Kamion> http://librarian.launchpad.net/4126254/3wpP6vZEOR10PTJ96thymFJtrOO.txt <-- weird
[05:18] <heno> dholbach: any idea where onboard has wandered off to?
[05:18] <Kamion> you need to expect a new/accept message following an upload
[05:18] <Kamion> and complain if you don't get one
[05:19] <Kamion> heno: it's (probably) not dholbach's problem - I'm investigating on the archive side
[05:19] <dholbach> heno: chris jones should have got a mail, not me
[05:19] <Kamion> no, the signer gets a mail too
[05:19] <dholbach> oh, hum - maybe I deleted it
[05:19] <janimo> heno: so for the xubuntu-at-sok metapckage I should depend on onboard at-spi and gnome-orca?
[05:19] <dholbach> it should be the virtkey and onboard source package
[05:20] <heno> tortoise_: did you get such an email? ^
[05:20] <Kamion> stop flailing :)
[05:20] <Kamion> soyuz broke, see the URL I posted above
[05:20] <Kamion> I'm just going to try reprocessing the upload and see if it fails a second time
[05:20] <Kamion> I love Soyuz
[05:21] <heno> ah, that reading of 'wierd', ok :)
[05:21] <Kamion> give it four minutes for the cron job to fire
[05:21] <tortoise_> heno dholbach: nope
[05:22] <heno> janimo: actually if by 'sok' package you mean for the onscrean keyboard, you only need python-virtkey and onboard
[05:22] <heno> janimo: it doesn't need at-spi 
[05:22] <G0SUB> pitti: I have fixed a lot of bugs today ... just trivial ones remaining
[05:22] <heno> janimo: the Orca stuff on the other hand does
[05:22] <Kamion> virtkey failed too
[05:22] <Kamion> what's going on
[05:23] <heno> janimo: btw, did I send you the link to this post of a guy using Xubuntu as a magnifier system? http://www.ubuntuforums.org/showthread.php?t=236401
[05:23] <Kamion> well, I've dumped it back into incoming too - we'll see
[05:24] <heno> It seems to work quite well for him
[05:26] <janimo> heno, you did not, I am looking at the link now
[05:26] <Kamion> there, that landed in NEW now
[05:27] <janimo> heno, by sok package I mean one of the three packages (mag, speech, sok) mentyioned in the xubuntu a11y spec you drafted in paris. The other two ar e uploaded already
[05:29] <heno> janimo: right, in that case my explanation above is correct. onboard does not need at-spi (and has changed name from sok). It needs gconf from gnome though (tortoise_ will know the details of where and why)
[05:32] <Kamion> virtkey and onboard source both accepted
[05:33] <dholbach> nice - thanks Kamion
[05:36] <jdong_> where are my fancy charge/discharge graphs in gnome-power-manager
[05:36] <jdong_> one of my laptops shows them, the other doesn't
[05:36] <heno> thanks Kamion (/me adds them to the main inclusion queue)
[05:38] <Kamion> woo, gen_keymap ran successfully
[05:40] <pitti> G0SUB: niced
[05:40] <thom> Kamion: after how long? :)
[05:42] <Kamion> thom: once I brutally reduced the list of keymaps it was working on, not that long
[05:42] <Kamion> it's taken me all day though
[05:52] <Kamion> zul: around? (xen-edgy)
[05:52] <carlos> seb128: hi, around?
[05:53] <doko> ogra: are you currently upgrading for the OOo test on powerpc?
[05:53] <zul> Kamion: yep
[05:53] <Kamion> zul: I was wondering if you could elaborate on the "still blocked on the builds" comment in the spec's status whiteboard
[05:54] <zul> Kamion: its not blocked anymore i think it was when it was sitting in new
[05:54] <Kamion> zul: ok, could you update it then? is the spec basically Implemented, or is there more to do?
[05:58] <janimo> heno, in that case does it make more sense to name the package xubuntu-at-keyboard?(onboard?)Or do you have a better suggestion?
[05:59] <heno> janimo: xubuntu-at-keyboard sounds good
[06:00] <zul> Kamion: its basically implemented i have to add some 3rd party drivers but thats what Im working on now
[06:00] <zul> now/tonight
[06:02] <Kamion> zul: ok, thanks
[06:02] <_ion> Btw, what's the point of mounting /lib/modules/$(uname -r)/volatile as tmpfs and putting 18 megs of stuff in there?
[06:03] <_ion> s/point/purpose/
[06:03] <mjr> don't really know tho
[06:04] <jdong_> _ion: to make you buy more RAM?
[06:04] <mjr> to make you create a swap partition? :] 
[06:04] <zul> Kamion: done
[06:05] <Kamion> zul: great, thanks
[06:05] <bluefoxicy> damnit
[06:05] <_ion> Sure, it can be swapped away, but that's still 18 MiB worth of RAM or swap space.
[06:05] <bluefoxicy> OOo is broken.  Horribly.
[06:06] <_ion> bluefoxicy: What else is new? ;-)
[06:06] <bluefoxicy> _ion:  I mean more than usual.  Export to PDF; Open; and Save As cause crashes
[06:08] <hunger> Keybuk: Still around?
[06:11] <tortoise_> janimo: At the moment onboard needs gconf and the python bindings pull in most of gnome so i'm not sure if onboard is suitable for xubuntu at this point.  I'm hoping to change that though.
[06:13] <bluefoxicy> aaargh.
[06:13] <bluefoxicy> How do I convert .odt to Word XP with broken OOo
[06:13] <seb128> carlos: pong
[06:14] <carlos> seb128: hi, do you know why the network monitor applet takes a huge space of my panel when monitoring a wireless device?
[06:15] <kozz> Kamion: The DVD works on Pegasos now, nice :)
[06:15] <seb128> carlos: you want to speak to dholbach, iz artwork bog
[06:15] <dholbach> seb128: is not
[06:15] <dholbach> seb128: the artwork didn't change
[06:15] <dholbach> seb128: at least that part of it
[06:16] <dholbach> i meant to look at it, but didn't do it yet
[06:16] <seb128> dholbach: but it probably doesn't provide an icon to the required dimension
[06:16] <seb128> you can then argue if the panel should manage
[06:16] <seb128> or if we should ship the icon for it :p
[06:16] <carlos> so it's a known bug
[06:16] <dholbach> the code in netstatus applet changed - I'll look it up
[06:16] <carlos> right?
[06:16] <dholbach> there's a bug filed already, yes
[06:16] <carlos> ok
[06:16] <seb128> carlos: yep
[06:16] <dholbach> thanks carlos
[06:16] <carlos> dholbach: do you know the bug number?
[06:17] <janimo> tortoise_: are you planning the change before edgy?
[06:18] <janimo> which parts of gnome besides gconf are you using?
[06:18] <tortoise_> I'm hoping to start it soon but I doubt it will be.  I only use gconf at the moment.
[06:18] <janimo> I need to fix the pygconf pulls in whole gnome issue because it affects an increasing number of apps
[06:19] <janimo> tortoise_: fixing that is easy in the python-gnome binding package, it only needs the agreement of the debian maintainers of that
[06:19] <seb128> carlos: https://launchpad.net/distros/ubuntu/+source/gnome-netstatus/+bug/57626
[06:19] <Ubugtu> Malone bug 57626 in gnome-netstatus "Network Monitor Icons look stretched out on 1280x768 resolution" [Untriaged,Confirmed]  
[06:19] <carlos> seb128: thanks
[06:19] <seb128> np
[06:19] <Sp4rKy> hi
[06:20] <Sp4rKy> please does anyone using edgy could paste me the result of apt-cache policy libxss1
[06:20] <tortoise_> janimo:  True but I need to do it for Kubuntu too.  Gconf itself pulls in orbit etc.
[06:20] <janimo> tortoise_: does your app use gcoinf keys which are visible/used by other apps in the desktop or just as a convenient way to store some private settings?
[06:21] <janimo> if you need it to read system wide a11y settings I am not sure there's a better way
[06:21] <janimo> if it
[06:21] <janimo> is the latter though gconf can be avoided
[06:21] <tortoise_> janimo: no I dont.  Just for storing private settings
[06:21] <tortoise_> yeah
[06:22] <janimo> tortoise_: then maybe ConfigParser or what it is called could be good enough, uses INI style config files
[06:22] <tortoise_> Currently the settings and the keyboard are handled in separate apps so I would need to notify onboard of changes made by the settings app
[06:23] <janimo> oh, so you use gconf notification api
[06:23] <tortoise_> janimo: yep, it's very convenient.  Someone really needs to write a desktop neutral gconf that uses dbus.
[06:24] <janimo> tortoise_: there is such a gconf but was not in kainline I think nokia uses it
[06:24] <janimo> s/kainline/mainline/
[06:25] <jdub> tortoise_: gconf will eventually hit gtk+/glib anyway, most likely the d-bus version
[06:25] <jdub> tortoise_: so all this effort going in to avoiding it is pretty wasteful
[06:25] <janimo> tortoise_: anyway if it's convenient use it, I need to fix the gnome-python bindings anyway, can't expect all pygconf users to find other APIs
[06:25] <jdub> tortoise_: particularly when you benefit from notification, etc.
[06:25] <tortoise_> jdub: In the long time, yes
[06:25] <tortoise_> */time/term
[06:26] <seb128> Kamion, Mithrandir, mvo: what is the recommended way to change the default language on a ubuntu installe, changing /etc/environment by hand?
[06:26] <jdub> might not be as long term as you think :-)
[06:26] <janimo> jdub: you think it could be 2.12?
[06:26] <jdub> probably not 2.12
[06:27] <janimo> then that is long-term :). two gtk release cycles at least, and those are open source years ;)
[06:28] <tortoise_> I was hoping to make onboard toolkit neutral too.  I do all the rendering in cairo so I was hoping to have the event handling pluggable qt/gtk.
[06:28] <janimo> tortoise_: kde guys will probably not want cairo but you'd have to use qt/arthur
[06:28] <tortoise_> doesn't kpdf use cairo?
[06:28] <janimo> I don;t know
[06:28] <Kamion> kozz: woo, glad to hear it
[06:29] <janimo> kpdf uses poppler, so whatever poppler uses
[06:29] <Kamion> seb128: language-selector, I thought; it ought to change /etc/environment and /etc/default/locale (if not, it's a bug)
[06:29] <janimo> I think it has both cairo and arthur (at least experimental) in it
[06:29] <seb128> Kamion: ok, thank you
[06:30] <Riddell> tortoise_: no
[06:30] <tortoise_> Riddell: Ok, scratch that then
[06:31] <tortoise_> does Kubuntu ship gtk?
[06:31] <jdub> tortoise_: you are entering a world of pain... :-)
[06:31] <tortoise_> lol
[06:31] <janimo> tortoise_: it doesn't, it has the 'k' in the wrong position
[06:32] <jdong_> LOL
[06:32] <Riddell> tortoise_: no
[06:35] <Riddell> trappist: .ini style config files should be easy enough to read from both gnome and kde apps
[06:43] <trappist> Riddell: that makes sense - this relates to a bug I commented on a long time ago?  I have a very vague recollection.
[06:44] <mdz> morning
[06:45] <_ion> Evening.
[06:45] <zul> afternoon mdz
[06:45] <mvo> hello mdz
[06:46] <pitti> hey mdz
[06:46] <pygi> ho pitti :)
[06:46] <mdz> the inbox hits!  the inbox hits!  you feel hemmed in.
[06:46] <pygi> pitti, we've got first released app using libburn, whee :)
[06:47] <pitti> pygi: \y/
[06:47] <pygi> Brasero (aka old Bonfire)
[06:47] <_ion> You die. Do you want your possessions identified?
[06:47] <pygi> _ion, ahm? :)
[06:49] <pygi> pitti, we need libburn package soon so we could make brasero package libburn-enabled
[06:50] <pygi> and libisofs while we're at that :)
[06:55] <Kamion> mdz: evening
[06:57] <tseng> breezy had gtk+ with cairo right?
[06:58] <jdub> tseng: yeah
[06:59] <jdub> tseng: 2.8.6
[06:59] <keescook> pitti: you seen this? http://www.suse.de/~krahmer/no-nx.pdf remote code execution even with nx bit.  :P
[07:00] <tseng> jdub: cool
[07:00] <pitti> keescook: ouch
[07:00] <tseng> jdub: i just updated to dapper and xming windows clients are running slower
[07:00] <pitti> keescook: hello, nice to see you again
[07:00] <tseng> jdub: trying to think of why
[07:00] <keescook> hiya!  :)
[07:01] <janimo> pitti: what is till's nick?
[07:01] <pitti> janimo: till
[07:01] <pitti> janimo: he's not online ATM
[07:01] <janimo> ok
[07:16] <Kamion> Mithrandir: http://people.ubuntu.com/~cjwatson/tmp/console-setup-keymapper.diff
[07:17] <Kamion> Mithrandir: first cut at gluing keymapper/cdebconf-keystep into console-setup; installer integration totally untested; review welcome
[07:17] <Kamion> (this is for sane-installer-keyboard, for those playing along at home)
[07:18] <Kamion> you'll need to wait for buildds to get around to keymapper 0.5.3-5 in order to build it
[07:19] <HiddenWolf> Kamion: will that improve user experience or just reduce maintainer headaches? 
[07:22] <Kamion> HiddenWolf: hopefully a bit of both
[07:22] <Mithrandir> Kamion: it looks good to me, but I've read it, not tested it.
[07:22] <Kamion> HiddenWolf: no more skew between console and X keymaps
[07:24] <Kamion> there will likely be a certain amount of teething trouble of course - it's a big change
[07:24] <Kamion> and it's still possible it won't make edgy - we're kind of close to the wire
[07:26] <heno> Kamion: heh, looks like I need to do some fresh Live CD a11y testing :)  I can't remember how that works now
[07:26] <heno> I swear I've tested ubiquity with at-poke in the past and that most stuff shows up though
[07:27] <heno> But my memory is not always to be trusted ...
[07:27] <Kamion> heno: I can't figure out how, that's all :)
[07:28] <Kamion> I thought this was why you'd made livecd-access depend on sudo-admin-atspi
[07:28] <janimo> heno, thinking more about it, I think the two keyboard apps could be installed by default in xubuntu w/o any metapackage for them
[07:28] <janimo> as soon as I fix their python-gnome dep
[07:29] <heno> Kamion: yep, and people have reported problems using it. (anyway I'll check later to make sure)
[07:30] <heno> The good news is that people are happily using the Knot 2 CDs to test Ubuntu as a platform with screen reading, which is great in itself
[07:30] <heno> janimo: that would be very cool!
[07:32] <heno> janimo: I remember tortoise_ was looking into that as well
[07:32] <heno> tortoise_: which part was it that pulled in most of gnome?
[07:32] <Kamion> Mithrandir: any chance you could fix the way that xkeyboard-config is listed as a transitional package but yet contains all the data that should be in xkb-data? I think that's what's breaking the console-setup build.
[07:33] <Kamion> bug 59220
[07:33] <Ubugtu> Malone bug 59220 in xkeyboard-config "Falsely claims it can be safely removed" [High,Confirmed]  http://launchpad.net/bugs/59220
[07:33] <Mithrandir> Kamion: oh, true dat, I need to fix it.
[07:34] <janimo> heno: I talked to tortoise about it an hour ago
[07:35] <janimo> it's only gconf but the rest of gnome libs come in via the python-gnome bindins
[07:35] <janimo> I know how to fix this
[07:35] <heno> janimo: ah ok, all in good hands then :)  Thanks for looking at this
[07:35] <heno> Cool, that will help us for other DEs too
[07:36] <heno> janimo: I guess removing gnome deps from gtk apps is one of your specialities :)
[07:36] <Kamion> Mithrandir: should we be putting console-setup-fonts-udeb in the d-i initrd? I notice that console-setup-udeb doesn't depend on it
[07:37] <janimo> heno, yeah :). The hard poart is talking to gnome people about it not the technical bits ;)
[07:37] <Kamion> it doesn't seem necessarily required, since we have unifont for the installer UI
[07:40] <Mithrandir> Kamion: unsure, I think we might want it?
[07:42] <dieman> ben collins rocks for getting custom kernel builds so easy now
[07:43] <jdong> dieman: how easy is it?
[07:46] <janimo> dieman: what changed besides dropping the various architecture flavours?
[07:48] <dieman> the KernelCustomBuild wiki page has deets
[07:48] <dieman> took me about 15 minutes to pull down his git archive, roll in a patch, and start compiling
[07:53] <zyga> hey
[07:56] <janimo> dieman: thanks for the link, indeed looks easier than it used to be
[07:56] <jdong> yeah, looks promising for those blessed with the task of modifying edgy's kernel :)
[07:56] <dieman> yeah, i need the web100 patch for a set of machines
[07:57] <dieman> and netem, but netem was already there
[07:57] <jdong> (probably not)
[08:03] <G0SUB> pitti: I got two people to test the app and they found no bugs
[08:03] <pitti> cool
[08:06] <G0SUB> pitti: in fact they are still testing it as we speak
[08:07] <Seveas> mjg59, (irl got in the way today -- final theming patches, including example theme will arrive in a few hours)
[08:07] <mjg59> Seveas: No problem
[08:08] <mjg59> Seveas: I uploaded a package with a 256 colour image, just to see if anyone would notice
[08:15] <janimo> mvo: ping
[08:16] <mvo> hello janimo
[08:16] <janimo> mvo: hello
[08:16] <janimo> at one point update-manager was using gamin directly?
[08:17] <mvo> janimo: update-notifier? sort of, it used the fam interface. but I switched to gnome-vfs because gamin was not very reliable
[08:17] <mvo> gnome-vfs uses inotify directly and I haven't had trouble since
[08:17] <janimo> hmm ok. wasn't sure whether gvfs used gamin or inorify. thanks
[08:18] <mvo> janimo: gnome-vfs is not ok for xubuntu?
[08:18] <janimo> mvo, we tried avoiding it so far
[08:18] <janimo> even though now it is somewhat leaner w/o bonobo
[08:19] <mvo> janimo: I use basicly only gnome_vfs_monitor_add()
[08:19] <janimo> mvo, right I rememebr checking out the code and wondering what was going wrong with gamin
[08:19] <mvo> I had all sort of problem of it not detecting changes
[08:19] <janimo> as thunar and xfce desktop use gamin directly we are porbbaly exposed already to whatever nastiness it holds :)
[08:21] <janimo> I wonder how evil it would be to use inotify directlty. That would mean linux only though
[08:22] <mdz> slomo: I saw a bunch of scrollkeeper errors from the current tomboy; do you know what the cause is?
[08:22] <mvo> janimo: I looked into this for update-notifier, its not terrible bad, but a bit cumbersome
[08:22] <mvo> janimo: while I have you here, do you know why "xubuntu-desktop" is in section: misc?
[08:24] <mvo> kamion, mdz: what do you think about adding a "metapackage" section for the various "ubuntu-base,minimal," etc? this way we could enable --install-recommends by default for this section only and dodge the big "cleanup-recommends" work for now
[08:25] <janimo> mvo: hmm I think there is no section for xfec like ther is for gnome and kde IIRC
[08:27] <mdz> mvo: I like that idea
[08:28] <zyga> did anyonce notice udev libsane problem at boot?
[08:28] <mvo> mdz: the apt patch is ready, now I need to look into germinate
[08:28] <jdong> zyga: yes, there is an error about it on bootup
[08:30] <crimsun> jdong: it's a comment anyway, so it doesn't matter ultimately.
[08:30] <ogra> grmbl ... db_subst doesnt like me :/
[08:30] <crimsun> there's a bug reported on it
[08:31] <jdong> crimsun: ok, that makes me feel better :)
[08:35] <slomo> mdz: will  look at it later... but afaik they were there from the first version with scrollkeeper docs,i.e. 0.3.9 or something
[08:36] <HiddenWolf> mjg59: and, did they notice? :)
[08:37] <G0SUB> pitti: it'd be great if you do a review again tonight
[08:37] <bluefoxicy> all that just to get UID 1000 on the LiveCD
[08:37] <pitti> G0SUB: I'll try to manage
[08:37] <G0SUB> pitti: if you tell me I will be there when you are free
[08:39] <G0SUB> pitti: any way I am going to mail you in a while mentioning the changes made today
[08:39] <mdz> slomo: I guess they're only visible because of the way that tomboy is running scrollkeeper in postinst
[08:51] <janimo> heno: which are the main differences from an a11y user pov between a gnome and a gtk only app? For instance do gaim and abiword lack something because they do not use gome libs?
[08:52] <jdong_> janimo: less ram usage? :)
[08:52] <jdong_> maybe less integration with gnome
[08:53] <janimo> jdong_: I am intersted in the psecifics, especially if there is some drawback with gtk only
[08:53] <heno> janimo: I don't think there is much in it. Certainly gaim is known to work well with Orca
[08:54] <heno> I have no idea if it relies on using any Gnome libs in doing that. I would think not
[08:54] <janimo> heno, because one of the reasons some maintainers seem to stick to gnome deps is that gnome_program_init sets up some extra a11y stuff
[08:55] <janimo> but I think I'll check out the source as this may not be as serious as it is said to be
[08:55] <janimo> there is a libgail-gnome module which is only used in gnome apps
[08:56] <janimo> and I think bonobo using apps rely on this initializations too. but most gnome apps are thankfully no longer using bonobo
[08:56] <ogra> Kamion, http://people.ubuntu.com/~ogra/netscript.templates and http://people.ubuntu.com/~ogra/netscript, any hint why the interface selection doesnt work ?
[08:57] <Surak> Hi guys
[08:57] <janimo> heno, it seems that if apps use widgets defined in gnome libs as opposed to proper gtk ones they need that gnome_program_init otherwise not
[08:58] <janimo> I am gathering ammo for fighting gnome_program_init you know ;)
[08:58] <heno> janimo: I think bonobo is somehow part of the a11y chain though
[08:58] <heno> heh, right
[08:58] <heno> I just tested Abiword with Orca and it seems to work well
[08:59] <janimo> heno: yes,  it is used by at-spi
[08:59] <janimo> bonob that is
[08:59] <Surak> I've noticed that sl-modem-daemon is not being shipped for amd64 both in ubuntu and debian. Anyone knows why? I've seen people at linmodems patching it for amd64...
[08:59] <heno> janimo: so how do you get around that?
[09:00] <janimo> the idea is if a11y can take advantage of these gnome bits _runtime_ there is no need for hard depending on them
[09:00] <heno> I see
[09:00] <janimo> heno: gtk_init() does all of program initialization which is needed by a gtk app
[09:00] <janimo> most gnome apps still; use gnome_app_init though, because of historical reasons I guess
[09:01] <janimo> it has the convenience of seting up bug buddy in case of crashes
[09:01] <janimo> soem of the a11y stuff for apps that use custom gnome widgets and maybe some more
[09:02] <janimo> instead of relying on bug buddy I'd rather keep the code leaner and hence less buggier :)
[09:02] <heno> wise
[09:03] <heno> From a a11y perspective, simple non-custom widgets are preferred to ensure that the AT apps know how to read them
[09:04] <janimo> exactly, that is what most apps use anyway, so no extra a11y setup is needed besides having the environment set up
[09:11] <Kamion> ogra: debconf choices need to be separated by comma-space, not just space
[09:11] <Kamion> ogra: so:
[09:12] <Kamion> db_subst ltsp-client-builder/dhcp-interface choices "$(echo "$IFACE_LIST" | sed 's/ /, /g')"
[09:12] <ogra> yep 
[09:12] <ogra> thats somethig i figured since i asked, but the choice still isnt presented
[09:12] <Kamion> if that still doesn't work, I suggest capturing a trace with DEBCONF_DEBUG=5 set as a boot parameter
[09:13] <jdong> any idea what would cause a usb mass storage device to give me " reset high speed USB device using ehci_hcd and address" errors?
[09:13] <jdong> it happens only in linux on this laptop
[09:13] <jdong> not in windows
[09:14] <jdong> does not happen in linux on my desktop
[09:14] <mdke> who has access to changing the html on cdimage.u.c?
[09:14] <Kamion> ogra: 'set -x' wouldn't hurt either - you'd get most of the information from that, and you don't need to reboot to do that
[09:14] <Kamion> mdke: me
[09:14] <mdke> Kamion: can you take bug 59161 pls?
[09:14] <Ubugtu> Malone bug 59161 in ubuntu-website "knot 2 page empty" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59161
[09:15] <ogra> ooooh
[09:15] <ogra> Kamion, thanks a lot ... ! works :)
[09:15] <Kamion> mdke: send it over to the ubuntu-cdimage product
[09:16] <Kamion> ogra: what was wrong?
[09:16] <mdke> Kamion: ok. Ditto for bug 58800?
[09:16] <Ubugtu> Malone bug 58800 in ubuntu-website "Release page should link to other derivatives" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/58800
[09:16] <Kamion> mdke: yeah
[09:16] <mdke> okies
[09:16] <mdke> thanks
[09:16] <Kamion> that one's harder to fix quickly, it needs to be scripted in order to stay fixed
[09:16] <ogra> i had the interfaces hardcoded with comma at the top *and* substituted the commas at the same time ...
[09:17] <ogra> hmm, now the output is wrong ... but i'm getting there :)
[09:17] <Kamion> ogra: ah, right
[10:15] <jdong> hmm, after the usplash upgrade, my screen turns off during the Loading Hardware Drivers stage
[10:15] <jdong> never to turn on again
[10:16] <jdong> turning off splash makes everything work again
[10:26] <Keybuk> Mithrandir: ping?
[10:26] <Mithrandir> Keybuk: hi
[10:26] <Keybuk> Mithrandir: odd mailman problem, you might know the answer
[10:26] <Keybuk> it's base64-encoding someone's e-mails in the archive
[10:26] <Mithrandir> you're using pipermail?
[10:26] <Keybuk> http://lists.netsplit.com/pipermail/upstart-devel/2006-September/000046.html
[10:26] <Keybuk> e.g.
[10:26] <Keybuk> yes
[10:27] <Keybuk> dapper packages
[10:27] <Mithrandir> oh, special
[10:27] <Mithrandir> what does it look like in the mbox?
[10:27] <Keybuk> where do I look?
[10:27] <_ion> The common things in the emails: they contain a character not representable by ISO-8859-1, and they are signed with PGP.
[10:28] <Mithrandir> /var/lib/mailman/archives/private/upstart-devel.mbox, iir
[10:28] <Mithrandir> +c
[10:29] <Keybuk> they look normal in the mbox
[10:29] <Keybuk> Content-Type: text/plain; charset=utf-8
[10:29] <Keybuk> Content-Disposition: inline
[10:29] <Keybuk> Content-Transfer-Encoding: quoted-printable
[10:31] <Amaranth> It's not just that guy's emails
[10:32] <Mithrandir> Keybuk: hmm, weird.  Looking
[10:35] <Keybuk> I could try updating it to the edgy packages, and see if that works better?
[10:36] <Mithrandir> they're the same, iirc
[10:36] <Keybuk> 2.1.5 vs 2.1.8
[10:36] <rodarvus> Keybuk, I tried to subscribe to upstart-devel a few days ago (using my @ubuntu.com mail address), but got no response from mailman
[10:36] <Keybuk> rodarvus: try again
[10:36] <rodarvus> *nods*, will do
[10:37] <Keybuk> you're not in the list
[10:37] <rodarvus> yes, I haven't even received the confirmation message from mailman
[10:37] <Keybuk> rodarvus: how did you subscribe?
[10:38] <rodarvus> via the web interface
[10:38] <rodarvus> on lists.ubuntu.com
[10:38] <rodarvus> I just retried so again, just for the sake of it
[10:38] <_ion> Do you mean lists.netsplit.com?
[10:38] <rodarvus> yeah.
[10:38] <rodarvus> thinko
[10:39] <Keybuk> Sep 06 21:37:07 2006 (28772) upstart-devel: pending Rodrigo Novo <rodavus@ubuntu.com>  200.146.22.15
[10:39] <Keybuk> Sep 06 21:37:27 2006 (288
[10:39] <rodarvus> :D
[10:40] <rodarvus> haha
[10:40] <Keybuk> you made the same mistake on Friday too
[10:40] <Mithrandir> Keybuk: sorry, I can't find any mention of it doing base64 encoding in the archiver at all
[10:40] <Keybuk> Mithrandir: kooooky
[10:40] <Keybuk> Mithrandir: md5 perhaps?
[10:41] <rodarvus> Keybuk, I used autocompletion to fill the form today
[10:41] <Keybuk> no, it's definitely base64 ... has the silly ==s
[10:41] <Keybuk> not rot13, doesn't look like Welsh
[10:41] <_ion> keybuk: It is base64. When decoded, it outputs the message plus some pipermail content.
[10:41] <rodarvus> Keybuk, thanks
[10:42] <Mithrandir> Keybuk: it decodes to:
[10:42] <Mithrandir> Hi
[10:42] <Mithrandir> Where (or by whom) is the process of migrating stuff from Ubuntu edgy
[10:42] <Mithrandir> rcS to upstart jobs planned or coordinated?  I'd like to help where i
[10:42] <Mithrandir> can, but i don't want to step on anyone's shoes.
[10:42] <Keybuk> _ion: yeah, so it does
[10:42] <Mithrandir> so it's certainly decodable and base64.
[10:42] <mjg59> elmo: Are you guys doing Linuxworldexpolala this year?
[10:42] <Mithrandir> Keybuk: can you put the list archive mbox somewhere I can get at it?
[10:42] <Keybuk> Mithrandir: I can give you temporary root on the box, if you want :p
[10:43] <elmo> mjg59: in the UK?
[10:43] <Mithrandir> Keybuk: I'd rather just have the mbox, if I can't get anything useful out of that (and maybe getting the message bounced to me), I can't really get anything useful out of it.
[10:43] <mjg59> elmo: Yeah
[10:43] <Keybuk> Mithrandir: ok
[10:44] <Keybuk> http://paperboy.netsplit.com/upstart-devel.mbox
[10:45] <elmo> mjg59: I'm not sure, sorry - I can ask tomorrow
[10:45] <mjg59> elmo: Ok, no problem
[10:48] <slomo> mdz: right... but this should only happen when updating from 0.3.9+dfsg-0ubuntu1 or 0.3.9+dfsg-0ubuntu2. and iirc most of those errors were in scrollkeeper stuff not from tomboy. what shall i do about it? normally nobody would ever see this unless he had one of the two broken versions installed
[10:48] <mdz> slomo: no worries
[10:49] <ge_ubuntu> is upstart already in Dapper updates?
[10:49] <Burgwork> ge_ubuntu, no
[10:49] <ge_ubuntu> it will be first in Edgy?
[10:49] <ogra_> Kamion, http://people.ubuntu.com/~ogra/netscript and http://people.ubuntu.com/~ogra/netscript.templates ... i think i'm done and will add it to the postinst if you dont find any blockers
[10:49] <Burgwork> ge_ubuntu, yes
[10:49] <ge_ubuntu> may I be little off topic?
[10:50] <Burgwork> not really
[10:50] <ge_ubuntu> this is about ubuntu still :)
[10:50] <Burgwork> ok
[10:50] <ge_ubuntu> I just don't know where to post, where Ubuntu team will hear
[10:51] <ge_ubuntu> What will be your filling, if
[10:51] <ge_ubuntu> Ubuntu will be chosen to be the OS in all the schools in the whole country?
[10:52] <ge_ubuntu> the country is Georgia, a small country though. about 4 mln
[10:53] <ge_ubuntu> with about 3000 schools
[11:00] <Mithrandir> Keybuk: it doesn't really make any kind of sense.. :-/
[11:01] <Keybuk> Mithrandir: to me neither :p
[11:01] <Keybuk> does it do the same to you then?
[11:02] <Mithrandir> Keybuk: I'm just trying to understand the code paths; it's been a while since I poked at mailman
[11:04] <Kamion> ogra: yes, looks mostly ok. you should change Description to _Description in the .templates file; assuming you're just going to add that to ltsp-client-builder.templates (are you?), then run debconf-updatepo after you've done so to update the .pot and .po files
[11:05] <ogra> yup
[11:05] <ogra> its just for running it locally
[11:05] <Kamion> ogra: "please select the interface you want to to use for the Thin Client Network." -> "Please select the interface you want to use for the thin client network", IMHO
[11:05] <ogra> if i run it in terminal with the underscore debconf complains
[11:05] <mjg59> Kamion: The new gparted stuff is much nicer, BTW
[11:05] <ogra> ok
[11:05] <Kamion> mjg59: 0.3?
[11:05] <ogra> corrected locally
[11:05] <Kamion> mjg59: and how so? I haven't played with it yet
[11:05] <mjg59> Kamion: Uh, no, whatever's happening in the installer now
[11:06] <mjg59> Rather than in dapper
[11:06] <Kamion> ogra: right, debconf needs Description without underscore - po-debconf does the translation
[11:06] <Kamion> mjg59: oh, right, good
[11:06] <ogra> yup
[11:06] <Kamion> it's not as good as I'd like (I'd like to kill gparted, but that's edgy+1 now), but it's a bit less rough
[11:07] <ogra> mjg59, did you get my ping this morning ?
[11:07] <Kamion> since ubiquity-advanced-partitioner is deferred, I'm going to have to add reiserfs support without libreiserfs to qtparted
[11:07] <mjg59> ogra: Nope
[11:07] <Kamion> how much fun
[11:07] <ogra> the glowing consoles on my laptop are gone ... they are "no-chars and stripes" now 
[11:07] <ogra> mjg59^^
[11:08] <mjg59> ogra: I'm looking into that
[11:08] <ogra> ok, so you know about it ... fine then :)
[11:08] <mjg59> Kamion: So I've just booted an amd64 daily, and in the live environment I'm certainly not on sub-pixel
[11:08] <mjg59> anti-aliasing
[11:08] <Kamion> mjg59: ok, must be a fontconfig bug
[11:08] <ogra> mjg59, start firefox .. thats even worse
[11:09] <mjg59> Yeah
[11:09] <mjg59> Oh, hurrah
[11:09] <mjg59> Latest HP bios fixes the "My nx6125 runs REALLY REALLY SLOWLY" problem
[11:10] <mvo> mdz: I just talked to cprov about the new metapackages section and he said it would be very easy to add this section to the database. we just need your ok :)
[11:14] <mdz> mvo: "ok"
[11:15] <ogra> heh
[11:15] <mdke> hehe, with two simple letters, mdz sets wheels in motion
[11:22] <mjg59> Kamion: Hmm. ubiquity crashed.
[11:22] <mjg59> Filing a bug now.
[11:25] <Seveas> mjg59, changes pushed to lp but will probably take some time to appear (LP builtin delay)
[11:27] <Seveas> mjg59, to sum up: add a -v flag and suppress text by default, 'verbose' on the kernel command line will enable text. Split off a -dev package for out-of-tree theme building and added bling bling example theme with instructions on how to create a theme
[11:27] <Seveas> I also changed INPUT to make sure text was displayed if usplash is runnign without -v :) 
[11:27] <mjg59> Seveas: Rocking
[11:27] <Seveas> is usplash_down actually used? If so, I'll fix that one up too
[11:28] <mjg59> It's meant to be, isn't it?
[11:28] <mjg59> I'm not sure offhand, though
[11:28] <Seveas> because a) it doesn't yet use /etc/usplash so resolution on shutdown != resolution on startup and b_ it doesn't know about -v
[11:29] <Seveas> mjg59, also, INPUT needs a bit more bling bling love, but that won't make feature freeze
[11:29] <mjg59> Kamion: Looks like it's unhappy if the attempt to get security updates fails
[11:29] <mjg59> Seveas: Yeah, no worries
[11:29] <Seveas> anyway, artwork people will be happy
[11:31] <pitti> mjg59: did you see my ftbfs fix patch? I hope it doesn't break on i386
[11:31] <mjg59> pitti: Getting to that
[11:32] <Kamion> mjg59: hmm, ok, I'll look later - buried in console-setup and friends right now
[11:33] <mjg59> Sure, no problem
[11:50] <Keybuk> Seveas: why "verbose", why not "remove quiet" ?
[11:50] <Seveas> Keybuk, point
[11:51] <Seveas> mjg59, I'll change it to Keybuks suggestion when fixing usplash_down (next on my list after grabbing something to drink)
[11:52] <Keybuk> Seveas: which bzr archive is this in?
[11:52] <mjg59> Seveas: Sure
[11:52] <Seveas> bazaar.launchpad.net/~dennis/usplash/theming iirc
[11:52] <Seveas> (bzr remembers push urls, so I don't have to :))
[11:53] <Keybuk> ok, just wanted a look
[11:53] <Keybuk> need to work out how to harmonise this with the usual boot messages
[11:53] <Seveas> this probably isn't visible via lp yet
[11:53] <Seveas> it has a dela
[11:53] <Seveas> y
[11:55] <Keybuk> Seveas: what are the main changes?
[11:56] <Keybuk> mjg59: your branch still exhibits the "nukes the console" bug :p
[11:56] <Seveas> Keybuk, since the current edgy package or everything I did in the past few days? (current edgy package already has the bling bling)
[11:56] <Keybuk> Seveas: what's the bling bling?
[11:57] <Seveas> mainly support for 256 colors in the code and themes being able to override drawing functions for text and progress bar
[11:57] <Keybuk> what are the text changes ?
[11:58] <Seveas> no text by default, usplash can handle text input (not my work), no other changes
[11:58] <slomo> Seveas: as you're working on usplash... do you know if someone is working on fixing it for ppc? i still get bogl saying that it doesn't like my framebuffer (if you want the real output i'll write it down tomorrow)
[11:58] <Keybuk> mjg59: meh, is there any particular reason you added "Edgy Eft Edition" to the usplash testcard?  The intent was that the testcard be completely neutral so that other distributions didn't need to fork the usplash package
[11:59] <Keybuk> Seveas: what happens when it gets the TEXT message?
[11:59] <Seveas> it scrolls existing text up and displays new text
[11:59] <Seveas> (only if running as usplash -v)
[11:59] <Keybuk> right, what happens if the TEXT is "Press ENTER to reboot." or "fsck failed", etc.?
[12:00] <Seveas> you would want to change TEXT to INPUT 
[12:00] <Seveas> or rather the still hypthetical INPUT_ENTER which doesn't echo anything
[12:00] <mjg59> Keybuk: To provide evidence that something was actually happening in the development
[12:01] <mjg59> Keybuk: Sorry, you never mentioned that feature
[12:01] <mjg59> Keybuk: "My branch"?
[12:01] <mjg59> I work on trunk...
[12:01] <Keybuk> mjg59: the one you last committed to :p
[12:01] <Keybuk> sladen and I did talk about making a new hi-res "C" testcard
[12:02] <Trae> Keybuk, https://launchpad.net/distros/ubuntu/+source/acpi-support/+bug/22336/comments/55 see my comment re: fixed for Edgy?
[12:02] <Ubugtu> Malone bug 22336 in acpi-support "laptop overheats when performing CPU intensive tasks." [High,Confirmed]  
[12:02] <Keybuk> we'd need a blackboard, some chalk, a fluffy tux, a digital camera, and you for that though :p
[12:02] <sladen> mjg59: testcard=unbranded and means that usplash never gets revved, only artwork
[12:02] <Keybuk> Trae: still "Confirmed", so no, not fixed for edgy yet
[12:02] <sladen> mjg59: mind if I roll it back?
[12:02] <sladen> mjg59: (possibly replacing it with some overridden artwork?
[12:02] <Trae> ahh ok...  My question is whether or not we should expect it to be fixed for Dapper or Edgy.  
[12:02] <mjg59> sladen: Sure, no problem, but it would be nice to leave more than 16 colours in it
[12:02] <Keybuk> Trae: and unless there's a patch that fixes it on that bug, there's no known way to fix it
[12:03] <mjg59> Trae: Quite possibly fixed in edgy
[12:03] <Trae> mjg59, k
[12:03] <Keybuk> mjg59: what did you change?
[12:03] <mjg59> But depends on what the failure is in your case
[12:03] <mjg59> Keybuk: powernowd bails and uses ondemand where possible
[12:03] <Keybuk> mjg59: yeah, that hasn't fully fixed it for Trae
[12:03] <mjg59> Ok, then no, it's not fixed
[12:04] <Trae> mjg59, my failure is a problem where the laptop shuts off when viewing a video.  I thought it was fixed one time.... but it came back later.
[12:04] <Trae> Keybuk, did you ever see what I posted on the bug ?  about my /etc/rc.local 
[12:04] <Trae> Keybuk, perhaps I put something in wrong.
[12:04] <Trae> cause it only came back after the reboot
[12:04] <Keybuk> Trae: oh
[12:04] <Keybuk> yeah
[12:04] <Keybuk> sh -c 'echo -n ondemand > sys/devices/system/cpu/cpu0/cpufreq/scaling_governor'
[12:04] <Keybuk> should be just
[12:05] <Keybuk> echo -n ondemand > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
[12:05] <Trae> Keybuk, heh, if that's the case then perhaps it _IS_ fixed.
[12:05] <Keybuk> note the extra "/" :p
[12:05] <Trae> OMG
[12:05] <Trae> eye am sofa king we todd did
[12:05] <Keybuk> (and you don't need the sh -c ' ' bit)
[12:05] <Trae> Keybuk, let me try that again for grins
[12:05] <mjg59> Keybuk: I think I've got a handle on the console restoration
[12:05] <mjg59> At least, one that works here for me
[12:06] <Keybuk> I'm guessing by the sudden requirement of an INPUT option that usplash somehow displays on something *other* than /dev/console ?
[12:06] <Trae> Keybuk, should I need the path to echo ?
[12:06] <Seveas> mjg59, when is usplash.conf created?
[12:06] <mjg59> Seveas: On installation
[12:06] <Keybuk> Trae: no, that's a builtin :p
[12:06] <Trae> Keybuk, ok.. hehe
[12:06] <mjg59> bzr: ERROR: Unsupported protocol for url "sftp://bazaar.launchpad.net/%7Eubuntu-core-dev/usplash/ubuntu/": Unable to import paramiko (required for sftp support): No module named paramiko
[12:07] <Kamion> install python-paramiko?
[12:07] <mjg59> Yes
[12:07] <sladen> Recommends:
[12:07] <Kamion> well, paramiko isn't needed for e.g. bzr get http://
[12:07] <mjg59> Yeah
[12:08] <Trae> Keybuk, http://www.shorttext.com/3a6i7
[12:08] <Kamion> I'm not sure, it's one of those difficult edge cases
[12:08] <mjg59> Now, where was I...
[12:08] <sladen> needs a Really-Recommends:
[12:08] <Kamion> like syslinux and mtools
[12:08] <Trae> Keybuk, can you triple check that for me please.
[12:08] <Keybuk> sladen: Recommends is intended to be "really"
[12:08] <Keybuk> Suggests is the "if you want"