[12:02] <ajmitch> yes?
[12:04] <trulux> ajmitch: check the package again
[12:04] <ajmitch> package?
[12:09] <trulux> ajmitch: err, page
[12:09] <trulux> :)
[12:09] <ajmitch> yes, I saw the updates
[12:09] <ajmitch> I'm just trying to see why this makefile is broken
[12:10] <trulux> ajmitch: what's done right now?
[12:10] <trulux> anything changed?
[12:10] <ajmitch> working on pam
[12:12] <Mithrandir> thom: because thunderbird is broken.  I'm going to fix it, I just haven't gotten around to it yet.
[12:14] <trulux> ajmitch: ok
[12:16] <dholbach> brb
[12:18] <dholbach> re
[12:19] <trulux> good night
[12:19] <dholbach> sleep tight, trulux
[12:20] <ajmitch> night trulux 
[12:43] <amu> just joking :)
[12:43] <amu> g8 all
[12:44] <dholbach> sleep tight amu :-)
[12:46] <dholbach> i'm off to bed, too *yawn widely*
[12:48] <dholbach> *wave*
[12:49] <sivang> me too
[12:50] <sivang> night all!
[12:50] <dholbach> bye sivang
[12:50] <mdz> jdub: pong
[01:03] <Kamion> like nis or whatever
[01:04] <jdub> Kamion: hopefully next release we go with libuser :)
[01:04] <jdub> i don't think pulling universe packages is good
[01:05] <Kamion> jdub: these are people with preexisting setups
[01:05] <jdub> perhaps just break nicely if those can't be satisfied?
[01:05] <jdub> yeah, but they're porting them to ubuntu
[01:05] <Kamion> saying "but look, we have this better option!" is precisely no good at all
[01:05] <jdub> there is no better option
[01:05] <Kamion> they're porting one piece to Ubuntu, if we won't work with the rest of their machines they may well just say "oh well, screw 'em" and ditch us
[01:05] <jdub> we don't support some of those things
[01:05] <jdub> (nis is in main though)
[01:06] <Kamion> hm, true
[01:06] <jdub> the right fix is to support those things
[01:06] <jdub> night sivan
[01:06] <Kamion> hm, ok
[01:06] <Kamion> kind of screws them over for explicitly selected packages still, but that's not so much of a problem
[01:07] <Kamion> I think I'd like to have a way to automatically enable universe in kickstart
[01:07] <Kamion> although I am not quite sure what the syntax should look like
[01:07] <zenwhen> hey Kamion is K3B currently broken?
[01:07] <Kamion> how should I know? :)
[01:07] <sivang> night jdub, I'll have the patch ready first thing morning. (too tired for it now)
[01:07] <zenwhen> Oh I thought you were the resident KDE guy
[01:07] <zenwhen> lol
[01:07] <Kamion> zenwhen: not even slightly or remotely
[01:07] <zenwhen> lol
[01:08] <Kamion> I used KDE once in 1999
[01:08] <Riddell> zenwhen: hello
[01:08] <Kamion> or thereabouts
[01:08] <zenwhen> It crashes when I drop in some mp3s
[01:08] <zenwhen> figured I would see if anyone had noticed.
[01:08] <Riddell> zenwhen: possibly restricted format issues
[01:09] <Riddell> although it shouldn't crash
[01:09] <zenwhen> Well it worked fine before I upgraded to hoary
[01:09] <jdub> Kamion: yeah, explicitly enabling it would be very useful
[01:10] <Kamion> jdub: kind of annoying that the format is owned by Anaconda; I'm not sure how to extend it correctly
 ;-) ;-)
[01:11] <Kamion> I guess I'll just add a 'components' directive or something so you can say 'components main restricted universe multiverse'
[01:11] <Kamion> or something like that
[01:11] <Kamion> kickstart != xml ;)
[01:11] <jdub> components files are tho
[01:11] <jdub> hrm, are you handling those?
[01:12] <Kamion> no
[01:12] <Kamion> well, not particularly
[01:12] <Kamion> I will have to provide some kind of crappy shim for system-config-kickstart
[01:13] <jdub> how painful was porting the system-config-kickstart app itself (not so much the rh->debian kickstart assumptions)?
[01:13] <Kamion> jdub: in answer to your previous question, yes, it's easy to break nicely on unsupported stuff, already doing that
[01:13] <Kamion> jdub: mostly just irritating, it's not a very well-written application and has an awful lot of RH/Fedora assumptions besides
[01:13] <jdub> i wonder how sane it would be to use a bunch of the rh system config thingies
[01:13] <Kamion> I had a look - it didn't really help
[01:14] <Kamion> the only one I'm currently wondering about is the stuff in system-config-securelevel, for firewall configuration
[01:14] <Kamion> but I think really that sort of stuff ought to go in base-config and do more Ubuntuish things
[01:15] <Kamion> for the moment I've implemented only 'firewall --disabled' :)
[01:15] <jdub> heh
[01:16] <bob2> erk, icky i810 audio ickiness
[01:16] <daniels> heh :)
[01:16] <Kamion> jdub: oh, by RH assumptions, I mean more stuff like the rhpl libraries
[01:16] <daniels> bob2: you don't have to hack the ALSA driver to get it working on your chipset :\
[01:17] <jdub> Kamion: ah, yeah
[01:17] <jdub> Kamion: mostly localised to those?
[01:17] <jdub> nice doesn't have a huge effect on IO load, does it?
[01:17] <bob2> daniels: it's got weird distortion after like 20 sleep/resume cycles
[01:17] <Kamion> rhpl stuff was primarily gettext plus a couple of hardware detection pieces
[01:17] <Kamion> I did the gettext stuff differently (and cheesily), and stubbed out the hardware detection for now
[01:17] <daniels> bob	bong
[01:18] <bob2> otoh, I'm happy sleep Just Works all the time now
[01:18] <bob2> except I'm not game to try suspend-to-disk again so I can swap batteries
[01:23] <Kamion> jdub: any idea what /usr/bin/htmlview might do, and what it should be replaced by?
[01:24] <Kamion> /usr/bin/x-www-browser maybe?
[01:25] <jdub> isn't htmlview like html2text?
[01:27] <Kamion> jdub: doesn't look like it from the code, it's connected to Help -> Contents
[01:30] <mdz> jbailey: what is the word on the initrd features you were working on?
[01:30] <mdz> jbailey: will they make it in time for the freeze?
[01:31] <jbailey> mdz: The DSDT stuff was uploaded earlier today (as well as fixes for most of the bugs I have assigned)
[01:32] <jbailey> mdz: The NFSRoot initramfs is iffy.  I either need to update linux-kernel-headers (ugly) or find a better solution (we've talked about linux-libc-headers in #debian-glibc) for klibc.
[01:32] <jbailey> mdz: I'm planning on hacking on the NFSRoot stuff a bit tonight to see if I can quickly compile something sane or if I should just punt it.  It's doing NFSBoot stuff fine for me, though.  integrated hotplug/busybox/tie into the kernel build isn't for Hoary.
[01:33] <jdub> Kamion: hrm, i don't have htmlview
[01:33] <jdub> oh
[01:33] <jdub> oh
[01:33] <jdub> sorry
[01:33] <Kamion> I think it's an RHism
[01:34] <jdub> yes, that's a red hat wrapper for the browser
[01:34] <jdub> i'd suggest using gnome-open
[01:34] <Kamion> what does that provide over x-www-browser? a GNOME dependency? :)
[01:35] <Kamion> ew, /usr/bin/gnome-open is provided by a library package
[01:35] <Kamion> so wrong
[01:36] <Kamion> next time libgnome's soname changes that will cause upgrade hell
[01:36] <lamont> jdub: thoughts on the postfix->shipseed mail?>
[01:37] <mdz> jbailey: how would linux-libc-headers differ from linux-kernel-headers (apart from eliminating a lot of confusion...)
[01:37] <jdub> lamont: i read it and cried
[01:38] <jdub> Kamion: iz gtk boog :)
[01:38] <lamont> jdub: it's a good thing - it lets postfix be functional at install.
[01:38] <Kamion> jdub: what would the correct way of depending on gnome-open be? kickstart does not currently depend on anything GNOME, and I don't want to add an explicit dependency on libgnome2-0 because that's asking for trouble
[01:38] <lamont> but I wanted to let you have a good cry before I did ti...
[01:38] <jdub> Kamion: isn't this for s-c-k?
[01:39] <Kamion> jdub: sorry, yeah, s-c-k. it only depends on gtk and glade at the moment.
[01:39] <jdub> lamont: the thing that worried me the most was removing everything that may need an mta, like mutt
[01:39] <Kamion> removing mutt kind of concerned me
[01:39] <lamont> understandable
[01:39] <lamont> but it's still on the CD...
[01:39] <jdub> Kamion: perhaps we should chat to seb about gnome-open
[01:40] <jbailey> mdz: linux-kernel-headers is the package that gotom, doko, and I put together to try and solve thge problem, but haven't really had time to keep up to date.  Most distro's have their own magic mess of the kernel headers tuned to userspace.  linux-libc-headers (http://ep09.pld-linux.org/~mmazur/linux-libc-headers/) is someone else's effort to do the same thing.  He seems to be doing things well, has been keep
[01:40] <jbailey> ing it up to date, and has been doing it for a year so far.
[01:40] <Kamion> jdub: ok, I'll file a bug if I remember, I'll use x-www-browser for now
[01:40] <jdub> ok
[01:40] <jdub> Kamion: perhaps you could suck it out of gconf via gconftool-2? :)
[01:42] <jbailey> mdz: My thought for now was to package l-l-h to go into /usr/include/l-l-h.  Tell klibc to use it for now.  That would give me exposure to the package and make it easier to propose replacing l-k-h
[01:42] <jdub> etc/X11/sysconfig/synaptic.desktop <- wtf?
[01:42] <jbailey> mdz: (Either here or in upstream Debian.  I still don't know my relationship to glibc in Ubuntu. *g*)
[01:42] <Kamion> jdub: I'll refrain for killing you because you're useful ;)
[01:42] <Kamion> er, refrain *from*. my English good, I learn him from a book.
[01:43] <jdub> heh
[01:43] <jdub> from a booo-ook
[01:43] <Nafallo> daniels: ping?
[01:43] <daniels> http://itax.sourceforge.net/itax3.png <- notice http://ubuntu/itax/
[01:43] <daniels> Nafallo: pong?
[01:44] <Nafallo> daniels: #5925. Is my last comment something to investigate?
[01:46] <daniels> Nafallo: well, that the drm changes in -11 broke stuff has always been known
[01:47] <elmo> Kamion: can you update little to also trigger syowa.ubuntu.com pls?
[01:47] <elmo> preferably first
[01:48] <elmo> err
[01:48] <elmo> sorry, synxprocy.ubuntu.com
[01:48] <Nafallo> daniels: and are there any solutions in sight or should I try to compile the kernel with CONFIG_DRM=m?
[01:48] <Kamion> elmo: syncproxy?
[01:49] <elmo> Kamion: yah, it's what mirrors are going to sync off of
[01:49] <elmo> well are
[01:49] <Kamion> elmo: I was checking the typo correction :)
[01:49] <elmo> oh, duh, right
[01:49] <daniels> Nafallo: well, I have other things to do today (mainly, upload xorg), but it's on my todo list to look at.  if you can beat me to a solution, sweet deal
[01:50] <jdub> daniels: dude
[01:50] <jdub> daniels: rock!
[01:50] <daniels> jdub: ?
[01:50] <jdub> daniels: itax
[01:50] <daniels> oh, right
[01:50] <Nafallo> daniels: oki. as written I have to study, but I'll look into it this weekend if you haven't practiced magic and solved it :-).
[01:50] <daniels> I thought you were stoked about itax being on my TODO
[01:51] <daniels> cool, thanks
[01:51] <Kamion> elmo: done
[01:51] <lamont> jdub: I suppose we _could_ have mutt Recommend MTA, but that just feels wrong.. :-)
[01:51] <jdub> daniels: itax is on your todo?
[01:51] <elmo> Kamion: tnx
[01:52] <daniels> jdub: s/itax/fglrx/
[01:52] <Kamion> oh, crap
[01:52] <Kamion> mdz,jdub: can I put system-config-kickstart in supported?
[01:53] <Kamion> part of the kickstart feature goal, I'd forgotten it was still in universe
[01:53] <lamont> jdub/mdz: permission to make the seed changes for the mta moving?>
[01:54] <lamont> or do we  need to discuss it more?
[01:54] <lamont> I think that the recommendation is consistant with the decision in the tech board.
[01:54] <elmo> Kamion: fyi, we now have a kick ass us.archive.u.c, if you want a real world use case for the installer changes ;)
[01:55] <lamont> elmo: will we also have a them.archive.u.c? :-)
[01:55] <Kamion> elmo: oh damn, didn't get round to doing that today ..
[01:55] <elmo> lamont: we.have.whatever.you.god.damn.want.archive.u.c
[01:55] <lamont> lol
[01:56] <elmo> see now, I have to go update my sources.list to use that
[01:56] <jdub> Kamion: yes
[01:57] <elmo> hmm, woops, helps if apache knows about the wildcard
[01:57] <Kamion> jdub: done, thanks
[01:57] <jdub> lamont: feels a bit early, given lack of feedback
[01:57] <lamont> jdub: is ok post-feature freeze, then?
[01:58] <lamont> assuming feedback goes that way, of course.
[01:58] <jdub> lamont: i think so, it is a change we've planned for
[01:58] <jdub> and not really a feature
[01:58] <jdub> it's a defeature if anything ;)
[01:58] <lamont> well, one could argue that point either direction, but OK.
[01:58] <lamont> well, having postfix uncrippled would be a nice bugfix/feature
[01:59] <jdub> heh
[02:02] <mdz> Kamion: yes
[02:03] <lamont> Kamion: the liveCD's from this morning have the -15 kernel in them, yes?
[02:03] <mdz> lamont: it's after 1700 and I haven't made it to ubuntu-devel (~3rd in mailbox order:)
[02:03] <Kamion> argh, complicated parsing bug in kickseed
[02:03] <lamont> Kamion: that's what you get for parsing complicated things, no? :-)
[02:04] <Kamion> lamont: -15> yes
[02:04] <elmo> mdz: slacker
[02:04] <Kamion> well, kickstart files use shell-style quoting, so I was running them through eval as a cheap way to dequote them
[02:05] <mdz> thom: gtimelog?
[02:05] <Kamion> but now I discover that they have unquoted metacharacters (specifically pre-crypted md5 root passwords start with $1$ ...)
[02:05] <lamont> mdz/jdub: but it's ok to go change 'exim4 | mail-transport-agent' to 'postfix | mail-transport-agent' everywhere in main/restricted, yes?
[02:05] <daniels> Kamion: d'oh
[02:06] <lamont> Kamion: ouch
[02:06] <mdz> Kamion: what do you think about the sudo group patch?
[02:06] <Kamion> mdz: I still have trouble with the name, I think adm/admin is asking for confusion
[02:07] <jdub> mdz: (rock!)
[02:07] <mdz> Kamion: I agree, but I haven't a better idea
[02:07] <mdz> jdub: how about you?
[02:07] <mdz> jdub: (name for the group)
[02:07] <dilinger> not bad.  57 seconds between the grub prompt and gdm prompt
[02:07] <Kamion> how about 'rootsudo'?
[02:07] <elmo> mdz: svn co http://mg.pov.lt/gtimelog/svn/ gtimelog
[02:07] <mdz> sudo is taken for something else
[02:07] <dilinger> (including hostap firmware uploading and dhclient pauses
[02:08] <mdz> elmo: is it usable?
[02:08] <Kamion> although rootsudo is a bit insanely implementation-specific
[02:08] <mdz> I've tried some other programs like that and found them less than ideal
[02:08] <tseng> sudoers might be valid
[02:08] <elmo> mdz: dunno, haven't tried, but the lunchpadders seem keen
[02:08] <jdub> mdz: admin seems confusing to me too
[02:08] <jdub> mdz: we really want to avoid 'wheel'?
[02:08] <elmo> mdz: it's written by Marius, who was at Mataaro
[02:08] <Kamion> jdub: wheel is gid 0 on BSD, ultra-confusing
[02:08] <elmo> so if it's usable enough, he's probably persuadable to fix it the rest of the way
[02:08] <jdub> Kamion: mmm, true
[02:08] <jdub> WDOD? (what does osx do?)
[02:09] <Kamion> OS X uses the wheel group, I think
[02:10] <elmo> I guess 'r00t' isn't an option?
[02:10] <Kamion> oh, wheel -> admin in OS X >= 10.3 apparently
[02:10] <mdz> jdub: yes, wheel has too much baggage
[02:10] <jdub> Kamion: hmm
[02:10] <Kamion> (http://developer.apple.com/documentation/Security/Conceptual/Security_Overview/Concepts/chapter_3_section_9.html#//apple_ref/doc/uid/TP30000976-CH203-TPXREF121)
[02:11] <mdz> jbailey: so your hotplug upload, does it address the ide-generic-loaded-last issue, or not yet?
[02:11] <Kamion> I also think <= 8 characters would be a good idea
[02:11] <mdz> agreed
[02:12] <elmo> why do we need a group again?
[02:12] <mdz> elmo: adding users to a group is simpler than modifying /etc/sudoers
[02:13] <mdz> lets users add new users with admin privileges using the GUI
[02:13] <mdz> prevents us from having to edit /etc/sudoers on the fly during install
[02:13] <Kamion> I dunno, I guess admin is OK, it's really adm that's misnamed
[02:13] <mdz> though i don't think the current patch implements that last bit yet
[02:13] <mdz> admin is the least evil of those we've discussed
[02:13] <jdub> i endorse admin
[02:13] <mdz> confusion >> accidental vulnerabilities
[02:14] <jbailey> mdz: No, but because PCI is loaded first, I think that should take care of it.
[02:14] <infinity> You can't use "wheel"... "wheel" is gid 0 on many systems (same as "root" on Debian/Ubuntu)
[02:14] <infinity> That's just confusing as all get out.
[02:14] <jbailey> mdz: (I'm just running off for food, I'll reply as soon as I'm back)
[02:14] <mdz> jbailey: for coldplugging anyway, which is all we really care about I think
[02:14] <Kamion> infinity: yeah, I mentioned the BSD case above and I think everyone agreed
[02:15] <Kamion> OK, I'll implement it with admin then
[02:15] <jdub> mdz: so, would it be dangerous to assume that people who should have sudo access (admin group) also be able to read logs and so on (adm group)?
[02:15] <Kamion> argh, the patch is evil though
[02:15] <jbailey> mdz: For this round.  I'm going to provide a setup and work with dilinger to try and solve the ide hotplug thing properly.  No real idea on timing for it, though.
[02:15] <Kamion> it randomly allocates gid 80 for itself
[02:15] <jdub> or am i missing some of the love that adm gets?
[02:15] <mdz> jdub: I don't think we can safely change ownership of log files to admin, if that's what you mean
[02:16] <Kamion> jdub: one change at a time ...
[02:16] <mdz> jdub: what Kamion said, and also "feature freeze" :-)
[02:16] <jdub> mdz: no, don't go too far forward, only asking exactly what i asked :)
[02:16] <Kamion> hm, so, I think I'm going to implement the admin group as a normal system group, not one in the 0-100 range
[02:16] <Kamion> 0-99
[02:16] <Kamion> although I could be persuaded otherwise
[02:17] <mdz> I don't think it needs to be in base-passwd
[02:17] <Kamion> however I was kind of hoping to avoid branching base-passwd for Ubuntu
[02:17] <infinity> Kamion : There's no real reason to have a static group at all, is there?
[02:17] <mdz> addgroup --system seems fine to me
[02:17] <Kamion> since I'd have to allocate the gid in both places anyway
[02:17] <Kamion> infinity: can't think of one
[02:18] <infinity> I thin kthe only problem you may run into is that a couple thousand users out there may already have groups named "admin".
[02:18] <Kamion> that is indeed uncomfortable; we won't be making this change on upgrades though
[02:18] <infinity> In which case, you want to name yours "Homage-to-Debian-Exim"
[02:19] <Kamion> (it's just too complicated and error-prone to try to do it on upgrades)
[02:19] <daniels> inf	heh
[02:19] <infinity> Kamion : Oh, is this a base-config change, not owned by sudo?
[02:19] <Kamion> ... which means that users-admin probably needs to check for %admin in sudoers and print "you lose, kthxbye" otherwise
[02:19] <Kamion> infinity: no, it's owned by sudo, but the sudoers file won't be changed on upgrade
[02:20] <Kamion> we can release-note it anyway
[02:20] <infinity> What if I install sudo after having created an admin group?
[02:20] <Kamion> sudo's in Ubuntu base so that would only happen on Debian->Ubuntu sidegrades
[02:21] <infinity> Check.  Let the Debian users work it out.  They're supposed to be smart.
[02:21] <Kamion> I could print a scary warning though
[02:21] <daniels> (in which case, you've already lost on group memberships anyway)
[02:21] <mdz> Kamion: do you have an opinion on lamont's MTA proposal?
[02:22] <infinity> Kamion : Well, if you're already testing upgrade vs. fresh install to determine whether or not to add the bits to sudoers, you may as well just test for the existance of the group too, and treat that as an upgrade.
[02:22] <infinity> Kamion : No warnings required then, just a small loss in fucntionality.
[02:24] <Kamion> mdz: it still feels wrong to me to ship without a mail-transport-agent and I'd much rather have Keybuk's dummy one, but if it can be made to work I think it's probably better than the current broken-postfix situation
[02:25] <Kamion> infinity: the corner cases worry me, because both sudo and passwd need to ensure that the group exists
[02:26] <mdz> lamont: I say let's do it; if we decide that it was a huge mistake, it's a simple matter to migrate to Keybuk's mini-MTA
[02:26] <Kamion> infinity: I think I might just print a warning; if you're installing sudo from scratch, chances are it's part of a relatively small upgrade anyway
[02:26] <daniels> Kamion: alternately, you could touch /var/lib/sudo/.have-i-created-the-group-yet
[02:27] <mdz> that could very well happen
[02:27] <mdz> but at least we'll have tried it
[02:27] <infinity> daniels : How does that help?
[02:28] <infinity> Kamion : What does base-passwd do on upgrades if a change it wants to make clashes with a group/user already on the system?
[02:32] <Kamion> infinity: base-passwd always asks, but its situation is different
[02:32] <Kamion> I'm just going to go with the scary warning, I think
[02:33] <infinity> Well, given the sensitive nature of a "default admin group", that would be the only reason I'd treat it carefully.
[02:33] <jdub> Kamion: sounds good
[02:33] <jdub> Kamion: really happy this is going in :-)
[02:33] <infinity> Make the scary warning blink. :)
[02:34] <Kamion> WARNING: An admin group already exists! This group is used to grant root
[02:34] <Kamion> privileges. Please audit the membership of your admin group and make sure
[02:34] <Kamion> this is appropriate.
[02:35] <Kamion> added "using sudo" after "root privileges"
[02:35] <jdub> that was my only suggestion
[02:35] <infinity> That should do.
[02:35] <jdub> which i didn't even have time to fully type :)
[02:36] <Kamion> actually
[02:36] <Kamion> sudo's postinst already has code to ask you whether you want to abort the installation if gid 27 is not sudo
[02:37] <Kamion> so it clearly cannot be a problem to do the same thing for admin :)
[02:37] <infinity> I'd say not.
[02:37] <infinity> Since admin is far more important than sudo.
[02:38] <infinity> Wait, gid 27 is sudo?
[02:38] <Kamion> yes
[02:38] <infinity> Is that an Ubuntu-only thing? :)
[02:38] <Kamion> hell no
[02:38] <Kamion> <cjwatson@cairhien ~/src/debian/base-passwd/trunk/base-passwd>$ grep sudo group.master
[02:38] <Kamion> sudo:*:27:
[02:38] <infinity> Oh, wait.
[02:38] <Kamion> that has been in Debian since at least 1996
[02:38] <infinity> Wrong system.
[02:39] <infinity> I was on a NetBSD box.
[02:40] <mdz> is Keybuk in .za?
[02:46] <jdub> hrm
[02:47] <jdub> is there a shell builtinish way of determining if a string is in a variable?
[02:47] <daniels> mdz: not that I know of
[02:47] <Kamion> jdub:
[02:47] <Kamion> case $var in
[02:47] <Kamion>     *string*)
[02:48] <Kamion>         do stuff
[02:48] <Kamion>         ;;
[02:48] <Kamion> esac
[02:48] <jdub> ah, sensible - thanks!
[02:48] <lamont> jdub: if it's at the start or end, you can use ${var#string} or ${var%string}
[02:48] <lamont> and then compare to $var
[02:49] <jdub> no, variable will be \n delimited
[02:51] <lamont> ew
[02:51] <jdub> i could do space delimited
[02:51] <jdub> other option is grep, but i'd like to avoid that
[02:51] <jdub> this is a rarely true branch in a tight loop
[02:52] <mdz> it's shell; how tight could it be? ;-)
[02:52] <Kamion> actually, I have an even better idea: let's do the %admin user business in passwd.config, where the initial user is created, and not in sudo.postinst at all
[02:53] <Kamion> that way there is no risk that it might affect upgrades
[02:55] <infinity> jdub : You know you'll just have to optimise it later when that shell script reaches 500k.  Why not start early?
[02:55] <mdz> daniels: when do you expect to upload new xorg?  the duplicate bugs are coming fast and furious
[02:56] <mdz> Kamion: that sounds right
[03:02] <daniels> mdz: to be fair, most of the bug noise was me triaging ;) but hopefully today.  i'm just getting all the stuff I need to build and test X installed on the amd64 (given the implosion of my desktop yesterday), so hopefully tonight?
[03:02] <lamont> "never unroll loops" heh.
[03:17] <Kamion> mdz: ok, simpler shadow-only version uploading
[03:19] <tritium> trulux, I'm back.  How's the LaTeX going?
[03:20] <lamont> is there a working undelete for ext3?
[03:21] <robertj> lamont: not that I know of, is there a reason for that question?
[03:21] <lamont> yeah
[03:22] <robertj> what happen?
[03:22] <Kamion> not to my knowledge either, save for trawling through the disk for blocks you recognise
[03:22] <lamont> I wanna undelete a bunch of files I just nuked
[03:22] <Kamion> AFAIK ext3 zeroes the inodes
[03:22] <lamont> not difficult to recover, just annoying
[03:22] <mjg59> lamont: There's no undelete
[03:22] <lamont> oh well.
[03:22] <mjg59> You might get some stuff by grepping through the disk
[03:22] <lamont> not worth the pain
[03:22] <robertj> yeah I've grepped stuff back into existance
[03:23] <lamont> it's just a few downloads from today that will have to be done again.
[03:25] <Kamion> night all
[03:26] <lamont> night Kamion 
[03:26] <lamont> actually, there were some dups, so I can actually recover much of it
[03:30] <jdub> $ dpkg -L xscreensaver | grep s.desktop$
[03:30] <jdub> /usr/share/control-center/capplets/screensaver-properties.desktop
[03:30] <jdub> /usr/share/control-center/Desktop/screensaver-properties.desktop
[03:30] <jdub> /usr/share/control-center-2.0/capplets/screensaver-properties.desktop
[03:30] <jdub> /usr/share/applications/screensaver-properties.desktop
[03:30] <jdub> 
[03:30] <jdub> elite!
[03:36] <AndyFitz> skills that killz
[03:36] <daniels> jdub is the vanillah killah
[03:37] <jdub> and i make yo momma cry
[03:37] <srbaker> anyone here know JL Boers?
[03:38] <srbaker> anyone here using a T22 and getting middle clicking or scrolling?
[03:59] <mdz> daniels:ok, mumble mumble feature freeze mumble mumble
[03:59] <jdub> ogra: xss patch in :)
[04:00] <mdz> daniels: so VT switching actually works for you on the live CD?
[04:01] <robertj> are the gnome vfs handlers universally b0rk?
[04:40] <Rotund> I want to get Ubuntu to auto-detect multiple video cards.  what is the program the installer uses to call xresprobe?
[04:43] <mdz> Rotund: xserver-xorg.config
[04:43] <mdz> Rotund: and the person maintaining it is daniels
[04:43] <Rotund> great.  
[04:43] <Rotund> thanks
[04:46] <Rotund> how do I get that file?  I'm still on warty and don't want to add repositories
[04:46] <Rotund> would it be in xorg_6.8.1.orig.tar.gz?
[04:48] <farruinn> Rotund: you're talking about mixing warty and hoary, not a fun thing
[04:48] <Rotund> yeah.  I know.  I just want the latest xserver-xorg.config
[04:49] <Rotund> I upgraded to hoary a while ago and it crashed too much for my tastes
[04:49] <Rotund> (I'm sure it's better now, but still little desire)
[05:01] <mdz> Rotund: apt-get source xorg
[05:01] <mdz> with a hoary deb-src entry in /etc/apt/sources.list
[05:01] <Rotund> ohhh.  that's smart, but I just found it too
[05:01] <Rotund> thanks
[05:03] <Rotund> how would Ubuntu deal with such things
[05:04] <Rotund> I know the general idea is "it just works"
[05:06] <Rotund> obviously, with dual-head would mean one would need to at least need to determine the monitor configuration (which is 1, 2, etc)
[05:06] <jdub> Rotund: that'd be a very welcome hack
[05:06] <Rotund> and the no NVIDIA driver in the default makes it even harder
[05:06] <Rotund> jdub: great.  I know I missed it on my wife's computer w/ 3 monitors
[05:06] <Rotund> I need to talk out how it should work
[05:07] <Rotund> would a Windows style "display properties" be easier?
[05:07] <jdub> i thought about this a while back
[05:07] <Rotund> great.  what was your thoughts?
[05:07] <jdub> setting up a configuration by default would be most useful
[05:07] <jdub> the only way you can really do that sanely is:
[05:07] <jdub> - only enable heads that have monitors attached (ie. you can get ddc info)
[05:08] <Rotund> great idea
[05:08] <Rotund> BTW: which discover does hoary use?
[05:08] <HrdwrBoB> AGP->first pci card->second pci card left to right default config
[05:08] <jdub> - only enable multiple heads if you can determine a 'primary' output in a reasonable way (ie. on dual head cards, that's easy, on an agp+pci machine, you'd be making a reasonable assumption, on other machines it would be tougher)
[05:09] <jdub> Rotund: xresprobe and ddcprobe, you'd have to run these on each available video card output
[05:09] <zul> there kernel team stuff added to the wiki
[05:09] <jdub> - if you can't enable it by default given the above points, add the configuration to xorg.conf but comment it out
[05:09] <Rotund> okay.  currently it looks like xserver-xorg.config uses discover
[05:10] <jdub> - having a display properties page that was gnomey and sexy and handled this kind of stuff would be ideal, but also very hard
[05:10] <Rotund> okay
[05:11] <Rotund> Another question (as people have asked me).  How do you have ubuntu auto detect the vid card again?
[05:11] <jdub> should all be in the .config
[05:12] <jdub> hotplug loads the right drivers, though
[05:12] <daniels> jdub: actually, you don't get to do multi-card DDC outside the X server
[05:12] <daniels> only the primary one, since we call via VIE
[05:12] <daniels> er, VBE
[05:12] <jdub> daniels: so you'd have to run X to do the ddc? that doesn't sound terrible
[05:13] <daniels> jdub: it's not terrible, but parsing the log output is pretty horrific
[05:14] <jdub> so, assuming you can parse the log output sanely, is running X for ddc something you'd want to avoid in the common case (1 output)?
[05:18] <Rotund> so, the installer would have to boot x?
[05:18] <HrdwrBoB> I would think so, but you couldn't tell without starting X
[05:18] <daniels> jdub: it can be done, but 'sanely' is pushing it
[05:18] <jdub> Rotund: the install process already does
[05:19] <Rotund> really?
[05:19] <Rotund> did it in warty?
[05:19] <daniels> well, not for DDC
[05:19] <daniels> but if you're on a laptop, second-stage installer fires X up
[05:20] <infinity> daniels : I assume the latest nv driver knows how to make my card go?
[05:20] <infinity> daniels : warty on my new machine was less spectacular than I'd hoped. :)
[05:21] <infinity> (Well, until I installed the binary driver..)
[05:21] <Rotund> the nv driver isn't too good
[05:21] <daniels> infinity: Yeah, nv in Hoary should be fine.
[05:22] <infinity> Rotund : It's good enough to make a livecd work.
[05:22] <Rotund> I'm assuming the nvidia driver will never be on the default install CD?
[05:22] <infinity> Having a system try really, really hard to boot to X, then dump to a console and flip you the bird is no fun.
[05:22] <Rotund> =)
[05:22] <Rotund> I didn't have that issue w/ mine, but I had an old card at the time
[05:23] <infinity> Yeah, mine was a shiny new 6800GT.
[05:23] <infinity> It's been a long time since I had a card newer than the X driver I was trying to run.
[05:23] <infinity> I'd almost forgotted that happens.
[05:23] <Rotund> oh yeah.  no chance on the ancient XFree drivers =)
[05:27] <daniels> 686-smp covers that just fine
[05:27] <Rotund> really?  I just saw i686 and thought an original Pentium wouldn't work
[05:28] <Rotund> If it supports i586, why not call it i586?
[05:28] <daniels> Rotund: 686 is original pentium, iirc
[05:28] <Rotund> no, 686 is the P2 (perhaps PPro too)
[05:28] <daniels> interesting.  in any case, what's out ther eas '586' certainly isn't the pentium.
[05:29] <Rotund> 586 is the Pentium, Pentium MMX, x586, MII (I think), K5
[05:31] <Rotund> I think the Media GX was 586 (guess when I was really into hardware)
[05:31] <Rotund> is XOSD supported on all video cards?
[05:38] <Rotund> general X question.  Does xinerama need for each screen to be the same display resolution?
[05:39] <Rotund> (all be 1024x768)
[05:39] <HrdwrBoB> no
[05:39] <Rotund> great
[05:40] <HrdwrBoB> xinerama has massive limitations though
[05:40] <Rotund> oh?
[05:40] <HrdwrBoB> eg: no 3d
[05:40] <Rotund> yup
[05:40] <HrdwrBoB> video overlay can be .. weird, or not work at all
[05:40] <infinity> daniels : Why bother enumerating unsupported cards?... if it's unsupported and lspci thinks it looks like a VGA adapter, just throw a signal at it.  How bad can that go? :)
[05:40] <Rotund> yup.  and the sleep stuff too
[05:40] <infinity> daniels : Windows has been managing that for ages with very little side effect.
[05:41] <infinity> daniels : The odds of it being both unsupported AND not a PCI VGA device are so slim, you'd cover pretty much all cases of unsupported cards being able to display /something/.
[05:43] <Rotund> what is the program that dumps the hard to parse info about the current screen?
[05:44] <Rotund> HrdwrBoB: I also noticed xrandr didn't work in xinerama
[05:44] <HrdwrBoB> xdpyinfo
[05:45] <Rotund> BRB
[05:45] <daniels> infinity: 'Swhat we do, but if it's from a specific vendor (ati, nvidia, intel), and we don't know about it, we throw it at the default driver for that.
[05:46] <infinity> daniels : ... Which breaks if the default driver doesn't support my new PCI ID, so why bother?
[05:47] <jdub> daniels: is there a known problem with agpgart not loading in current X?
[05:47] <daniels> infinity: discover1-data was in sufficiently bad shape, even after I loved it quite extensively, that this was IMO the best choice.
[05:47] <daniels> jdub: Er, loading agpgart isn't X's job.  It's hotplug's.
[05:48] <jdub> daniels: i am being sleepy
[05:48] <jdub> daniels: s/X/kernel/
[05:48] <daniels> jdub: not sure, sorry -- seems to work fine
[05:51] <marcin_ant> jdub: hi
[05:51] <marcin_ant> jdub: could you tell me something about "ImprovedPanel"
[05:52] <marcin_ant> jdub: from "secondary goals" section
[05:52] <marcin_ant> jdub: from http://www.ubuntulinux.org/wiki/HoaryGoals
[05:53] <daniels> fuck me
[05:53] <daniels> jdub: http://gnome-look.org/content/preview.php?preview=1&id=18178&file1=18178-1.jpg&file2=&file3=&name=Tobacco+Sky&PHPSESSID=09b3ea2d25052f3ba8ff2b0c8d93d8e6
[05:53] <daniels> jdub: also, http://www.gnome-look.org/content/show.php?content=17346&PHPSESSID=c43d53d9d12edc60e4ae0849b45a9c93
[05:55] <infinity> Purdy.
[06:01] <Rotund> which would be more "gnomey" an Advanced tab or an Advanced button for configuring the finer details of the video card?
[06:01] <Rotund> (adding options, changing driver, etc)
[06:01] <Rotund> or none of that at all?
[06:06] <daniels> wow, this amd64 si frighteningly fast.
[06:06] <daniels> Rotund: neither, really
[06:07] <Rotund> okay.  so someone should have to manually edit their config file to change from nv to nvidia?
[06:07] <Rotund> (Okay this is probably somewhat outside what GNOME should be doing in the first place... more of a distro thing)
[06:10] <daniels> Rotund: that said, a graphical utility would be absolutely welcome
[06:10] <Rotund> =)
[06:10] <Rotund> okay
[06:10] <daniels> the plan for hoary was to have the package just set up, out of the box, something that *worked*
[06:10] <daniels> if you wanted to get fancy, we could have an all-singing, all-dancing graphical utility
[06:10] <Rotund> how do I get xresprobe to do something?
[06:10] <daniels> the former has been largely satisfied ...
[06:10] <daniels> er, run it? :)
[06:10] <daniels> sudo xresprobe $drivername
[06:11] <Rotund> what is the $drivername?
[06:11] <Rotund> =) sudo =)
[06:12] <jdub> the name of the driver for your card
[06:12] <Rotund> what happens if I have 2 nvidia cards installed?
[06:12] <jdub> no idea
[06:12] <Rotund> or in my wife's case, 3 ati cards =)
[06:12] <daniels> Rotund: many things, none of which are the right thing
[06:12] <Rotund> SWEET
[06:12] <daniels> well, it's the right thing for the installer, but not the right thing to set the cards up
[06:12] <daniels> it's kind of misleading, because all it does is asks the VESA BIOS 'hey, what up'
[06:13] <Rotund> NICE
[06:13] <daniels> this means it will only ever ask one card (i.e. $drivername is meaningless unless you're on a laptop), and that card is determined by the BIOS
[06:13] <Rotund> okay.  So, what should I use instead?
[06:13] <daniels> errrrrr
[06:13] <HrdwrBoB> Rotund: or in my case
[06:13] <HrdwrBoB> the name of one card
[06:13] <HrdwrBoB> and the attached monitor of the other
[06:13] <HrdwrBoB> in any case, it doesn't produce meaningful results
[06:13] <daniels> you probably want to look at xprobe.sh and lcdsize.sh, and look into starting X with -probeonly, and grepping the log
[06:14] <daniels> unfortunately, the grep logic in this case will be totally batshit insane
[06:14] <Rotund> Also, mine is stupid as it has the right monitor and the wrong specs anyways =)
[06:14] <daniels> if you're using warty, then yes, the sync ranges calculated are not necessarily what your monitor can do
[06:14] <Rotund> no, it only goes to 1024x768.  of couse, my monitor is currently at 1600x1200
[06:15] <daniels> cool
[06:15] <Rotund> ummm.  something like that
[06:16] <daniels> if that's hoary doing it, i'd love to know about it
[06:17] <Rotund> no, I'd have to reboot to the live CD
[06:17] <Rotund> too lazy =)
[06:17] <Rotund> Does it use a lookup table or just ask?
[06:20] <daniels> it probes the monitor for its list of supported resolutions
[06:20] <daniels> in the case of a laptop, it starts up X and looks for a rather complicated list of magic strings
[06:20] <daniels> if it really has no idea, it asks
[06:21] <Rotund> okay. 
[06:21] <Rotund> it's a desktop
[06:21] <Rotund> CRT screen
[06:22] <Rotund> DELL P780.  It hasn't detected it right yet
[06:23] <Rotund> why does this work?  sudo xresprobe nvidia23
[06:26] <daniels> because the driver argument is unused unless you're on a laptop
[06:26] <Rotund> It's hard to try to get a display properties that doesn't have the ugliness of the advanced screen from Windows
[06:26] <daniels> in the future, it may be used
[06:27] <Rotund> 23?
[06:27] <daniels> it depends how seriously we go down the grep-x-log-for-ddc-results insanity path
[06:27] <daniels> er, yes.  dude, it's unused.  nothing looks at it, unless you're on a laptop.
[06:27] <daniels> i routinely enter x in there
[06:27] <fabbione> morning guys
[06:27] <Rotund> hahah
[06:27] <Rotund> morning
[06:33] <jdub> anyone got openoffice.org2 packages installed atm/
[06:33] <jdub> ?
[06:33] <tritium> jdub, yes
[06:34] <jdub> tritium: can you see if the .desktop files are separated into the subpackages, ie. openoffice.org2-writer
[06:34] <tritium> jdub, sure
[06:34] <Rotund> daniels: xresprobe only looks for the following above 1024x768.  1280x1024x75
[06:34] <Rotund> THAT'S IT
[06:34] <Rotund> even the newest version
[06:35] <daniels> ok, so your monitor's lying
[06:35] <tritium> jdub, they are separate
[06:35] <tritium> ooo2-calc.desktop, etc.
[06:35] <jdub> tritium: ie. that particular file is in openoffice.org2-calc?
[06:36] <tritium> jdub, yes, checked with dpkg -L
[06:36] <jdub> thanks
[06:36] <syn-ack> Hey everyone.
[06:36] <tritium> jdub, sure.
[06:37] <Rotund> daniels: my monitor can only do 1280x1024 at 72, so it didn't lie
[06:37] <daniels> Rotund: ok, so what's the actual probleM?
[06:38] <Rotund> xresprobe doesn't detect much above 1024x768
[06:39] <Rotund> does it detect anything better on your monitor?
[06:41] <daniels> what?
[06:41] <daniels> dude, it detects exactly what your monitor reports
[06:41] <tritium> jdub, out of the available 13 packages, calc, draw, impress, math, and writer have .desktop files
[06:41] <Rotund> what does sudo xresprobe return on your monitor?
[06:41] <daniels> for me, this is 1600x1200 and everything below it, in nice little increments
[06:41] <daniels> right now?  absolutely nothing, since I'm on amd64
[06:41] <daniels> but on the i386, it returned a *lot* of resolutions, many of which were above 1024x768
[06:42] <daniels> i can assure you that there is no bug in xresprobe
[06:42] <daniels> it feeds through exactly what your monitor reports back
[06:42] <Rotund> daniels: interesting. because looking at the code, I can't see how that's possible
[06:42] <daniels> if you dislike the resolution reported, take it up with your monitor manufacturer
[06:42] <daniels> you can't see how it's possible?
[06:42] <Rotund> daniels: look at ddcprobe.c
[06:42] <daniels> i can't see how it wouldn't be, but there you go
[06:43] <daniels> Rotund: which bit?
[06:43] <jdub> tritium: rocking, thanks
[06:43] <tritium> yup
[06:44] <Rotund> daniels: line 179-209
[06:44] <daniels> Rotund: established timings are irrelevant
[06:44] <Rotund> oh?
[06:44] <daniels> Rotund: everyone uses custom timings or detailed timings, which are freeform in terms of the values you can put in there
[06:44] <daniels> yes
[06:44] <Rotund> ahh
[06:45] <Rotund> okay.  so that's what that last part does?
[06:46] <daniels> yes
[06:46] <daniels> the bits that spew out ctiming: and dtiming: lines
[06:46] <Rotund> that makes sense then
[06:47] <Rotund> where's the code for xresprobe?
[06:48] <lamont> Kamion: yesterday's ia64 livecd doesn't find the cdrom... :(
[06:48] <Rotund> ahh... script w/o a .sh neverming
[06:51] <fabbione> hey lamont
[06:51] <fabbione> mind to send me your ssh pub key?=
[06:51] <fabbione> so i can setup the access to the sparc buildd
[06:52] <lamont> fabbione: sent
[06:52] <Rotund> daniels: actually, ddcprobe gets the info right and ddcprobe does not
[06:52] <fabbione> lamont: thanks
[06:52] <Rotund> daniels: .sh does not
[06:52] <Rotund> daniels: would you like my output from ddcprobe
[06:52] <fabbione> lamont: i will send you an email back with how to access.
[06:52] <lamont> thanks
[06:52] <daniels> Rotund: yes, although this may merely be a feature
[06:53] <lamont> and remind me what I'm doing and when...
[06:53] <Rotund> =)
[06:53] <daniels> Rotund: if you look at ddcprobe.sh, it will throw away the top resolution since this is never the native resolution on crts
[06:53] <Rotund> oh
[06:53] <daniels> s/native/optimal/
[06:53] <daniels> if you have an lcd, it will take the top mode and run with it
[06:53] <Rotund> okay.  that's it then.  it's the top one as the 1280 ones aren't mentioned
[06:54] <Rotund> that works then.  thanks
[06:55] <daniels> np
[06:58] <fabbione> night lamont
[06:58] <daniels> lam	night dude
[06:59] <fabbione> lamont: you don't have ipv6, do you?
[07:00] <daniels> mako: http://mako.yukidoke.org/copyrighteous/images/muppet_hat_2-small.png
[07:01] <daniels> mako: dude, that is so emo.  you look like you're about to cry.
[07:03] <mako> jordi: dude
[07:04] <mako> daniels: is that emo?
[07:05] <daniels> mako: crying is very emo
[07:06] <tseng> you cant be emo wearing white
[07:06] <tseng> its a strict rule
[07:06] <mako> i was trying to look very serious
[07:07] <mako> so i realized i look a lot like the EPA guy from ghostbusters.. i'm looking for pix
[07:09] <tseng> hm youre right
[07:31] <Rotund> is there a language that the x log parser should be in?
[07:31] <Rotund> I'd assume something higher level would be better
[07:31] <daniels> Rotund: probably python, but dude, it's a bitch to pars
[07:31] <Rotund> I'm sure
[07:31] <Rotund> =)
[07:32] <daniels> the problem is that you can detect when you're getting xf86PrintEDID() output, but you need driver-specific smarts to work out which head that DDC info is for, e.g.
[07:32] <Rotund> BTW: NVIDIA prints WAY different output
[07:33] <daniels> way to go nvidia
[07:33] <Rotund> =)
[07:33] <daniels> most of the DDC stuff should be in one standard form (i.e. that from xf86PrintEDID())
[07:33] <Rotund> I'm a tester during the day.  all I do is parse
[07:33] <daniels> but, that being said, I can't say that I've seen an nvidia (i.e. driver nvidia, not driver nv) log file recently
[07:33] <Rotund> (**) NV(0):  Default mode "1152x768": 65.0 MHz, 44.2 kHz, 54.8 Hz
[07:33] <Rotund> (II) NV(0): Modeline "1152x768"   65.00  1152 1178 1314 1472  768 771 777 806 +hsync +vsync
[07:34] <daniels> er, that's not what you're looking for
[07:34] <Rotund> (**) NVIDIA(0):      Default mode "800x600": 49.5 MHz, 46.9 kHz, 75.0 Hz
[07:34] <Rotund> (**) NVIDIA(0):      Default mode "800x600": 50.0 MHz, 48.1 kHz, 72.2 Hz
[07:34] <Rotund> it's not?
[07:34] <daniels> nope
[07:34] <daniels> you're looking for the DDC output
[07:35] <daniels> grep for 'EDID Version'
[07:35] <Rotund> I think that's it
[07:35] <dilinger> i'm not sure whether it's the hardware or warty's kernel, but this machine keeps freezing :/
[07:35] <Rotund> Validated modes for display device CRT-0:
[07:36] <Rotund> okay
[07:36] <daniels> Rotund: no, you do not want the validated modes
[07:36] <daniels> you want the DDC output
[07:36] <Rotund> what's a validated mode?
[07:36] <Rotund> there are a bunch of resolutions in there that are not described in my config file
[07:39] <daniels> this is well offtopic for #ubuntu-devel at the moment.  but basically all you need to know is that you're after what the monitor says is can do, which is in the DDC packet.  to find this, grep for 'EDID Version' and look around there'.  the mode stuff is useless as it's every mode X knows about, constrained by your configuration.  so if it's not set up right, it's useless to you.  and if you're probing, it isn't set up right by definition.
[07:40] <zenwhen> daniels, is gnome-volume-applet currently iconless for you?
[07:40] <Rotund> Really?  That's odd as it actually mentions a lot of things not defined in my configuration (resolutions and refresh rates)
[07:41] <daniels> zenwhen: works fine for me
[07:41] <daniels> Rotund: 'every mode X knows about, constrained by ...'; please read what I'm saying
[07:41] <zenwhen> odd, It has no icon for me and I have to hunt for it. Its clickable area is an invisible line.
[07:41] <zenwhen> lol
[07:41] <zenwhen> newest hoary packages
[07:43] <dholbach> good morning
[07:49] <tseng> hey there, dholbach 
[07:52] <dholbach> hai tseng
[08:14] <dholbach> those guys at dumbledore.hbd.com want to "own" the channel ;-)
[08:16] <morgs> dholbach: Eh, we're all sitting at the other end of a long thin pipe, and we want to see our updates before we all try to grab the debs at the same time!
[08:18] <dholbach> morgs: you then should get yourself some apt-proxy :-)
[08:26] <Rotund> interesting.  the nv driver supports running from either of the two outputs on my vid card, but can't do both at the same time!
[08:27] <daniels> that's common
[08:27] <Rotund> really?
[08:27] <Rotund> weirf
[08:27] <Rotund> BTW: could you answer my last question for the night from the pm
[08:34] <magnon> woah, someone just yelled "APPROVED" in my face :)
[08:34] <magnon> that one was a bit over the top
[08:35] <daniels> Rotund: what was that?
[08:37] <tseng> http://higgs.djpig.de/ubuntu/www/ < hmm
[08:37] <tseng> rock.
[08:38] <dholbach> tseng: cool
[08:38] <jdub> heh
[08:38] <jdub> rad ;)
[08:38] <jdub> elmo: http://higgs.djpig.de/ubuntu/www/
[08:39] <jdub> Mithrandir: did you end up doing guifications? what were the cool features in that?
[08:41] <Mithrandir> jdub: yes, but it's still stuck in Debian's NEW.  I can just upload it to ubuntu; I assume elmo's not too happy about doing debian NEW processing to get stuff into ubuntu?
[08:41] <drbyte> jdub: kernel maintainership on ubuntu/ppc, you still don't have anyone ?
[08:43] <fabbione> drbyte: we still don't have an official responsable for the port
[08:44] <drbyte> fabbione: hmm, okay. if you're still stuck say in March (when back in Melb, and my ppc hoard), i'd gladly help out. i do fedora/ppc (mainly still ppc32), and am aiming for them to run on pegasosppc hardware as well
[08:44] <fabbione> drbyte: that's nice. thanks
[08:45] <fabbione> i like the idea of multi distro kernel team
[08:45] <fabbione> and that's what is coming up now
[08:45] <drbyte> fabbione: though we're pretty anal about making sure we stick to upstream kernels (very few ppc specific patches)
[08:45] <drbyte> fabbione: are ya'll comfortable with adding the new iBook/powerbook g4 sleep patches, and so on? that would rock, but we can't mainline it (policy-wise)
[08:46] <jdub> Mithrandir: heh, don't think so, no ;)
[08:46] <fabbione> drbyte: it depends. we discuss patches in the team
[08:47] <pitti> Morning
[08:47] <fabbione> clearly upstream bugfixes are always welcome
[08:47] <drbyte> fabbione: ah, okay. so i'm guessing i should joing ubuntu-devel-list soon. yal
[08:47] <fabbione> for external patches/features we are a bit more conservative
[08:47] <fabbione> in terms that we include stuff that has an active upstream
[08:47] <fabbione> but kill dead stuff
[08:47] <fabbione> if upstream dies
[08:48] <drbyte> ok.
[08:48] <drbyte> fabbione: i'm no DD, i make rpm's well, but is the preferred debian/ubuntu method make-kpkg ?
[08:49] <fabbione> drbyte: kinda.. yed
[08:49] <fabbione> yes
[08:49] <fabbione> it's a debian package, but there is a "packager" in the team
[08:49] <fabbione> so you don't really need to worry too much about it
[08:49] <fabbione> in terms that you can help the ppc side just providing patches or link to the Changesets
[08:49] <fabbione> or stuff like that
[08:50] <drbyte> ok. sounds good. that means i might not even need ubuntu on ppc yet :P
[08:50] <fabbione> drbyte: well clearly testing the kernel is kinda mandatory
[08:50] <drbyte> fabbione: i know.. i was just joking ;-)
[08:51] <fabbione> specially if you want to take responsability towards the community :)
[08:51] <drbyte> fabbione: so discussion on the list, i presume?
[08:52] <fabbione> drbyte: yes. we are using ubuntu-devel as a starting point with [kernel]  prefix in the subject
[08:52] <fabbione> if there will be enough traffic, we will ask for a mailing list
[08:52] <fabbione> but most of the work is done on irc (here)
[08:52] <fabbione> and stuff will be soon on the wiki
[08:52] <fabbione> (after yesterday's kernel team meeting)
[08:53] <drbyte> ok. i'll suck hoary down soon(ish) for ppc
[08:53] <fabbione> cool
[09:07] <dholbach> jdub: could you give a brief etymological talk on the word "rad"? i read it everywhere nowadays and since i'm a non-native speaker, you could surely enlighten me a bit :-)
[09:08] <jdub> dholbach: it's short for radical, which is in the dictionary ;)
[09:08] <dholbach> jdub: ah... yes - should be :-)
[09:08] <jdub> dholbach: stupid english speakers ;)
[09:08] <dholbach> jdub: stupid non-native english speakers - i had no imagination 
[09:09] <tseng> dholbach: have coaster debs online somewhere?
[09:10] <dholbach> tseng: i'm working on them, bryan just released 0.1.4.2 and i'm still sorting the auto*-doing-{post,pre}{rm,inst}-stuff out
[09:10] <tseng> ok.
[09:10] <tseng> whatever the issue is, ill blame him
[09:35] <dholbach> hai mvo_
[09:36] <mvo_> hi dholbach 
[09:36] <mvo_> morning all
[09:38] <jordi> mako: dude?
[09:42] <jdub> mvo_: yo
[09:43] <jdub> mvo_: do synaptic and apt-get share their pinning/on-hold status?
[09:45] <haggai> mdz: ping
[09:46] <bob2> amusingly, aptitude doesn't use the same pinning system as dpkg/apt
[09:46] <Treenaks> bob2: sounds logical!
[09:46] <mvo_> bob2: it should honor /etc/apt/preferences?
[09:46] <bob2> bah
[09:46] <bob2> hold, not pinning
[09:46] <bob2> it's too early or late or something for me
[09:46] <mvo_> :)
[09:47] <Treenaks> doesn't apt-get honor the dpkg selections stuff?
[09:47] <bob2> selections, somewhat, holds, yes
[09:51] <thom> mdz: a pretty sweet pygtk app that keeps track of work done and so on
[09:56] <mvo_> jdub: synaptic honors the apt pins but uses it's own config file to implement "locks" on versions
[10:02] <Mithrandir> pitti: glibc change is ok with mdz on the condition that it makes it in before feature freeze.
[10:02] <Mithrandir> pitti: can you do that?
[10:02] <pitti> Mithrandir: sure, I can upload it
[10:02] <Mithrandir> great, thanks.
[10:02] <pitti> Mithrandir: i. e. it should be done today
[10:02] <fabbione> not another libc6 update!
[10:02] <Mithrandir> pitti: yes
[10:02] <fabbione> :)
[10:03] <Mithrandir> fabbione: MULTIARCH update. :)
[10:03] <fabbione> why am i afraid that it will break sparc?
[10:03] <pitti> fabbione: please look at the patch, it looks safe
[10:03] <pitti> fabbione: http://arch.err.no/index.cgi/tfheen@idi.ntnu.no--2005/pkg-glibc--multiarch--0--patch-2?cmd=cs_new&file=debian/patches/99_multiarch-ld.dpatch
[10:04] <fabbione> looks ok....
[10:05] <Mithrandir> fabbione: it should be harmless, but I could test-build it on sparc
[10:05] <fabbione> Mithrandir: nah.. let the buildd munge it
[10:05] <fabbione> i am checking one thing only..
[10:06] <fabbione> gcc -dumpmachine
[10:06] <fabbione> sparc-linux
[10:06] <fabbione> is that what you expect?
[10:06] <Mithrandir> yeah; it should be sparc64-linux or something if you have a sparc64 userland.
[10:06] <Mithrandir> but sparc64 uses a different loader, so that should be fine, I imagine.
[10:06] <fabbione> sparc64 gcc -dumpmachine
[10:06] <fabbione> sparc-linux
[10:06] <Mithrandir> does gcc -m64 -dumpmachine say the same?
[10:07] <Mithrandir> well, shouldn't be a problem anyhow
[10:07] <fabbione> yes
[10:07] <fabbione> but i have no dirs like /usr/lib/sparc-linux/
[10:07] <Mithrandir> that's ok, you'll have them later.
[10:07] <fabbione> ok
[10:07] <Mithrandir> it just adds them to the path, it doesn't remove anything.
[10:08] <Mithrandir> /usr/lib can't be removed from path in a _long_ time.
[10:08] <Mithrandir> uhm, ld-linux' path, that is
[10:09] <fabbione> why on heart ppc is always a porting bitch
[10:09] <fabbione> 2.6.11 compiles everywhere other than ppc
[10:09] <daniels> mono is the most hateful evil shit ever
[10:10] <Treenaks> daniels: you've obviously never tried a jvm
[10:10] <fabbione> none of you have ever tried a new kernel release on 6 arches
[10:11] <tseng> novell sure makes it hard to find their srpms
[10:12] <fabbione> [RFC]  Changing COW detection to be memory hotplug friendly
[10:12] <Treenaks> Moo
[10:12] <fabbione> next will be: Changing PIGS and BULLS detection to be wifi friendly
[10:13] <Treenaks> I want CHICK detection!
[10:13] <Simira> :-)
[10:14] <bob2> fabbione: you should apply the sleep patch again!
[10:15] <daniels> VICTORI!!! TAKE THAT, MONO
[10:15] <daniels> also, VICTORY
[10:16] <daniels> mono has corrupted my ability to spell
[10:16] <bob2> From: VICTORY ...
[10:16] <daniels> bob2: heh
[10:16] <thom> daniels: what are you doing to mono?
[10:18] <Mithrandir> thom: torturing it
[10:19] <fabbione> bob2: there are no updates from benh and today is Feature Freeze. kthxbye
[10:19] <bob2> dammit
[10:19] <daniels> thom: dbus-mono
[10:19] <bob2> no MAD PHAT LAPTOP SUPPORT for ppc then
[10:20] <thom> daniels: ahr
[10:20] <thom> bob2: wha?
[10:20] <fabbione> most his stuff is merged upstream
[10:20] <fabbione> but not all of it
[10:20] <bob2> thom: well, for modern ppc laptops
[10:20] <bob2> ones with radeon 9200 need special love, aiui
[10:20] <daniels> bob2: sure there is -- a dialog box saying 'should've bought an x40'
[10:20] <bob2> rofl
[10:21] <thom> daniels: so not fixoring mono for amd64 then
[10:22] <jdub> mono 1.1 should be ok
[10:22] <daniels> thom: no, cool as it would be
[10:22] <daniels> jdub: doable, with an uvf break? :)
[10:22] <daniels> as long as it doesn't spaz out and decide to change /usr/lib/mono *again*
[10:22] <thom> jdub: yes, but it needs total repackaging!
[10:22] <jdub> daniels: universe
[10:22] <seb128> is somebody working to fix muine ? 
[10:22] <jdub> thom: ouch
[10:22] <jdub> thom: doable for hoary though
[10:22] <thom> jdub: one source package versus about 3000
[10:22] <daniels> jdub: i assume that means motu+own time, then?
[10:22] <thom> that's why i didn't do it
[10:23] <jdub> daniels: yeah
[10:23] <daniels> jdub: ah, crap
[10:23] <daniels> thom volunteers
[10:23] <martink> jdub: mono 1.1 in universe would rock. No more manual mono compiles to use tomboy on amd64
[10:23] <jdub> yeah, and beagle on all arches, etc.
[10:23] <fabbione> jdub: is 1.1 buildable on sparc too?
[10:23] <daniels> jdub: fwiw, the invocation required is /libdir /usr/lib /gacdir /usr/share/dotnet
[10:24] <jdub> fabbione: i believe so
[10:24] <daniels> afaict
[10:24] <daniels> (as in, dbus-mono makes beagle work, and I don't care beyond there)
[10:24] <jdub> oh man
[10:25] <jdub_> pants off
[10:25] <Treenaks> uh?
[10:25] <jdub> mr 6dub is very laggy
[10:25] <bob2> 4in6?
[10:26] <jdub> 6to4, sipper
[10:26] <jdub_> so i can say stuff
[10:26] <jdub_> but not see stuff
[10:26] <jdub_> or change nick
[10:26] <bob2> 6inyourPANTS.
[10:27] <jdub> mvo_: aha
[10:27] <jdub> mvo_: dude
[10:27] <jdub> mvo_: installation from web browser
[10:27] <fabbione> goody.. amd64, ppc, i386, ia64 are GO
[10:28] <jdub> mvo_: i'm using gnome-app-installer infrastructure to do the web side of that, was going to seed your brain with it ;)
[10:29] <mvo_> jdub: hehe :) you are subscribed to the packagemanagment wiki page I'm just editing :) ?
[10:30] <jdub> mvo_: i'm subscribed to all ;)
[10:30] <jdub> mvo_: so i can do a spiffy webpage for you
[10:30] <mvo_> jdub: that would be cool
[10:31] <jdub> fun proof of concept, anyway
[10:31] <EvanCarroll> i just upgraded the style sheet for my website, anyone want to give input
[10:31] <dholbach> seb128: is http://bugzilla.ximian.com/show_bug.cgi?id=70976 what we would need to get evolution back on amd64?
[10:31] <mvo_> yeah and cheap to implement
[10:31] <EvanCarroll> www.evancarroll.com
[10:32] <bob2> EvanCarroll: this is a technical channel, dude
[10:32] <seb128> dholbach: no, we already have this fix
[10:32] <dholbach> seb128: oh ok
[10:35] <pitti> elmo: xemacs21 sync, please
[10:35] <pitti> elmo: oops, wait. Actually I just need 21.4.16-2
[10:35] <pitti> elmo: but sid already has 21.4.17
[10:36] <pitti> elmo: although 21.4.17 does not contain much more (only other fixes), and it is universe
[10:37] <pitti> elmo: thus if you don't want 21.4.17-1, please sync 21.4.16-2 (it's still in the archive)
[10:39] <thom> pitti/jdub: can i break UVF for apache2, please; two security bugs and a lot of changes from debian we want
[10:39] <pitti> thom: microrelease?
[10:40] <thom> 2.0.52->2.0.53
[10:40] <pitti> thom: debian changes = bugfixes?
[10:40] <thom> (there are a bunch of segfaults and two CANs fixed)
[10:41] <thom> and yes, debian changes are stuff i'd be taking anyway
[10:41] <pitti> sounds sane to me, however, I don't have the power to have the final word :-)
[10:41] <thom> ber.
[10:41] <thom> mdz/jdub: ping!
[10:42] <daniels> jdub: dbus-mono 0.23-3, have at it
[10:43] <jdub> thom: approve
[10:43] <Mithrandir> seb128: jdub thinks we should have gaim with a non-palindromic version number in hoary, what do you think?
[10:43] <daniels> jdub: OH MY GOD YOU ARE WORSE THAN THE X SERVER
[10:43] <daniels> or something
[10:43] <thom> jdub: do i need both you and mdz to approve?
[10:44] <seb128> Mithrandir: go for it :)
[10:45] <jdub> thom: no
[10:45] <thom> score
[10:47] <Mithrandir> when writing daemon code, please don't close std{in,out,err} without reopening them to /dev/null, kthxbye.
[10:47] <daniels> heh
[10:48] <jdub> oh man
[10:48] <jdub> STOP DELETING FEATURES ON FEATURE FREEZE DAY
[10:49] <fabbione> does anybody have a link to "infiniband" technology?
[10:49] <Mithrandir> jdub: hmm?
[10:49] <jdub> Mithrandir: i just deleted a feature.
[10:50] <fabbione> oh here it is
[10:50] <daniels> jdub: 6.8.1-1ubuntu16 to be uploaded today with some new features; 6.8.2-1 will come later, breaking uvf (not that it matters, since there's less code churn between .1-1ubuntu16, which has 6.8.x branch, and .2-1 than in most releases)
[10:50] <daniels> s/releases/ubuntu revisions/
[10:51] <jdub> man
[10:51] <jdub> i am such a tool
[10:51] <daniels> yes
[10:51] <daniels> but why, specifically?
[10:51] <jdub> i deleted a feature.
[10:51] <daniels> which feature?
[10:52] <jdub> the desktop file sucker which will seed gnome-app-install
[10:53] <fabbione> http://www.infinibandta.org/ibta/
[10:54] <fabbione> up to 100Gb/sec data transfer
[10:56] <thom> INFINIBAND: ADRENALINE FOR DATA CENTERS
[10:56] <rubenv> mjg59: or some X/dpms person: ping
[10:56] <rubenv> my laptop screen does funny stuff I don't trust
[10:56] <daniels> rubenv: oh?
[10:57] <rubenv> when i do xset dpms force <something>
[10:57] <rubenv> it doesn't blank & stuff
[10:57] <sivang> morning all!
[10:57] <rubenv> it shows lines & lights up
[10:57] <rubenv> i don't dare to wait what'll happen in the end
[10:57] <rubenv> it goes very very scary bright :-)
[10:58] <daniels> rubenv: ooo.  what sort of video chipset?
[10:58] <rubenv> nvidia, with nv driver now
[10:58] <daniels> 'cool'.
[10:59] <daniels> unfortunately, that's the most productive answer I can give
[10:59] <daniels> that driver is a black box :\ no-one other than nvidia can do anything with it
[10:59] <rubenv> no, not the binary nvidia
[10:59] <rubenv> the open source nv
[10:59] <rubenv> the binary one does it right
[11:00] <daniels> yes, even the 'open source' nv, is not open source at all
[11:00] <rubenv> but the open source one allows suspend to disk
[11:00] <Treenaks> daniels: scary
[11:00] <rubenv> perhaps i should go back to binary evil (with no suspend)
[11:00] <daniels> for instance, fixing a bug on radeon the other day, I changed added save->fp2_gen_cntl = save->fp2_gen_cntl | RADEON_FP2_ON & ~(RADEON_FP2_BLANK_EN);
[11:01] <daniels> the nv equivalent would ne pNv[0x1234]  = 0x5678;
[11:01] <daniels> where no-one but nvidia has any idea what those numbers mean
[11:01] <rubenv> aha, right :-)
[11:01] <rubenv> unknown specs
[11:01] <daniels> yeah.  the only people to ever change the nv driver have been nv, really.
[11:01] <rubenv> damn you nvidia :-(
[11:02] <Treenaks> daniels: so it's a good thing the laptop I want is Ati 9700 then?
[11:02] <Treenaks> is/has
[11:02] <rubenv> so no power management for me then :-/
[11:02] <daniels> Treenaks: well, you won't get 3d without fglrx, which sucks horribly
[11:02] <daniels> rubenv: you would appear to lose, yah
[11:02] <Treenaks> daniels: yeah, but there aren't many other laptops which are almost fully supported AND have 1680x1050 resolution
[11:04] <daniels> Treenaks: for 2d it'll be fine
[11:04] <daniels> and you'll get power management and all that goodness
[11:04] <daniels> (and I have the specs and full 2D DDK, so when stuff goes wrong, I can fix it, or at least have a go)
[11:05] <Treenaks> daniels: the reason for not opening up is all patent license stuff, right?
[11:07] <daniels> Treenaks: ish.  part of it's cross-licenced technology, part of it's just not wanting to let nvidia or whatever get at their chips (raising the barrier to entry to reverse-engineering everything they have)
[11:07] <daniels> part of it may also be the soundblaster thing -- just not wanting people to make cheap knockoffs of their chips that work with their drivers
[11:07] <Treenaks> argh
[11:08] <sivang> rehi all
[11:08] <rubenv> i'd rather not have my screen burned while i'm showering
[11:09] <daniels> Treenaks: either way, it sucks, yeah
[11:09] <Treenaks> daniels: yeah.. and sticking with Intel/Via/S3 sucks too
[11:10] <daniels> Treenaks: ironically, the most advanced open-source 3D driver is i915
[11:11] <Treenaks> daniels: VIA sort-of works on sid + unofficial "r30" unichrome patches
[11:11] <Treenaks> uh patches=patched packages
[11:11] <daniels> Treenaks: 6.8.1-1ubuntu16 will have unichrome r30
[11:11] <daniels> we currently have r29 or something
[11:11] <thom> holy crap, novell just hired the xgl dude?
[11:12] <daniels> thom: davidr?
[11:12] <Treenaks> daniels: problem is DRM support in the kernel.. which is too insecure to be included by default (shared-mem video or something)
[11:12] <daniels> woah yeah, they hired davidr.  jesus.
[11:12] <daniels> Treenaks: yeah.  something about dri access being equivalent to a free run over /dev/mem.
[11:13] <daniels> woah, they're hiring him for xgl.  sweet!
[11:13] <thom> yeah
[11:13] <thom> that's fully awesome
[11:14] <daniels> now someone needs to hire anholt, ajax, idr, and brianp specifically for mesa-solo
[11:14] <daniels> oh, and probably jonsmirl also
[11:18] <Mithrandir> Kamion: happy with the answer from the sun guy?
[11:19] <Kamion> Mithrandir: yep, fantastic, thanks
[11:19] <Mithrandir> Kamion: are you sending a "thanks a lot" letter or should?
[11:19] <Treenaks> Kamion: what are they about then?
[11:19] <fabbione> Mithrandir: did you have any chance to install ubuntu on sparc?
[11:19] <Mithrandir> fabbione: just dist-upgraded, I haven't gotten around to doing a full install.
[11:19] <Kamion> Mithrandir: I will do
[11:19] <fabbione> ah ok
[11:19] <Mithrandir> fabbione: I can look at it tomorrow if you remind me.
[11:19] <fabbione> sure i think i can
[11:20] <Kamion> Treenaks: usually either X or how the installed system works in general
[11:20] <fabbione> tomorrow i will not be around 100%
[11:20] <Treenaks> Kamion: ah, so you can just blame daniels and be happy :)
[11:20] <thom> Treenaks: i think he does that anyway
[11:20] <fabbione>   NUMA support (ACPI_NUMA) [N/y]  (NEW) 
[11:20] <fabbione> hmmmm
[11:21] <Mithrandir> fabbione: just prod me about it, but I'm busy today, so.
[11:22] <fabbione> no problem
[11:25] <daniels> Kamion: btw, since you clearly don't have enough work, want to take care of #6296 and #5894? :)
[11:31] <Kamion> daniels: I have approximately zero idea how keymaps get from d-i to X
[11:31] <Kamion> I assumed you were guessing them from console keymaps, I think
[11:32] <daniels> Kamion: afaict we're guessing it from $LANG?
[11:32] <Kamion> !
[11:32] <Kamion> that's fucked up
[11:32] <Kamion> and so, so wrong
[11:33] <daniels> # Warty change: try to guess keyboard layout from $LANG
[11:33] <daniels> case "$LANG" in
[11:33] <daniels>   "bs_BA.ISO8859-2" ) LAYOUT="us,bs" XKBOPTIONS="grp:alt_shift_toggle" ;;
[11:33] <daniels> ...
[11:33] <Kamion> maybe we should just use the work Konstantinos has done in Debian on this, and be done with it
[11:33] <daniels> although that doesn't explain how the dvorak damage gets there
[11:33] <pitti> daniels: that means that currently an English-speaking guy in front of a Russian keyboard would get an English layout in X?
[11:33] <Kamion> ok, that needs to die painfully
[11:33] <Kamion> d-i has a separate keyboard chooser for a reason
[11:33] <daniels> pitti: seemingly
[11:33] <pitti> d'oh
[11:34] <daniels> maybe there's something I'm missing though, otherwise no-one would get a 'dvorak' layout?
[11:34] <Kamion> I don't see how anyone does
[11:34] <Mithrandir> magggic?
[11:34] <Mithrandir> or nobody but mdz uses dvorak?
[11:35] <daniels> Kamion: me neither
[11:35] <daniels> Kamion: if you want to merge Konstantinos's stuff, that would be awesome :)
[11:35] <fabbione> thom: how can i avoid a specific client to access pages on my web server?
[11:35] <Kamion> I think I might do
[11:35] <daniels> Kamion: phat
[11:35] <fabbione> there is a new search engine that doesn't respect robot.txt
[11:36] <seb128> jdub: around ?
[11:36] <Kamion> mdz,jdub: how would you feel about pulling localization-config into main?
[11:36] <jdub> Kamion: i'll leave this one for matt
[11:36] <pitti> Mithrandir, daniels: I think ddaa uses dvorak, too
[11:36] <jdub> seb128: yeah
[11:37] <daniels> pitti: yeah, azerty dvorak.  bong.
[11:37] <Mithrandir> fabbione: deny from $ip in htaccess?
[11:37] <fabbione> Mithrandir: they use the same technics as google
[11:38] <fabbione> if you ban the domain, they switch to non resolvable addresses and if you ban the address they switch to another net
[11:38] <fabbione> FUCKING ANNOYING
[11:38] <seb128> jdub: insight on #1080 appreciate :)
[11:39] <daniels> Kamion: https://bugzilla.ubuntu.com/show_bug.cgi?id=6296
[11:39] <thom> fabbione: ew. 
[11:40] <Mithrandir> fabbione: ban on user agent?
[11:41] <fabbione> Mithrandir: exactly
[11:41] <thom> fabbione: see query
[11:41] <fabbione> but i am more evil
[11:41] <Kamion> daniels: yeah, you mentioned that, seems to be the same class of bug
[11:41] <thom> it's mod_rewrite, sadly
[11:41] <fabbione> Redirect on user Agent
[11:41] <fabbione> thom: yes i am reading
[11:41] <thom> redirect? where?
[11:41] <fabbione> goatse.cx
[11:41] <Kamion> daniels: if we don't want to pull in localization-config (which might make sense, it does quite a few things), then we should at least pull the logic from localization-config into xserver-xorg
[11:41] <fabbione> let's them cache some good old shit
[11:42] <thom> fabbione: they won't. you're just wasting someone else's bandwidth, and that's kinda rude
[11:42] <fabbione> not really...
[11:43] <daniels> thom: to be fair, i hardly doubt goatse.cx was set up for anything other than being put in slashdot comments
[11:43] <daniels> kami	sure
[11:43] <thom> true
[11:43] <fabbione> daniels: point
[11:43] <Mithrandir> fabbione: look at http://www.neilgunton.com/spambot_trap/
[11:44] <Kamion> daniels: if we go that way, I'll put together some sample code for you
[11:45] <fabbione> thom: can i put that rule in the generic httpd.conf or does it need to be x vhost?
[11:45] <daniels> Kamion: dude, if you do, I'll buy you a night out down here on some *real* beer
[11:45] <daniels> fabbione: rewrite is per-vhost
[11:45] <daniels> fabbione: don't forget RewriteEngine On
[11:46] <daniels> Kamion: heh, I'd say a case of Coopers, but that might be tricky on the plane.  but that offer's still valid if you like. :)
[11:47] <fabbione> daniels: still x vhost, right?
[11:47] <daniels> fabbione: yeah
[11:47] <fabbione> ok
[11:47] <jdub> seb128: hrm, what do you want me to say? :)
[11:48] <haggai> jdub: was it you I was talking to at the con about dbus-qt bindings?
[11:48] <jdub> haggai: possibly
[11:48] <seb128> jdub: is there any way to get bug-buddy speaking to bugzilla ?
[11:48] <jdub> seb128: i thought we were going to use the sendmail gateway
[11:48] <seb128> jdub: I don't know what this gateway is ...
[11:48] <daniels> haggai: if you test them in the next two hours, I'll upload before FF ends
[11:48] <daniels> haggai: http://people.ubuntu.com/~daniels/dbus/
[11:49] <seb128> jdub: where is the gateway ? how does it work ?
[11:49] <daniels> haggai: asked riddell and amu to test a few days ago, but no response
[11:49] <haggai> daniels: ah cool
[11:49] <jdub> seb128: no idea :)
[11:49] <daniels> haggai: tick tick :)
[11:49] <seb128> jdub: ok, thanks anyway :)
[12:04] <haggai> daniels: seems there is some problem with dependencies and we're not gonna manage to fix it all so don't worry
[12:09] <daniels> haggai: what's the dependency problem?
[12:10] <pitti> Mithrandir: just uploaded new glibc with your patch
[12:10] <Mithrandir> pitti: thanks.
[12:10] <Treenaks> eek @ ubuntu-user.. someone who tries "xhost +"
[12:11] <daniels> whoops
[12:12] <Kamion> mvo_: it would be really nice if new apt configure options came with documentation :)
[12:13] <Mithrandir> yay.  I just repaid my mortage.
[12:13] <daniels> Mithrandir: word!
[12:14] <mvo_> Kamion: good point :)
[12:15] <Kamion> (looks like -o Acquire::gpgv::Options::=--ignore-time-conflict works)
[12:15] <mvo_> at least that's good news :)
[12:16] <elmo> why are you ignoring time conflict, JOII
[12:16] <elmo> or JOOI even
[12:17] <Kamion> elmo: 'cos sometimes the clock is broken in the installer and we haven't got to the point of running ntpdate yet
[12:17] <elmo> ah
[12:17] <Kamion> elmo: daniels reported that he couldn't finish the installation because his clock was set to 2003
[12:17] <jdub> mine was set to 1985!!
[12:18] <jdub> o/~ and that's the power of love o/~
[12:18] <elmo> daniels in "two year behind the times" shocker... ;P
[12:18] <Kamion> jdub: your clock is *always* set to 1985
[12:19] <daniels> elmo: says the man from leeds
[12:27] <haggai> daniels: dbus-qt-1-dev depends on libqt3c102-mt-dev but it should be libqt3-mt-dev 
[12:28] <daniels> haggai: i can fix that in an instant.
[12:28] <daniels> haggai: if you --force-depends it, is the rest ok?
[12:31] <haggai> daniels: looks like it, yes
[12:32] <daniels> phat
[12:32] <daniels> thanks
[12:53] <sivang> hey jbailey 
[12:55] <jbailey> Heya sivang 
[12:55] <sivang> jbailey: how you been? :)
[12:55] <jbailey> sivang: I've been sleeping. =)
[12:55] <sivang> jbailey: ah well, that's ever good :)
[12:58] <thom> jbailey: nice initrd-tools upload dude!
[01:00] <thom> no, that was serious
[01:00] <jbailey> Cool. =)
[01:01] <jbailey> I need to spend some time over the next few weeks making sure that I have LVM and EVMS setups here to test on, so I'm always nervous uploading that my inbox will fill up with "You broke my system" emails =)
[01:01] <thom> you've not killed by evms
[01:01] <daniels> works with lvm on amd64, at any rate
[01:01] <thom> uh, my
[01:01] <daniels> thom: is lvm or evms recommended these days?
[01:02] <Kamion> I need to have a RAID/LVM install test system that isn't mission-critical; maybe I can do a small LVM on the amd64
[01:02] <ajmitch_> jbailey: I'll complain loud & long to you if you break my box, don't worry :)
[01:02] <thom> evms allows you to use lvm as part of it, it's pretty cool
[01:02] <Treenaks> thom: so evms is recommended?
[01:02] <daniels> thom: ahr.  so if I have an lvm /home, then I can get all the EVMS goodness without reformatting?
[01:02] <thom> daniels: afaik
[01:02] <thom> Treenaks: i guess, i just use it on /home though so not an expert
[01:03] <daniels> thom: phat
[01:03] <ajmitch_> sounds good, I've got LVM on everything but / at the moment
[01:06] <Keybuk> #6293: Only English people should be allowed to build packages
[01:07] <daniels> Keybuk: what, no Welsh allowed?
[01:07] <Keybuk> apparently noy
[01:07] <jbailey> Keybuk: *that's* what I should've titled the bug. =)
[01:08] <thom> we should just declare en_GB to be the one true sort order
[01:08] <Keybuk> thom: NO!
[01:09] <thom> pfft, madness
[01:09] <jdub> Keybuk: had you seen this?
[01:09] <jdub> http://itkitchen.info/2004/10/27/planet-free-software/
[01:10] <eruin> wasabi: Couldn't stat source package list http://archive.ubuntu.com hoary/universe Packages (/var/lib/apt/lists/archive.ubuntu.com_ubuntu_dists_hoary_universe_binary-i386_Packages) - stat (2 No such file or directory)
[01:10] <eruin> err
[01:10] <eruin> that wasn't directed at you :-)
[01:11] <eruin> could anyone tell me if that's a problem on my part or archive.ubuntu.com ?
[01:11] <daniels> thom: posix sort order, mofo
[01:11] <daniels> thom: did you see that particular frightening quirk of the iaudio?
[01:11] <thom> daniels: yeah
[01:12] <Keybuk> why's that frightening?
[01:13] <sivang> jdub: wow nic, dind't know you are one on the long history of blog development :) and scott also :)
[01:15] <Keybuk> jdub: heh, fame
[01:15] <jdub> Keybuk: authored by spiv
[01:15] <jdub> er
[01:15] <jdub> Keybuk: authored by spiv's other half
[01:15] <daniels> Keybuk: don't get me wrong, I love it, partially because it's how ~/music on my desktop sorts also
[01:15] <daniels> Keybuk: but on a piece of consumer hardware?  frightening.
[01:17] <Keybuk> daniels: I would've have probably freaked if it didn't do that
[01:17] <Keybuk> to be honest, I didn't notice
[01:19] <Keybuk> jdub: she has an allergy to 'w's
[01:31] <ogra> jdub.... psst....hoary changes ... boa constructor.....
[01:31] <ajmitch_> what did I break?
[01:31] <ogra> nothing.... i hope ;)
[01:32] <ajmitch_> well it still runs here with py 2.4 :)
[01:32] <ogra> just wanted to point out you first timer ;)
[01:32] <ajmitch_> aha
[01:33] <ogra> applause from germany.....
[01:36] <jdub> ogra: xss went in, btw.
[01:36] <jdub> ogra: (yeah, seen ;)
[01:37] <ogra> jdub: i saw it....,ade me really happy :-D
[01:37] <pitti> Hi ogra 
[01:37] <jdub> ogra: already had a few people comment on it
[01:37] <ogra> jdub now lets see if my hal patches get accepted by pitti
[01:37] <pitti> ogra: I just wanted to remember you about the feature freeze today :-)
[01:37] <pitti> ogra: can you send me a debdiff?
[01:37] <pitti> or put it somewhere?
[01:38] <ogra> pitti...just sending the first patch (lsb_release)
[01:39] <ajmitch_> jdub: what's the status of bug management for universe for now?
[01:40] <ogra> pitti, btw, what did you change between ubuntu1 and 3 ? it takes nearly 1min to get the first response from lshal after upgrade.....
[01:40] <ogra> (without my patches)
[01:40] <jdub> Kamion: nice reply
[01:41] <jdub> ajmitch_: "not" :|
[01:41] <pitti> ogra: I fixed the nasty "udev goes crazy" bug
[01:41] <Kamion> jdub: I had to bite my tongue in a couple of places
[01:41] <pitti> ogra: however, it should actually become faster with this fix
[01:42] <ogra> pitti: nope.... the first execution of lshal gives me 0 devices... after about 30sec-1min i see them appear then....
[01:42] <jordi> I wonder why the "About Ubuntu" menu item in the panel doesn't haev a Ubuntu logo.
[01:42] <pitti> that's odd
[01:43] <ogra> bearable...but not nice....
[01:43] <pitti> ajmitch_: congrats for your first upload, and welcome aboard :-)
[01:44] <doko> fabbione: just a toy for jbailey
[01:45] <pitti> ogra: works fine for me, after about 5 seconds I have all devices
[01:45] <ogra> hmm...probably my setup....
[01:45] <ogra> or my arch....
[01:46] <ajmitch_> pitti: thanks 
[01:47] <fabbione> Kamion: to which post?
[01:47] <fabbione> ah i found it
[01:50] <Keybuk> seb128: latest evo still broken wrt. secure smtp
[01:50] <Keybuk> and they've seriously broken quote-message-in-reply
[01:50] <jdub> sounds like GTK BOOG to me!
[01:50] <daniels> jdub: no, it's a feature
[01:51] <daniels> so seb has a few hours to fix it
[01:51] <seb128> hate hate hate hate evolution
[01:51] <seb128> we should ship an another MUA
[01:51] <Treenaks> seb128: like mutt?
[01:51] <seb128> every single release broken
[01:51] <jdub> where can i download the ubuntu uploader keyring?
[01:51] <dholbach> re
[01:51] <seb128> Treenaks: ...
[01:51] <sivang> hey dholbach 
[01:51] <seb128> Treenaks: you did that troll already yesterday
[01:51] <dholbach> hi sivan! :-)
[01:51] <sivang> dholbach: 'sup?
[01:51] <Treenaks> seb128: no, that was another one
[01:52] <jdub> aha
[01:52] <dholbach> sivang: just did the laundry :-)
[01:52] <dholbach> sivang: how are you?
[01:52] <seb128> Treenaks: seriously, that was the same, and that's not really a chan to troll
[01:53] <sivang> dholbach: fine, intend to fix 2 bugs today hopefully before FF is in effect :)
[01:53] <Treenaks> seb128: I know
[01:53] <dholbach> sivang: cool - which ones?
[01:54] <mvo_> jdub: I should eventually be in ubuntu-keyring, but it's not (yet?)
[01:54] <mvo_> s/I/it/ == the uploader keyring
[01:55] <jdub> mvo_: looks like you are
[01:55] <jdub> how else are you uploading? :)
[01:55] <sivang> dholbach: 1849, and the one I already know by heart - 6092 :)
[01:55] <mvo_> jdub: I mean, the uploader keyring should be available in the package ubuntu-keyring :) 
[01:55] <jdub> oh
[01:55] <mvo_> but it's not right now
[01:55] <dholbach> sivang: nice :-)
[01:55] <mvo_> ah :)
[01:56] <jdub> ah, so i am not so stupid!
[01:56] <jdub> ah ha ha!
[02:01] <mvo_> jdub: this was more a reminder to myself that I should add it to the ubuntu-keyring package :)
[02:02] <daniels> jdub: scorchio!
[02:02] <elmo> it's on the keyserverss too
[02:03] <mjg59> rubenv: Yo
[02:04] <jdub> Mithrandir: ping
[02:04] <rubenv> mjg59: pong
[02:05] <rubenv> mjg59: ever seen a laptop screen coming out of suspend lighting up abnormally?
[02:05] <mjg59> Yes, the POSTing will often result in the backlight being on but the video chip not doing anything useful
[02:06] <mjg59> Does it go back to normal eventually?
[02:06] <rubenv> even when I disable posting i think
[02:06] <rubenv> i don't wait to see what it does in the end :-)
[02:06] <rubenv> I need this laptop ;-)
[02:06] <rubenv> the nv driver also seems to do it if you do something normal dpms
[02:07] <jdub> ajmitch_: as if you wouldn't!
[02:07] <jdub> ajmitch_: what about UDU? :)
[02:08] <rubenv> mjg59: also, i've read somewhere on nvidia's site that they have power management support, or is this horribly outdated info?
[02:08] <sivang> Unpacking replacement login ...
[02:08] <sivang> Errors were encountered while processing:
[02:08] <sivang>  /var/cache/apt/archives/gnome-icon-theme_2.9.91-0ubuntu2_all.deb
[02:08] <sivang> E: Sub-process /usr/bin/dpkg returned an error code (1)
[02:08] <sivang> known probably?
[02:08] <bob2> scroll up, that's not the error message
[02:08] <ajmitch_> jdub: it'll be easier to get to that, once I find out details :)
[02:09] <ajmitch_> it's just whether I can afford 2 weeks in .au :)
[02:10] <sivang> bob2: err, lost the scrollback...he well, I can live with that until it's fixed..
[02:10] <ajmitch_> thankfully I've got a friend in canberra who I could stay with
[02:10] <sivang> apptemting apt-get install 0f
[02:11] <sivang> s/0f/-f/
[02:12] <mjg59> rubenv: The brightness is just because an uncontrolled TFT will turn white
[02:12] <mjg59> There's no way it can damage anything
[02:12] <thom> infinity: yes, it's known
[02:12] <thom> uh, s/infinity/sivang/
[02:12] <mjg59> But if the nv driver is doing it, then it's likely that your video BIOS is just broken
[02:12] <rubenv> mjg59: well crap
[02:12] <rubenv> nvidia doesn't do it
[02:12] <sivang> Unpacking replacement libdbus-cil ...
[02:12] <sivang> dpkg: warning - unable to delete old file `/usr/lib/mono/gac/dbus-sharp/0.23.0.0__9eef2692033670f5': Directory not empty
[02:13] <sivang> dpkg: warning - unable to delete old file `/usr/lib/mono/gac/dbus-sharp': Directory not empty
[02:13] <sivang> dpkg: warning - unable to delete old file `/usr/lib/mono/gac': Directory not empty
[02:13] <seb128> haggai: around ?
[02:14] <rubenv> mjg59: how the hell would my video BIOS break?
[02:14] <mjg59> rubenv: As in, it's been written wrongly
[02:15] <rubenv> mjg59: could very well be, it's dell software after all ;-)
[02:15] <Kamion> sivang: those are just warnings, not errors
[02:16] <mjg59> rubenv: What model is it?
[02:17] <rubenv> Inspiron 8600
[02:17] <sivang> Kamion: oh right :)
[02:20] <haggai> seb128: yup
[02:22] <seb128> haggai: for information OO.o is probably broken by the new eds (soname change in some libs)
[02:22] <pitti> ogra: why did you modify Makefile.in instead of Makefile.am?
[02:22] <pitti> ogra: you really wrote the additional rules yourself?
[02:22] <haggai> seb128: oh thanks, I'll make sure I test against the new one
[02:31] <ajmitch_> night all
[02:39] <Mithrandir> jdub: pong
[02:39] <elmo> daniels: what's this stupid libgl warning spam about anyways?
[02:40] <daniels> elmo: could you possibly be less specific?
[02:40] <elmo> libGL warning: 3D driver claims to not support visual 0x23
[02:40] <elmo> and yes, Icould be a lot less specific
[02:40] <daniels> well, your 3D driver is claiming not to support visual 0x23
[02:41] <daniels> what hardware?
[02:41] <elmo> MGA G550
[02:41] <elmo> it does it for all of them
[02:41] <daniels> dunno
[02:41] <daniels> worksforme
[02:41] <elmo> thanks, dude
[02:42] <daniels> in all seriousness, I would have to poke at the MGA DRI stuff to do that
[02:43] <daniels> and unless you want to rebuild my desktop for me and get a new version of xorg out before the feature freeze and make ddcprobe run just fine on amd64, I'm not going to spend an hour or so chasing up some random warning
[02:43] <elmo> no need to be a drama queen; I didn't ask you to spend anytime on it, least of all an hour
[02:43] <daniels> whatevah
[02:44] <Rotund> elmo: glxinfo | grep 0x23
[02:44] <mvo_> quick poll: do we want "auto-install-recommends" in synaptic on by default (like aptitude)?
[02:44] <Rotund> huh?
[02:44] <elmo> 0x23 24 tc  0 24  0 r  y  .  8  8  8  0  0  0  0  0  0  0  0  0 0 None
[02:45] <elmo> mvo_: given a lot of our recommends can't be satisified in main, I guess not
[02:46] <elmo> if it does get turned on, we should either fix the recommends or pull what we can into main
[02:46] <elmo> no one's done a recommends check since Brazil, AFAIK
[02:46] <Rotund> mvo_: I don't think so.  I noticed X recommends glide.  I don't have a 3DFX, so I really don't want glide
[02:48] <mvo_> thanks. we may turn it off in aptitude then as well
[02:48] <pitti> ogra: so you really want to introduce buffer overflows in hal? :-)
[02:57] <zul> hey
[02:58] <ogra> pitti ? 
[02:58] <pitti> ogra: I write a mail
[02:58] <pitti> ogra: I have some other comments, too
[02:59] <Rotund> when is the feature freeze date for ubuntu?
[02:59] <infinity> Today.
[02:59] <sivang> ogra: pitti always has, it's really hard to pass him on first try :-)) <smiles to pitti>
[02:59] <Rotund> oh.  ouch
[02:59] <sivang> Rotund: yes 
[03:00] <elmo> rotund: http://www.ubuntulinux.org/wiki/HoaryReleaseSchedule
[03:00] <sivang> infinity: do we have a scheduled time in which no uploads would be allowed after that?
[03:00] <Rotund> darn.  I was going to help w/ the auto detect stuff for X, but I gotta leave
[03:00] <infinity> sivang : I'm likely the wrong person to ask. :)
[03:00] <Rotund> day job.  Hopefully I'll have some time tonight
[03:00] <sivang> Rotund: well, maybe for hoary+1 then
[03:01] <sivang> infinity: I've bugged jdub too much in the last 24 hours, I am waiting for my jdub account to elapse :)
[03:01] <Rotund> okay.  I'll aim for then
[03:01] <Rotund> later
[03:01] <ogra> sivang, i'm fine with critics from pitti.... but i need to know what to change to make him happy ;) 
[03:02] <pitti> ogra: hold on, I'm still typing
[03:02] <ogra> yup...i'm patient ;)
[03:02] <zul> fabbione, ping
[03:02] <fabbione> zul: pong
[03:02] <sivang> pitti: lol 
[03:02] <zul> fabbione: i added some pages to the wiki did you get 2.6.11 uploaded yet?
[03:03] <fabbione> zul: not yet. fixing ppc ftbfs and waiting for ia64 to complete the build (and grab the configs)
[03:03] <zul> okie dokies
[03:03] <fabbione> zul: i think i will manage to upload tomorrow morning very early
[03:03] <fabbione> i need to leave quite soon to pick up my parents
[03:03] <zul> k, add your items to the todo list as well when you get a chance
[03:04] <fabbione> zul: yes i know.. it's in my TODO list :-)
[03:04] <zul> hehe
[03:05] <fabbione> this afternoon and tomorrow morning will be a bit hectic for me
[03:05] <Kamion> lunchtime
[03:07] <infinity> What.. The.. Fuck?
[03:07] <infinity> Also, -EWIN.
[03:07] <fabbione> infinity <- 0w3n3d
[03:07] <fabbione> :P
[03:08] <jbailey> fabbione: Is sparc going to be a supported arch in Hoary+1?
[03:08] <zenwhen> hey
[03:09] <fabbione> jbailey: i actually hope to get it for hoary :-)
[03:09] <fabbione> jbailey: sparc.u.c
[03:09] <fabbione> i will make the official announce during the next week
[03:09] <elmo> fabbione: speaking of sparc; how come it hasn't done the kernel yet?
[03:09] <fabbione> so that people can trash it while i am away
[03:09] <zenwhen> Hey I get the following error when installing gnome-icon-theme_2.9.91-0ubuntu2_all.deb
[03:09] <zenwhen> trying to overwrite `/usr/share/icons/hicolor/48x48/apps/gnome-run.png', which is also in package gnome-panel-data
[03:10] <zenwhen> and it dies
[03:10] <fabbione> elmo: because my mirror went of out of sync for 2 days thanks to the people having 5 concurrent sessions
[03:10] <jbailey> fabbione: Cool.  Means I need to get around to getting the sparc from a friend of mine soon.  He's giving it to me when he moves to replace my ultra that died.
[03:10] <fabbione> elmo: and it is picking up now
[03:10] <jbailey> fabbione: Are you doing usparc only, or also supporting classic sparc?
[03:10] <fabbione> jbailey: only sparc64
[03:10] <jbailey> fabbione: Yay. =)
[03:11] <fabbione> jbailey: since i am doing it in my spare time :)
[03:11] <fabbione> but patches for sparc32 are welcome
[03:11] <fabbione> jbailey: be carefull.. i am talking about kernel
[03:11] <fabbione> the userland is the same as debian
[03:11] <elmo> err, yeah, sparc 64-only userland would be real madness
[03:12] <fabbione> that's a complete suicide
[03:12] <jbailey> Ah well.
[03:12] <fabbione> madness doesn't describe it
[03:12] <fabbione> ;)
[03:12] <jbailey> It's no crazier, than, say, running ia64. =)
[03:12] <trulux> hi
[03:12] <trulux> ajmitch_: congrats
[03:12] <elmo> jbailey: err, sure it is?
[03:12] <Mithrandir> jbailey: well.. does that say much? :)
[03:13] <fabbione> now... the only 2 showstoppers for me to upload 2.6.11 are PPC and ia64
[03:13] <mjg59> jbailey: Does sparc64 actually have a useful compiler yet?
[03:13] <fabbione> which is the next arch are we going to drop?
[03:13] <jbailey> mjg59: Does ia64? =)
[03:13] <elmo> jbailey: ia64 is designed to be 64-bit, sparc is known to be more efficent with 32-bit code by default
[03:13] <elmo> mjg59: useful how?  it's enough to compile the kernel and few bi-arch libs Debian has
[03:14] <trulux> tritium: It's going fine
[03:14] <mjg59> elmo: It used to require an egcs snapshot
[03:14] <jbailey> fabbione: Which 'we'? 
[03:15] <trulux> tritium: have i sent you what I'm working on? Maybe you could give me your opinion on formatting and such
[03:15] <tritium> trulux, good.  Sorry I had to bail last night.  Sure, you can do that.
[03:15] <tritium> (no, you haven't sent it yet.  Go ahead and send it, if you like)
[03:15] <pitti> ogra: you have mail (three buffer overflows, two heap corruptions :-( ). I also have some other nitpicks, please don't take it personally
[03:16] <ogra> i wont... is it solvable ?
[03:16] <sivang> seb128: trying to fix 6092, dpkg-buildpackage: Build dependencies/conflicts unsatisfied; aborting.
[03:16] <sivang> seb128: darn, sorry, that was supposed to be:
[03:16] <seb128> ?
[03:17] <sivang> seb128: dpkg-checkbuilddeps: Unmet build dependencies: system-tools-backends (>= 1.1.91)
[03:17] <sivang> seb128: tried to install, no such pkg..
[03:17] <seb128> yeah, you need to backends
[03:17] <jbailey> elmo: As I understood it, it wasn't so much that sparc32 was more efficient as that sparc64 didn't add any wins beyond twice the data bandwidth.  So the efficiency loss is only from hauling twice as much data around when you don't really need to.
[03:17] <seb128> it's probably in NEW
[03:17] <seb128> ping elmo about it
[03:17] <seb128> arg
[03:17] <sivang> elmo: James my freind, how are you today? :-)
[03:18] <seb128> probably need a seed update
[03:18] <thom> elmo: fancy giving me some NEW love? ;-)
[03:19] <sivang> seb128: who's allowd to update the seeds? can you do it?
[03:19] <seb128> yep
[03:19] <elmo> system-tools-backends_1.1.91-0ubuntu1_source.changes
[03:19] <elmo> REJECT
[03:19] <elmo> Rejected: Unknown distribution `unstable'.
[03:19] <seb128> graaa
[03:19] <seb128> thanks elmo
[03:19] <sivang> seb128: woops :)
[03:20] <seb128> my dh_make put unstable by default, not the first time I screw with it :p
[03:20] <seb128> elmo: BTW that's targetted for main, I've to update the desktop seed first, right ?
[03:20] <ogra> pitti....its a lot of work....but doesnt leave me depressive ;) thanks, i'll try to inject your changes in the other patches too, so they will be cleaner, thanks for the help :)
[03:20] <pitti> ogra: Icurrently review your second patch
[03:20] <sivang> seb128: yeah, I can recall at least another time when this happened :) 
[03:21] <trulux> tritium: ok, DCC'ing...
[03:21] <ogra> pitti: leave it... i'll send a new one if i made the corrections to the fist one...dont waste your time :)
[03:21] <sivang> seb128: then I'll wait for the update..
[03:22] <tritium> trulux, okay, I'll take a look at it
[03:24] <seb128> sivang: http://pkg-gnome.alioth.debian.org/system-tools-backends_1.1.91-0ubuntu1_all.deb 
[03:25] <trulux> tritium: OK, thanks!
[03:25] <tritium> trulux, no problem.  Anything you want me to look for?
[03:26] <trulux> tritium: what do you think that should be fixed in the layout and such
[03:26] <trulux> it's going to be a large doc
[03:26] <trulux> is USENIX/IEEETran good for it?
[03:26] <trulux> or may i use book AMS or something alike?
[03:27] <fabbione> jbailey: ppc or ia64 :-) we as in ubuntu ;)
[03:27] <tritium> trulux, okay, I'll take a look
[03:27] <jbailey> fabbione: Ah.  I didn't know that we were looking at dropping arch's.  I'm a day or so behind on my u-devel reading.
[03:28] <jbailey> fabbione: I'd really prefer that ppc wasn't dropped. =)
[03:28] <fabbione> jbailey: i was just kidding
[03:28] <trulux> tritium: Many thanks
[03:28] <jbailey> fabbione: Oh good.  Did you guys find someone for PPC stuff in the kernel meeting yesterday?
[03:28] <fabbione> nope
[03:28] <sivang> seb128: it's the same right?
[03:28] <jbailey> One of those "I don't have the skills but would love to learn them" type of moments.
[03:29] <sivang> seb128: ah, I only need to the binary pkg to build the frontneds?
[03:29] <Kamion> seb128: you don't need to update seeds for build-dependencies
[03:29] <tritium> trulux, where's the big table that you were going to \usepackage{longtable} for?
[03:29] <seb128> Kamion: ok, thanks
[03:29] <trulux> tritium: I didn't used it finally
[03:29] <seb128> sivang: that's the package you are looking for no ?
[03:29] <tritium> trulux, probably good
[03:29] <Kamion> germinate sucks them in automatically once they're in universe
[03:30] <sivang> seb128: I think so, thanks :)
[03:30] <seb128> np
[03:30] <seb128> Kamion: k
[03:30] <tritium> trulux, which conference is this for?  got a url?
[03:31] <jbailey> elmo: What's the best way to tell you that nagios-radius-plugin can be removed now.  nagios-plugins-extra is in the archive now.  (I need to hunt down why nagios-plugins didn't make it, but that's another question)
[03:32] <elmo> nagios-plugins did make; I guess it's FTBFS
[03:32] <jbailey> I'll check it again, I didn't see it in the build logs atall.
[03:33] <elmo> oh, it may be dep-wait then
[03:33] <Kamion> http://people.ubuntu.com/~lamont/buildLogs/n/nagios-plugins/1.3.1.0-12ubuntu2/
[03:33] <elmo> and I removed nagios-radius-plugin
[03:33] <jbailey> Kamion: I'm waiting for ubuntu3.
[03:33] <Kamion> ah
[03:33] <jbailey> elmo: Thanks.  What's the best way to tell you that, though?
[03:33] <Kamion> build-dep on a virtual package probably means lamont has to clear the dep-wait by hand
[03:34] <ogra> pitti: is it ok to drop efd = fds[1] ; and: dup2 (efd, STDERR_FILENO); completely ?
[03:34] <elmo> jbailey: here, if I'm around, email if not
[03:34] <elmo> forcibly given back ubuntu3
[03:34] <pitti> ogra: I would not parse stderr with your "key: value" parser, that makes no sense
[03:34] <jbailey> elmo: Cool thanks.
[03:35] <elmo> oh, meh, that b-d probably means it'll go back into dep-wait
[03:35] <pitti> ogra: either ignore it completely (just redirect to /dev/null), or write it into the hal logs
[03:35] <ogra> pitti: so i throw it away then...thanks
[03:35] <pitti> ogra: I don't know whether lsb_release actually writes something to stderr
[03:35] <elmo> jbailey: if it d-w's again, you might need to fix the libmysqlclient-dev b-d
[03:35] <pitti> ogra: if it makes no sense, redirect it to /dev/null (don't just close it)
[03:35] <ogra> ok
[03:36] <jbailey> elmo: 'k, thanks
[03:36] <pitti> ogra: if it's closed and lsb_release indeed tries to write to stderr, this would crash with SIGPIPE
[03:42] <pitti> ogra: I reviewed the second patch, mailed to you
[03:44] <elmo> err
[03:44] <elmo> is oocalc known to be likely entirely SNAFU in hoary?
[03:45] <fabbione> s/calc//g :P
[03:46] <ogra> pitti: thanks....now i need a cigarette.....uff...as i said, i dont trust my c-foo ;)
[03:46] <ogra> pitti: thanks ;)
[03:46] <elmo> meh, it's not reproduceable except in my expenses
[03:46] <seb128> jdub: around ?
[03:47] <pitti> elmo: for me it crashes when deleting a row
[03:49] <dholbach> seb128: just changed  dh_make 's template to say "hoary" instead of "unstable" - want to upload? :-)
[03:49] <seb128> dholbach: what package ?
[03:49] <dholbach> dh-make 
[03:49] <elmo> pitti: I've got that + crashing when cut'n'pasting a block of data, crashing when trying to open a file
[03:49] <trulux> tritium: no con. , it's for be presented independently
[03:50] <pitti> elmo: d'oh
[03:50] <zul> gah...hate windows
[03:50] <pitti> elmo: the other stuff works for me
[03:50] <dholbach> seb128: deb-src http://ubuntu.gplan.info hoary main
[03:51] <seb128> dholbach: quite busy atm, but I'll have a look on it soon
[03:51] <jbailey> dholbach: It might be worth doing a hack to dh_make to look in /etc/lsb-release for DISTRIB_CODENAME
[03:51] <dholbach> jbailey: oh cool, you're right
[03:52] <jbailey> dholbach: That way the change doesn't have to be constantly changed, is based on the current system, and flows through to derivative distros with the right changes.
[03:52] <dholbach> jbailey, seb128: i'll rework it
[03:52] <dholbach> jbailey: thanks, master yoda :-)
[03:53] <jbailey> dholbach: I knew my wrinkles were showing!
[03:53] <tritium> trulux, it looks good.  I've been trying to find author information on usenix.  All I've found is this: http://www.usenix.org/events/samples/ 
[03:53] <dholbach> jbailey: i'll still have to train my jedi skills 
[03:54] <tritium> trulux, I've used latex-beamer for the talks I've given.  It's pretty sweet.  You might consider it for the presentation slides.
[03:56] <tritium> trulux, if you don't have format guidelines, I think IEEEtran looks nice.  If you use your current format, check your top margin.  It seems too large.
[03:59] <lamont> Kamion: need something cleared?
[04:00] <Kamion> lamont: nagios-plugins - elmo did it though
[04:00] <lamont> elmo: moving stars to multiverse?
[04:00] <lamont> Kamion: cool
[04:00] <daniels> elm	ping
[04:00] <Riddell> haggai: is there a MOTU e-mail list?
[04:01] <daniels> blah
[04:01] <daniels> elmo: ping
[04:04] <haggai> Riddell: no
[04:05] <Riddell> haggai: someone is moaning to me about a broken package, anywhere I can send him to?
[04:05] <lamont> daniels: I was wondering why you were pinging a tree....
[04:05] <daniels> lamont: misguided irssi feature
[04:05] <lamont> heh
[04:06] <trulux> tritium: yeah, how could I change it?
[04:06] <haggai> Riddell: ubuntu-devel.  We're waiting for bugtracking support still
[04:06] <Riddell> haggai: he's already tried there, poor soul will just have to wait or fix it himself
[04:06] <tritium> trulux, \documentclass{IEEEtran} (be sure to download the IEEEtran class)
[04:07] <trulux> tritium: I'm going to test it in a backup, lemme copy the files and try to hange it
[04:07] <tritium> trulux, or, margin setting is probably in the .cls file you're using
[04:09] <trulux> it works out of the box with IEEETran
[04:09] <trulux> without tweaking the tex
[04:09] <tritium> it should
[04:10] <trulux> still I prefer numbered sections and not ro. numbers
[04:10] <haggai> Riddell: hmm, sounds like it
[04:10] <tritium> trulux, your choice :)
[04:11] <lamont> fabbione: you mean I have to figure out how to use imap now? :-)
[04:11] <trulux> tritium: is it easy to change them?
[04:11] <trulux> mdz: ping
[04:11] <tritium> trulux, cuales?
[04:13] <trulux> tritium: el poner nmeros romanos o sequenciales, es que la AMS est genial, tiene montones de objetos, y las book me encantan, el doc, tendr bastantes pginas, as que no se que escoger, tu que dices?
[04:13] <elmo> lamont: meh, done
[04:13] <elmo> daniels: ?
[04:15] <tritium> trulux, oh, yes.  You can redefine how section numbers are produced.  /msg me for details
[04:15] <trulux> tritium: ok
[04:16] <daniels> elmo: is it possible to get a list of which xorg binary packages are in main and which are in universe?
[04:16] <lamont> elmo: thanks
[04:17] <elmo> lbxproxy, libxaw6-dev, proxymngr, twm, x-window-system, xdm, xdmx, xfonts-100dpi-transcoded, xfonts-75dpi-transcoded, xfonts-base-transcoded, xfonts-cyrillic, xfree86-common, xfs, xfwp, xmh, xprt
[04:17] <elmo> are in universe - everything else is in main
[04:18] <daniels> elm	thanks
[04:21] <Mithrandir> I _love_ how building gcc-3.4 kills stuff in /usr/include. :(
[04:21] <pitti> sjoerd: btw, pmount now already uses "quiet" for VFAT (just saw your hal commit)
[04:22] <elmo> Mithrandir: how can it?
[04:23] <Mithrandir> elmo: it removes /usr/include/bits when it's a symlink, at least.
[04:23] <Mithrandir> elmo: I'm doing sbuild shit as root -- I'll need to hack gcc-3.4 I guess.
[04:23] <Mithrandir> but the build system is soooo bad.
[04:24] <seb128> elmo: could you sync evince 0.1.4 from incoming ?
[04:25] <lamont> seb128: libbonobo ftbfs on ia64...  implicit func defn --> scan fail --> bummer
[04:26] <seb128> k
[05:03] <lamont> fabbione: what all did you change in the ia64 kernel?  it doesn't find the cdrom any more...
[05:03] <sivang> jbailey: wrinkles? :)
[05:07] <sivang> jbailey: btw, good idea to take it from lsb-release :-)
[05:10] <mjg59> Anyone with ops on #ubuntu about?
[05:12] <lamont> mjg59: yeha
[05:12] <lamont> yeah, even
[05:13] <lamont> had to check, you see...
[05:13] <sivang> mjg59: what's cracking there?
[05:13] <mjg59> lamont: The first time someone speaks, it carelos sends them a message
[05:13] <mjg59> He doesn't actually seem to be around, but it's rather annoying
[05:13] <lamont> mjg59: and what would you like done to him...
[05:14] <mjg59> Removal? :)
[05:14] <lamont> mdz about yet?
[05:15] <zenwhen> who handles the gnome panel applets?
[05:16] <zenwhen> The window list now picks up items from both of my monitors, the new modem lights applet is useless, and the volume applet acts really weird. lol
[05:19] <kent> zenwhen, you get respons from Ubuntus bugzilla realy fast, and i think that bugzilla is the best way for the devlopers aswell :)
[05:22] <zenwhen> yeah
[05:22] <zenwhen> I suppose
[05:22] <zenwhen> I just hate searching through it and always wind up putting it in thre wrong place or duplicating one
[05:23] <sivang> zenwhen: better put it that way then getting it lost in the trail of irc backlogs..:-))
[05:23] <kent> haha, for me aswell.  And even though i try not to, my first reaction is always to go out on irc and complain ;)
[05:23] <mjt> sorry for asking here, #ubuntu is silent on the topic... Anyone know what's up with hoary repository -- 2nd day now ?  I can't apt-get update, it complains about MD5Sum mismatch for Packages.gz files...
[05:23] <mjt> (or is it just me? :)
[05:24] <zenwhen> I am not having any issues with it
[05:25] <jbailey> mjt: No issues here ppc or i386
[05:25] <mjt> maybe it's apt-proxy on another machine does weird things...  i don't have direct access to the 'net from this machine so have to use apt-proxy
[05:27] <sivang> seb128: did you keep the backends patches I did on the ubuntu tool-backends package?
[05:28] <seb128> what is tool-backends ?
[05:28] <sivang> seb128: ah sorry, system-tools-backedns
[05:28] <zenwhen> buzilla is fun on dialup
[05:28] <zenwhen> kik
[05:28] <seb128> that's the first release
[05:28] <seb128> this is a new package
[05:28] <seb128> sivang: read the changelog ?
[05:29] <sivang> seb128: sure.
[05:29] <seb128> so what's the question ?
[05:29] <dholbach> bbl
[05:29] <zenwhen> hey seb128  https://bugzilla.ubuntulinux.org/show_bug.cgi?id=6322 is solved, btw
[05:29] <seb128> hi
[05:29] <seb128> k, thanks
[05:29] <haggai> mjt: I'm afraid that's an apt-proxy bug, that does happen sometimes
[05:29] <zenwhen> you just have to upgrade gnome panel
[05:30] <seb128> zenwhen: I know that
[05:30] <mjt> haggai: i think it's apt bug, really ;) -- after manually updating /var/lib/apt/lists/* stuff and running apt-get install apt apt-utils (from 0.6.30 to 0.6.32) it works now ;)
[05:31] <zenwhen> seb128, do you know what changed the Window lists's ability to know which apps are on which monitor in a twinview situation? 
[05:31] <zenwhen> It sorted them in Warty, and now it lumps both monitors.
[05:32] <zenwhen> oh
[05:32] <zenwhen> its a bug
[05:32] <zenwhen> lol
[05:32] <zenwhen> I suck and will shut up
[05:32] <haggai> mjt: that might be to do with apt-proxy's handling of the Modified-Since http headers
[05:33] <mjt> haggai: nope, the files on archive and in proxy cache are the same, I verified
[05:33] <mjt> haggai: and nope, apt-get update grabs them
[05:35] <mjt> either way, apt-0.6.32 fixed the issue
[05:35] <mvo_> mjt: I would be suprissed if it is a apt bug that is sovled with upgrading from 0.6.30->32. the changes are nearly all about apt-cdrom :)
[05:36] <mjt> well, not really, but indeed changes are quite small
[05:37] <mjt> still it works now and didn't work a few mins ago
[05:38] <doko> elmo: please can you sync python-mode from unstable?
[05:55] <zenwhen> I was really pleased with how well upgrading from a warty array install disk worked.
[05:55] <zenwhen> I mean hoary.
[05:55] <zenwhen> It went better than most Windows upgrades. ;)
[05:57] <Hwolf> zenwhen. Every time I doubt my choice of running hoary, I think about the pc I just installed for a neighbour, which caught a virus before she got about running windows update.
[05:59] <zenwhen> lol
[05:59] <zenwhen> I am thinking of install hoary on the laptop that is arriving in the next couple hours
[05:59] <zenwhen> installing*
[05:59] <zenwhen> Because I want hoary to work well on it and the only way to affect that is to file bug reports.
[05:59] <zenwhen> :)
[05:59] <Hwolf> zenwhen, she bought an old laptop. 366mhz, 64mb memory, Think I can get it to run ubuntu?
[06:00] <zenwhen> Hwolf, with XFCE, it should be fine.
[06:00] <zenwhen> With fluxbox, it might be quite snappy.
[06:00] <zenwhen> Woith gnome, its going to be sluggish.
[06:01] <mdz> lamont: here
[06:01] <Hwolf> ZenWhen. I call gnome sluggish on my pc. And it's a decent rig.
[06:01] <mdz> Kamion: is localization-config sane for ubuntu?
[06:02] <zenwhen> Oh, I am running a 3Ghz P4 and a Gig of PC3200 ram.
[06:02] <daniels> mdz: anything you can think of before I upload xorg 6.8.1-1ubuntu16?
[06:02] <zenwhen> Gnome is snappy for me.
[06:02] <Kamion> mdz: I was having a look, but got distracted; there's a fair bit of stuff in there
[06:02] <daniels> mdz: works fine for me on amd64 with DDC (although I have to select the resolution), i386 with laptop, and live CD through qemu
[06:02] <mdz> haggai: pong
[06:03] <mdz> daniels: changelog?
[06:03] <Kamion> mdz: I think we should look at it in the long term, but for Hoary I think the least-breakage approach would probably be to clone its keyboard mapping logic into xserver-xorg's config script
[06:03] <Kamion> and acknowledge Konstantinos in the changelog
[06:03] <Hwolf> zenwhen. I've got an amd 1800+ and 512mb ram, which should be more than plenty for any typewriter inet/mail rig. Yet gnome is sluggish.
[06:03] <mdz> Kamion: it strikes me as the sort of thing that we might be able to implement better by modifying the affected packages directly
[06:03] <Kamion> mdz: yeah, I agree
[06:04] <mdz> Kamion: I thought we already did that for Warty?
[06:04] <Kamion> mdz: did which?
[06:04] <mdz> (the xserver logic)
[06:04] <Kamion> mdz: no, I looked at what we did for Warty this morning, and it's the source of a lot of bugs
[06:04] <haggai> mdz: you asked me to ping you yesterday re bounties but I missed you when I got back
[06:04] <zenwhen> Hwolf, sluggish compared to what?
[06:04] <Kamion> mdz: we automatically pick a keyboard mapping based on the selected *language*
[06:04] <daniels> mdz: /msg
[06:04] <daniels> mdz: no, we guess based on $LANG
[06:05] <mdz> Kamion: the stuff we have now is going to be obsoleted by smurfix's layout selector
[06:05] <Hwolf> zenwhen, compared to windows xp if kept virus-free. sluggish compared to my expectations.
[06:05] <Kamion> mdz: how so? smurfix's layout selector will presumably set debian-installer/keymap and the mapping will need to be just the same
[06:05] <Kamion> I was under the impression that the layout selector was a replacement for kbd-chooser
[06:05] <mdz> Kamion: it will find both the appropriate console keymap and the appropriate XKB keymap
[06:06] <Kamion> mdz: hrm. ok.
[06:06] <daniels> mdz: is that going to be good to go for hoary?
[06:07] <Kamion> it's a hoary goal, and seems to be already well underway
[06:08] <Kamion> I haven't had time to try it out yet though :-(
[06:08] <bluefoxicy> OH NO YOU DIDN'T
[06:08] <mdz> daniels: he's quite far along with the console stuff, and XKB is next
[06:08] <bluefoxicy> bluefox@icebox:~$ cat /boot/config-2.6.10-3-amd64-generic | grep CAPA
[06:08] <bluefoxicy> CONFIG_SECURITY_CAPABILITIES=m
[06:08] <daniels> mdz: cool
[06:08] <Kamion> elmo: not to be are-we-there-yet, but could I have file-kickseed in main so I can upload d-i with it?
[06:09] <mdz> Kamion: I've tried it, it's quite nice
[06:09] <bluefoxicy> M?!
[06:09] <tseng> bluefoxicy: um
[06:09] <tseng> the bug is patched
[06:09] <daniels> bluefoxicy: please calm down, dude.
[06:09] <tseng> and it went back to being M
[06:09] <tseng> because its freaking useless
[06:09] <mdz> amu: what is the status of the kubuntu seeds?
[06:10] <bluefoxicy> tseng:  ah
[06:10] <elmo> Kamion: it'll go in next cron.daily
[06:10] <Kamion> elmo: ta
[06:11] <bluefoxicy> tseng:  I thought the kernel devs were moving away from root and more towards caps though
[06:11] <lamont> mdz: figured I'd lob mjg59's #ubuntu issue at you
[06:12] <tseng> bluefoxicy: uh
[06:12] <mjg59> Oh, he's quit now
[06:12] <lamont> mjg59: ah, good
[06:12] <elmo> [Updating]  python-mode (4.62-1 [ubuntu]  < 4.70-1 [debian] )
[06:12] <elmo> doko: UVF, please mail mdz, jdub & cc me
[06:13] <bluefoxicy> tseng:  check out kernel/module.c
[06:13] <elmo> doko: you know a bunch of main packages are now built from gcc-4.0 right?
[06:13] <bluefoxicy> tseng:  the pattern "uid" isn't found, but "CAP_SYS_MODULE" is required for inserting and removing modules
[06:14] <tseng> you dont need that stupid lsm caps thing to use caps
[06:14] <tseng> im not even sure what it does, tbh
[06:15] <bluefoxicy> tseng:  Oh, you don't?  I thought they moved capabilities out of the core and into a module "because it's bad to have policy in the kernel" or something
[06:15] <bluefoxicy> tseng:  thanks for the info
[06:16] <tseng> grsec and rsbac use caps w/o ever touching that lsm thing
[06:16] <tseng> suid apps drop their caps on start (postfix), etc
[06:16] <tseng> if you figure out what that module even does, lemme know
[06:17] <doko> elmo: yes, fastjar, libgcc1, libgcj-common. as discussed with jbailey and mdz
[06:17] <elmo> doko: and libstdc++6
[06:17] <amu> mdz: started 30min. ago   
[06:17] <elmo> doko: gcc-4 doesn't ABI bump from 3.4? 
[06:17] <doko> elmo: for i386, powerpc and ia64. correct.
[06:17] <pitti> Morning mdz
[06:17] <mdz> amu: what happened?
[06:18] <mdz> pitti: morning
[06:18] <doko> elmo: no
[06:18] <elmo> doko: neat
[06:18] <sivang> morning mdz 
[06:18] <mdz> doko: we did not discuss libstdc++6
[06:18] <mdz> but apparently it's only used by a few packages, the ones currently using gcc-3.4 I guess?
[06:19] <amu> mdz: just started working, 1h or so      
[06:19] <doko> mdz: correct, except for amd64, where we only had libstdc++6-0
[06:19] <haggai> am I allowed to use gcj-3.4 for OOo2?  That might let me build OOo with gcj...
[06:20] <trulux> mdz: there?
[06:21] <mdz> trulux: yes, but backlogged
[06:21] <doko> haggai: gcj-3.4 is still in universe. gcj-4.0 will enter main.
[06:21] <mdz> amu: are you feeling ok?  usually we do not start work at the same time :-)
[06:21] <mdz> trulux: I have that URL
[06:22] <haggai> doko: hmm, that should be ok too
[06:22] <trulux> mdz: ok, fine, lemme know if you need something
[06:23] <amu> mdz: ;) hehe i've flexile working hours, our hole family is ill atm :)    
[06:24] <bluefoxicy> pitti: have you looked into adding netdev-random to the hardened kernels
[06:24] <mdz> trulux: looking at the init patch; is it certain to be a no-op in the default case (selinux disabled)?
[06:24] <pitti> bluefoxicy: no, I didn't. Is there an URL for that?
[06:24] <mdz> it's not clear
[06:24] <bluefoxicy> pitti:  I dunno.  tseng might know
[06:24] <bluefoxicy> pitti:  I just yanked 3000 and 3001 from the hardened gentoo kernels :)
[06:24] <bluefoxicy> and put them in the Ubuntu ones
[06:24] <trulux> mdz: It will work w/o selinux too
[06:25] <bluefoxicy> no idea where they got 'em
[06:25] <tseng> the url is really old, used to be maintained by RML
[06:25] <tseng> he signed it over to albeiro, who seems to just mail them to me.
[06:25] <bluefoxicy> buh
[06:25] <tseng> bluefoxicy: if we want to do something with netdev-random
[06:25] <mdz> trulux: it looks like it will print an error about mounting selinuxfs
[06:26] <tseng> it really needs to be cleaned up and proposed upstream
[06:26] <mdz> it needs to completely disable itself if selinux is not in use
[06:26] <bluefoxicy> tseng:  I'm thinking more entropy == faster GPG key generation
[06:26] <tseng> bluefoxicy: erm
[06:26] <mdz> trulux: on the dpkg patch I defer entirely to Keybuk
[06:26] <tseng> bluefoxicy: gpg key isnt very slow here
[06:26] <tseng> bluefoxicy: but our main case for netdev-rand was SSP, which now uses urandom
[06:27] <tseng> and alot less bits
[06:27] <mdz> trulux: do you have links to the cron and pam patches?
[06:27] <bluefoxicy> tseng:  over here when I do it (I do 4096 bit keys sometimes) it spits out like 5 lines of junk to the terminal, then one char per second, unless I move the mouse
[06:27] <mdz> trulux: are you sure that cron isn't already done?
[06:27] <mdz> trulux: it build-depends: libselinux1-dev
[06:28] <trulux> mdz: libselinux1-dev needs to be in base, as well libselinux1 too
[06:28] <tseng> bluefoxicy: im not sure its right for the ubuntu kernel
[06:28] <trulux> mdz: cron is not done I think
[06:28] <tseng> bluefoxicy: alberio is barely around lately.
[06:28] <bluefoxicy> tseng:  what about mainline
[06:28] <tseng> if someone wanted to maintain it, and keep up with all the new drivers
[06:28] <tseng> then yes, it could be somewhat useful
[06:29] <bluefoxicy> heh
[06:29] <bluefoxicy> I'm just curious because people always say not to use /dev/random unnecessarily, like it's some precious finite resource that you need to use when you need great entropy
[06:30] <tseng> theyd be right
[06:30] <bluefoxicy> tseng:  so anything that pumps it up with more high-quality entropy would be good then?  Or am I leading a bad conclusion?
[06:31] <Keybuk> mdz: I've yet to see a dpkg patch that I would even consider
[06:31] <dholbach> re
[06:31] <tseng> bluefoxicy: netdev-random touches every net driver
[06:31] <bluefoxicy> tseng:  I noticed, I had to manually patch 3
[06:31] <tseng> pretty darn intrusive
[06:32] <bluefoxicy> tseng:  is there any other way to tell if it's coming from net
[06:33] <trulux> mdz: I've hacked out the behavior of libselinux and added a bootstrapping-capable package manager context setting helper that may help to avoid the dirty postinst hacks for dpkg
[06:34] <tseng> bluefoxicy: just by watching the pool and not doing anything else adding entropy afaik
[06:34] <trulux> mdz: I've sent it to Stephen (NSA) but dunno if it will get mainline
[06:34] <trulux> bluefoxicy: getting entropy from network devices is pretty a weak thing
[06:34] <tseng> its very effective, actually
[06:34] <trulux> bluefoxicy: don't bother with it, believe in me
[06:34] <tseng> just not that practical
[06:35] <tseng> in terms of maintainability
[06:35] <bluefoxicy> trulux:  well, they're affected by the load on routers along the path, noise, cable run length, load on other servers, everything half the internet is doing
[06:35] <bluefoxicy> I'd consider that fairly difficult to reliably poison
[06:36] <trulux> bluefoxicy: the Linux TCP/IP stack is somewhat a predictable design and this has been demonstrated many times, but, I can admit that it may require many more skills and a deep study on how it works
[06:36] <bluefoxicy> tseng:  reading from the soundcard's analog in is also neat, since you can never perfectly buff out the noise on the mic port.  Audio-entropyd I think does that
[06:36] <trulux> it's like gathering entropy from the sound card
[06:37] <bluefoxicy> trulux:  interrupt timing, hardware not software
[06:37] <trulux> would you trust in something that can get attached to a poisoned jack which sends a non-changing, repeated stream on a certan frequency which makes such amount of "true random" bits reallly true randomized?
[06:37] <bluefoxicy> heh
[06:37] <trulux> bluefoxicy: it's affected by software AFAIK :)
[06:38] <bluefoxicy> trulux: interrupts are hardware.  They go through two interrupt chips that poke a port on your CPU
[06:38] <trulux> have in mind a thing, and don't forget it: if you guess just one bit of the entrop, then the rest is completely in compromise
[06:39] <trulux> the atacker should just greab a frozen seed of the compromised entropy and try to use it against the application at issue
[06:39] <trulux> the standard random.c gives a good quality PRNG
[06:39] <trulux> if you don't like it
[06:39] <Kamion> that kind of depends on the quality of the input hash used for the entropy pool, of course
[06:39] <trulux> let me know and I will port Fortuna CSRNG to latest sources
[06:40] <Kamion> one ought not to be able to use knowledge of the input to work out what's in the entropy pool unless one knows *all* the input
[06:40] <Kamion> (I don't know if Linux' entropy hash is actually good enough for that, though)
[06:42] <lamont> so if package A gets absorbed into package B, should package B provide: A?
[06:42] <trulux> yep
[06:42] <trulux> Kamion: Linux uses SHA1
[06:43] <trulux> Fortuna is based on Schneier's CSRNG design, relies in SHA256
[06:43] <trulux> (real) implementation from CyptoAPI
[06:45] <Kamion> it's not the SHA part that's important AFAIK, it's the CRC done when mixing input into the entropy pool
[06:45] <Kamion> the comment in the kernel source says "This is not cryptographically strong, but it is adequate assuming the randomness is not chosen maliciously, and it is fast enough that the overhead of doing it on every interrupt is very reasonable."
[06:45] <Kamion> so sounds like netdev-random is not such a good idea with that
[06:46] <dholbach> wb ogra
[06:47] <pitti> Hi ogra! I made some further comments in the meantime
[06:47] <pitti> ogra: but it looks much better now :-)
[06:48] <ogra> great.... just getting the mail......
[06:48] <trulux> Kamion: check jlcooke's fortuna
[06:48] <trulux> jlcooke.ca/random
[06:49] <wasabi> (probably cuz hal restarts)
[06:50] <ogra> wasabi: rather because it doesnt ;)
[06:50] <wasabi> Ahh. Well, something. ;)
[06:50] <pitti> ogra: why it shouldn't restart?
[06:51] <wasabi> THis has got me thinking about some way to mark packages which require you to be either "logged out" on installation, or similar.
[06:51] <wasabi> So that theydownload, but don't install until you log out.
[06:51] <pitti> wasabi: actually gvm should reconnect to a new hal
[06:51] <wasabi> Yeah, in this case yes. ;)
[06:51] <pitti> wasabi: in warty this "just works"
[06:51] <pitti> wasabi: but it obviously got broken in Hoary
[06:52] <ogra> pitti: if hal doesnt restart gvm will be left without backend....dont take me to serious ;)
[06:52] <wasabi> Well, I'd hate to have a user upgrade... like, gaim or something, because of a security bug, but never inform them that they need to close/reopen gaim for it to take effect.
[06:52] <wasabi> Since we never have to actually reboot Linux, nobody will. ;)
[06:52] <pitti> ogra: right, but usually it does restart and thus gvm is probably failing to reconnect
[06:53] <pitti> wasabi: hmm, usually I add a note to the USN if something needs to be restarted manually
[06:53] <pitti> wasabi: I probably forgot it in gaim, sorry
[06:53] <ogra> pitti: btw, HAL_INFO and HAL_ERROR are not available in my testbed code ;) thus i commented these out....
[06:53] <wasabi> Naw, I'm just using it as an example.
[06:53] <pitti> ogra: why not?
[06:53] <wasabi> Does USN's pop up when you upgrade?
[06:53] <pitti> ogra: ./hald --verbose=yes --daemon=no
[06:53] <pitti> wasabi: no
[06:53] <wasabi> My mom doens't read them then
[06:54] <pitti> wasabi: they get filed to ubuntu-security-announce@list.u.c
[06:54] <pitti> :-)
[06:54] <wasabi> Actually.
[06:54] <pitti> wasabi: right, somehow this should be integrated
[06:54] <wasabi> She wouldn't read them even if they popped up.
[06:54] <wasabi> They woul dneed a "restart gaim now" button.
[06:54] <pitti> ;-)
[06:54] <wasabi> Like Windows has. =)
[06:54] <wasabi> "Restart Windows Now"
[06:54] <wasabi> hahah
[06:54] <pitti> wasabi: no, they would need a button "just click this unless you know what you are doing" :-)
[06:54] <wasabi> yeah
[06:55] <ogra> pitti: because i dont use hal for the basic development..... richard huges has a nice testbed that prints out the set_string and device valueshal would recieve andi save a lot of time only compiling 150 lines instead of the whole daemon
[06:55] <Kamion> trulux: ah, interesting
[06:55] <pitti> ogra: why you want to compile the whole daemon?
[06:55] <pitti> ogra: caling make will just recompile the changed stuff
[06:55] <ogra> pitti: i will remove the comments in the real patch
[06:55] <pitti> ogra: ok
[06:55] <pitti> ogra: anyway, calling make in the build tree and executing hal there works fine
[06:56] <ogra> jup.... but i dont see hats going on .... as i do if the set_string or callout device just print to stdout
[06:56] <ogra>  hats/whats
[06:57] <trulux> pitti: what's the current status of hardened kernels?
[06:57] <pitti> trulux: the current version works reasonably well; however, it does not ship any firmware images and vesa framebuffer breaks
[06:57] <ogra> pitti: just makes initial development easier.... later i only have to add the headers and HAL_ERR HAL_INFO
[06:58] <pitti> ogra: sure, that's fine :-)
[07:00] <ogra> pitti: stdin is not fds[0]  ??
[07:00] <pitti> ogra: no, when doing "int fds[2] " you must not access fds[2] 
[07:00] <ogra> pitti: (Ahem. Womit fangen wir an zu zhlen? :-))
[07:01] <pitti> ogra: yes, SCNR :-) I thought it would be obvious
[07:01] <ogra> h, ok... now i see it
[07:04] <bluefoxicy> is anyone else's hoary broke?
[07:04] <bluefoxicy> I can't install ubuntu-desktop because lsb won't install
[07:06] <bluefoxicy> locales depends on glibc-2.3.2.ds1-20ubuntu6
[07:06] <bluefoxicy> glibc-2.3.2.ds1-20ubuntu6 does not appear to be available
[07:06] <bluefoxicy> lsb depends on locales
[07:07] <jbailey> ls
[07:07] <jbailey> bah
[07:08] <bluefoxicy> tuxnes metroid.nes
[07:08] <trulux> tritium: ping
[07:08] <bluefoxicy> oh, you're not justin o.o
[07:08] <trulux> bluefoxicy: damn! dcc me the rom :)
[07:08] <Kamion> bluefoxicy: wait a bit, it'll appear on your architecture in time
[07:08] <bluefoxicy> Kamion:  *nod*
[07:09] <trulux> is Hoary ready for x86? stable for desktop?
[07:09] <bluefoxicy> trulux:  abril
[07:09] <trulux> I want to do the warty->hoary on my development and daily use box
[07:09] <bluefoxicy> trulux:  it's suitable for the desktop if you're a hacker
[07:09] <trulux> bluefoxicy: lo soy?
[07:10] <bluefoxicy> trulux:  I have no idea wtf you just said :)
[07:10] <Kamion> bluefoxicy: it's just waiting for the i386 buildd to get round to building new locales
[07:10] <trulux> bluefoxicy: Am I?
[07:10] <trulux> :)
[07:10] <bluefoxicy> :P
[07:10] <bluefoxicy> Kamion:  /me on amd64
[07:10] <Kamion> bluefoxicy: yes, and? the i386 buildd builds arch: all packages
[07:10] <trulux> bluefoxicy: be sure to know if you really want to answr, it's trulux's ego feeding
[07:10] <Kamion> such as locales
[07:10] <trulux> noy a good idea
[07:10] <trulux> :)
[07:10] <Kamion> oh, i386 failed for some reason
[07:11] <bluefoxicy> Kamion:  you cross-build all archs on i386?
[07:11] <tseng> bluefoxicy: no, all-arch packages
[07:11] <bluefoxicy> ah
[07:11] <tseng> not arch specific
[07:11] <bluefoxicy> was about to say, amd64 or PPC G5 is MUCH faster than x86 :)
[07:11] <Kamion> you don't know what architecture: all means?
[07:11] <bluefoxicy> Kamion:  no, I read all arch, not arch all
[07:12] <zenwhen> bluefoxicy, I dont know about MUCH faster.
[07:12] <zenwhen> Maybe a bit.
[07:12] <zenwhen> x86 is by no means a slow poke dead arch
[07:13] <Kamion> who cares, it's not like the buildds are running at full capacity
[07:13] <Kamion> (for now anyway ...)
[07:14] <ogra> pitti: any idea why i still dont get my prompt back after everything is closed ?
[07:14] <pitti> ogra: ECONTEXT
[07:14] <Kamion> Mithrandir: could you check out the glibc i386 build failure?
[07:14] <ogra> pitti: ....i.e. the command hangs after xecution
[07:14] <pitti> ogra: which command?
[07:15] <bluefoxicy> zenwhen:  G5?  :)
[07:15] <ogra> my compiled binary of lsb_release
[07:15] <bluefoxicy> zenwhen:  daed has a dual-opteron his boss bought him
[07:15] <pitti> ogra: hm, what do you mean? executing lsb_release does return to me...
[07:15] <bluefoxicy> he uses it for researching 3D graphics or something (i.e. he's writing his own 3D engine and it takes too long to compile on x86)
[07:15] <ogra> pitti: it should drop me at the prompt after all the strings are printed
[07:16] <pitti> ogra: right, works for me
[07:16] <bluefoxicy> anyway
[07:16] <ogra> pitti: as i said, i compile a single binary from the .c file which i run on the commandline.... normally the code returns after execution... but lsb_release_new hangs
[07:17] <pitti> ogra: strace it, where it hangs?
[07:17] <ogra> heh... in read
[07:17] <ogra> indeed
[07:18] <ogra> uuuh what is ERESTARTSYS
[07:18] <ogra> sounds weird
[07:20] <bluefoxicy> is evolution supposed to cause a floating point exception?  :o
[07:21] <ogra> bluefoxicy: only on amd64
[07:21] <ogra> bluefoxicy: its the new floating point exception generator ;)
[07:22] <sivang> ogra: lol
[07:23] <bluefoxicy> ogra:  will that increase the size of my entropy pool
[07:24] <ogra> bluefoxicy: depends how deep you dig ;)
[07:24] <bluefoxicy> heh
[07:25] <bluefoxicy> damnit debian doesn't have a .deb for sunbird either
[07:25] <bluefoxicy> "Once you are done building, run "make" in xpfe/communicator ( this is a bug )"  lol
[07:27] <tritium> trulux, pong
[07:28] <trulux> tritium: hey! how was your time? I was in the academy learning German
[07:28] <dholbach> trulux: Guten Abend, wie war Dein Deutschkurs?
[07:28] <tritium> trulux, fine, thanks.  I even got stopped by some Mormons since they saw the ashes on my forehead.
[07:29] <trulux> tritium: a lot of progress, printed IEEETran How To and did some tweaks on the layout
[07:29] <tritium> trulux, excellent
[07:29] <trulux> dholbach: Hallo, Mein Deutschkurs war gut
[07:30] <trulux> tritium: still I don't get the point on how to change the sections (and subsections) formatting (make them bold is a blocking issue, i need it)
[07:30] <trulux> btw
[07:30] <trulux> libsafe and Stack Shield are right broken
[07:30] <trulux> bluefoxicy: SSP is the final way to go for sure
[07:31] <tritium> I'll /msg you, trulux
[07:31] <dholbach> trulux: Woher kommst Du?
[07:33] <bluefoxicy> trulux:  I like libsafe in conjunction with SSP, though the overhead may not be appropriate.  I haven't had a chance to evaluate the overhead personally, but it's apparently "negligible".  Etoh managed to make 8% out of it on a test SSP did 4% on though, so it's likely something like applying SSP ~3 times :)
[07:33] <trulux> dholbach: Ich komme aus Spanien
[07:33] <Mithrandir> Kamion: will do
[07:33] <Kamion> thanks
[07:37] <Kamion> bluefoxicy: re ubuntu-devel, um, "Lightning" is not a misspelling
[07:38] <bluefoxicy> Kamion:  wow.  you're right
[07:38] <bluefoxicy> google said it was >:|
[07:38] <Kamion> dict > google
[07:39] <bluefoxicy> Kamion:  uhh, well uhh
[07:39] <bluefoxicy> Happy new year then?  :)
[07:39] <Kamion> :)
[07:46] <tritium> op to #ubuntu please?
[07:48] <sivang> tritium: what's going on there? still the one that PMs ?
[07:48] <tritium> sivang, it's okay now.
[07:48] <tritium> someone was changing nicks rapidly and flooding
[07:48] <sivang> tritium: ah
[07:48] <sivang> :-)
[07:49] <sivang> sorry,
[07:49] <sivang> that is  = :-(
[07:49] <tritium> no worries.  it's not happening any more. :)
[07:49] <sivang> good :)
[07:49] <tritium> thanks, sivang 
[07:52] <sivang> pitti : btw, take a look at #1849 tell me what you think? 
[07:54] <sivang> pitti: (probably better add commmenst then talking here though :)
[07:55] <pitti> sivang: mom, phone
[07:55] <sivang> pitti: cool :)
[07:56] <carlos> pitti: dude, postgres' gettext setup sucks too much or you have a bug in your script to extract .po and .pot files...
[07:57] <carlos> pitti: I only got one .pot file (from debconf templates) and a bunch of .po files inside different directories
[07:57] <pitti> carlos: wait, phone
[07:57] <carlos> ok
[07:57] <sivang> carlos: hmm, hoary mega import? ;-)
[07:58] <carlos> sivang: working on it, yes
[07:58] <carlos> :-)
[07:59] <sivang> carlos: yay!
[08:04] <pitti> carlos: the reason is that there are no pot files in the upstream source
[08:05] <pitti> carlos: there are only po files
[08:05] <pitti> carlos: I'm afraid you have to get along with that
[08:05] <carlos> pitti: and how do you get a .pot file when translating it into a new language?
[08:05] <pitti> sivang: I followup to the bug now
[08:05] <pitti> carlos: hmm, ask upstream?
[08:05] <pitti> carlos: I can look in the cvs, though
[08:07] <carlos> pitti: don't worry, will ignore postgres until I have time to look into it
[08:07] <carlos> pitti: I asked you because you work on its .deb package, in case you know how it works
[08:07] <sivang> pitti: thanks
[08:07] <carlos> don't lose your time with it, it's not a critical application to get translated
[08:07] <carlos> so it can wait
[08:08] <pitti> carlos: hmm, no pot in CVS either
[08:08] <carlos> that's normal, GNOME packages don't have it neither
[08:08] <carlos> pitti: but they create it on build time
[08:09] <pitti> sivang: cool, shadow and /etc/sudoers changes are already done? Great, thanks Kamion :-)
[08:10] <sivang> pitti: yes :) kamion rocks, and rcaskey did the original patch :)
[08:10] <pitti> sivang: followed up
[08:11] <sivang> pitti: thank you && http://muse.19inch.net/~sivan/g-s-t/sivan_uid_fix.diff <== can I "upload" ;-) ?
[08:15] <carlos> sivang: all people are hosted at muse?
[08:15] <carlos> :-D
[08:15] <pitti> sivang: please attach such things to the bug as patch in the future
[08:15] <sivang> carlos: hehe :)
[08:16] <pitti> sivang: something is wrong with this  patch, did you happen to swap the old and new dsc?
[08:16] <sivang> pitti: lemme see..
[08:16] <pitti> sivang: anyway, I don't know the Gnome API, but if that works, it looks good
[08:17] <sivang> pitti: ok, thanks
[08:17] <pitti> sivang: it just might look a little confusing
[08:18] <sivang> pitti: btw, I don't think I swapped the dsc's
[08:18] <pitti> sivang: since it has knobs for adjusting the value, which are just greyed out
[08:18] <pitti> sivang: are you sure that the value can't be changed in another way?
[08:18] <pitti> sivang: it would be nice to disable the uid changing code in addition, just to be sure
[08:19] <pitti> sivang: add comments signs around the code
[08:19] <Mithrandir> pitti: what other changes did you make to glibc than the multiarch patch?
[08:20] <pitti> Mithrandir: none
[08:20] <sivang> pitti: pretty much, I can unshow the widget all together if that's reasonable, I took that approach to differ as less as we can from upstream, but I can attempt remove the code form the frontend and the backend as well, but I don't think the uid can be changed that way unless someone fire up the backend by hand using it's new directive system.
[08:20] <pitti> Mithrandir: just added your dpatch, added it to 00list, added changelog
[08:20] <pitti> Mithrandir: however, I did not try to build it on i386
[08:20] <pitti> Mithrandir: I used a powerpc chroot since I still have this with glibc from my previous upload
[08:21] <pitti> sivang: don't remove the code, just comment it out
[08:21] <pitti> sivang: well, and having a greyed out box which displays the uid is okay for me
[08:22] <pitti> Mithrandir: can you reproduce the FTBFS on i386?
[08:22] <Mithrandir> pitti: haven't tried yet.
[08:23] <Mithrandir> I'm grabbing the build log now 
[08:25] <amu> Kamion: will the germinate-output updated after i upgraded the seed? 
[08:27] <Kamion> amu: it's updated daily at midnight; if you want it pushed, let me know
[08:27] <amu> Kamion: so fire in the hole ;)
[08:30] <ogra> pitti: any idea about my hanging program ?
[08:30] <rubenv> anyone knows if kiko ever comes on IRC?
[08:30] <pitti> ogra: no, sorry
[08:30] <Kamion> amu: done
[08:30] <ogra> pitti: but you see it too ?
[08:30] <rubenv> (read: somebody point me towards kiko, i need him now :-))
[08:30] <pitti> ogra: no, lsb_release returns fine for me
[08:30] <Kamion> amu: looks like you still have a lot of GNOME stuff to get rid of :)
[08:30] <amu> Kamion: i'm impressed how it works! nice sys    
[08:31] <ogra> pitti: the binary i sent you ? (i'm not talking about lsb_release)
[08:31] <Kamion> amu: if you're wondering where something comes from, see the rdepends/ directory
[08:31] <Kamion> e.g. http://people.ubuntu.com/~cjwatson/germinate-output/kubuntu-hoary/rdepends/gnome-desktop/gnome-about
[08:31] <Kamion> amu: thanks :)
[08:32] <Kamion> amu: ah, perhaps you should just drop ubuntu-desktop from your desktop seed for now so that you can see what's going on?
[08:32] <Kamion> amu: you can add kubuntu-desktop back in later
[08:32] <pitti> ogra: ah, I didn't see your mail
[08:33] <Kamion> rubenv: he quit #canonical nearly an hour ago
[08:33] <amu> Kamion: yep, this was the first try ... probably the -desktop is needed first
[08:33] <Kamion> yeah, you'll need to get rid of it before you can make the output make any sense at all
[08:34] <rubenv> Kamion: crap
[08:34] <rubenv> thanks anyway :-)
[08:34] <Mithrandir> lamont: about?
[08:34] <lamont> Mithrandir: yeah
[08:34] <Kamion> rubenv: he's in a sprint at the moment, I'm guessing he's gone for dinner since a bunch of people left at the same time
[08:34] <Mithrandir> lamont: I'm looking at the glibc build failure.. any great ideas?
[08:34] <rubenv> the cape town sprint?
[08:34] <Mithrandir> make[3] : *** No rule to make target `/build/buildd/glibc-2.3.2.ds1/build-tree/i386-libc/stdio-common/stamp.o', needed by `/build/buildd/glibc-2.3.2.ds1/build-tree/i386-li
[08:35] <Mithrandir> bc/libc.a'.
[08:35] <Mithrandir> that looks kinda bad.
[08:35] <lamont> doko was thinking it could be because it's doing parallel make
[08:35] <ogra> pitti: the code never seems to return from the fgets loop....but i dont see any logical error
[08:35] <rubenv> i'll head on to #canonical for stalking behaviour :-)
[08:35] <lamont> (i.e., borken dependenencies)
[08:35] <rubenv> bah, no stalking then :-)
[08:35] <pitti> ogra: I think I have a vague idea what's going wrong
[08:35] <amu> Kamion: the file all show me at the end the total diskuseage?
[08:36] <sivang> rubenv: you can also find him around #pygtk if you need him
[08:36] <pitti> ogra: I think there is some synchronisation missing between the two processes
[08:36] <rubenv> if you want, could you ping him for me?
[08:36] <Kamion> rubenv: private channel ...
[08:36] <rubenv> Kamion: yeah, i noticed
[08:36] <Mithrandir> lamont: yeah, possibly.. I haven't seen it on i386 before and I've built that glibc a fair number of times.
[08:36] <Kamion> amu: yeah
[08:36] <sivang> rubenv: this is where I always talk to him :)
[08:36] <Kamion> amu: "deb size" and "installed size" respectively
[08:36] <pitti> ogra: ahem, what shall I do with this binary?
[08:37] <pitti> ogra: in addition, I don't have amd64
[08:37] <ogra> pitti: you should just see the error......
[08:37] <Kamion> amu: the first in bytes and the second in kilobytes, for some strange reason
[08:37] <Mithrandir> /bin/sh: /build/buildd/glibc-2.3.2.ds1/build-tree/i386-libc/stdio-common/errlist-compat.cT: No such file or directory
[08:37] <ogra> pitti: argh, dumb me
[08:37] <Kamion> rubenv: cape town> right
[08:38] <amu> Kamion: thanks, and total size -50MB for d-i ? 
[08:38] <Mithrandir> lamont: I agree to the parallell build thing; want to try a manual build with NJOBS set to 1?
[08:39] <pitti> sivang: again, your patch is wrong; it _removes_ the changelog entry and the patch; you have to do it the other way round
[08:39] <ogra> pitti: what kind of sync do you mean ? the child is called, drops its output to the parent and if nothing is in handle anymore the fgets while loop must end ... i dont see a sync problem there
[08:39] <sivang> pitti: ok, I'll switch it
[08:40] <Kamion> amu: that's at the end of the 'installer' output; 16MB for i386, but it's more like 34MB for powerpc so allow a bit more
[08:40] <sivang> pitti: yeah, noticed it now :-/
[08:40] <lamont> Mithrandir: you're suggesting that I do that, or saying you will?
[08:40] <pitti> ogra: I had a similar problem when I did this for my "debcrash" project
[08:41] <pitti> ogra: I solved this with a timeout
[08:41] <ogra> hmm...weird
[08:41] <pitti> ogra: however, did it work with your old version with open/read?
[08:41] <Kamion> I should probably put germinate output up for all architectures
[08:41] <amu> Kamion: mdz: do we add theopencd packages to the CD's ? 
[08:41] <ogra> pitti: sure
[08:41] <Mithrandir> lamont: suggesting you do it; I don't think I've got an account on any i386 buildd boxes.
[08:41] <Kamion> amu: we intend to for Ubuntu; I don't know what's happening for Kubuntu yet
[08:41] <Kamion> amu: we don't yet, though
[08:41] <ogra> pitti: but this code was quite ugly compared to the one i'm looking at
[08:42] <amu> Kamion: it rocks, i think it's a must to have it
[08:42] <pitti> ogra: yeah, that's why I wanted to use a line-reading function
[08:42] <pitti> ogra: I strongly suppose that your client does not recognize the EOF of the stream
[08:42] <pitti> ogra: does the child process terminate?
[08:44] <ogra> pitti: hmm, strace says yes: --- SIGCHLD (Child exited) @ 0 (0) 
[08:44] <pitti> ogra: oh, that's interesting
[08:44] <Kamion> amu: seems to be a 107MB tarball at the moment (http://maitri.ubuntu.com/theopencd/ubuntu/winfoss/latest/), but Henrik said he could reduce that
[08:45] <ogra> pitti: and i saw a zombie this afternoon, remember ?
[08:45] <pitti> ogra: that was the parent process which did not wait for its child
[08:45] <pitti> ogra: you must call wait() somewhere in the parent
[08:45] <ogra> pitti: yeah, strace ends with: read(3,
[08:46] <ogra> and hangs there forever
[08:46] <ogra> pitti: ok 
[08:46] <pitti> ogra: however, this is a bit tricky. You can't call it before fgets(), you must call wait() after the loop
[08:47] <sivang> pitti: ok, I fixed it http://muse.19inch.net/~sivan/g-s-t/sivan_uid_fix.diff
[08:47] <pitti> ogra: please try to use select() before attempting to fgets()
[08:47] <ogra> pitti: what about waitpid ? and getting the childs pid ?
[08:48] <pitti> ogra: not necessary, you only have one child
[08:48] <elmo> Kamion/mdz: that URL isn't permanent btw; you guys know that right?
[08:48] <elmo> kamion/mdz: also which machines in the DC will you need to download it from.. just little?
[08:48] <Kamion> elmo: he didn't say that, although I kind of assumed that it might change
[08:48] <Kamion> elmo: just little
[08:48] <Kamion> (as far as I know)
[08:50] <pitti> ogra: the "I evade the problem" solution is just to use a temporary file
[08:50] <pitti> ogra: or just read /etc/lsb-release
[08:50] <ogra> bah
[08:51] <pitti> ogra: the other solution is to use select with a reasonably long timeout before the first fgets()
[08:51] <pitti> ogra: and then select with a short timeout
[08:51] <elmo> Kamion: okay, done little, you'll need to let me know if you need it for any others
[08:51] <pitti> ogra: and stop reading from the pipe (and do wait()) if select times out
[08:51] <ogra> pitti: i like to solve problems, not to work around them ... if possible ;)
[08:51] <pitti> ogra: that worked for me, dunno if there is a better method
[08:52] <pitti> ogra: google may help, LDP has a nice book where IPC is explained
[08:52] <ogra> pitti: i'll try it...
[08:53] <ogra> pitti: i'm fine with man and info ....
[08:53] <pitti> ogra: http://www.tldp.org/LDP/tlk/ipc/ipc.html
[08:53] <pitti> ogra: bah, that's the wrong one, sorry
[08:54] <Kamion> elmo: OK, it's not something I'm likely to manage to get done until about next week anyway I suspect
[08:54] <ogra> pitti: but nice to have a list ofthe signals handy, thanks ;)
[08:54] <pitti> ogra: http://www.tldp.org/LDP/lpg/node7.html#SECTION00700000000000000000
[08:54] <pitti> ogra: wrt signals: signal(7)
[08:57] <mdz> elmo: what URL?
[08:57] <pitti> ogra: rather than fiddling with select(), can you please try to check feof before?
[08:57] <mdz> amu: we do not yet, but we will
[08:57] <ogra> pitti: yup
[08:57] <pitti> ogra: while (!feof(fd)) { fgets(...); ... }
[08:58] <pitti> ogra: btw, if anything helps, just use popen
[08:58] <ogra> lol
[08:58] <pitti> ogra: usually it should be avoided, but since you don't supply any variable parameters to it, it would work
[08:59] <ogra> pitti: so i needed 2 weeks of work to get back there ? no, never
[08:59] <pitti> ogra: but please try the feof() first :-)
[08:59] <elmo> mdz: the one for the opencd tar ball
[08:59] <ogra> pitti: i will do everything, but you wont see me using popen 
[09:00] <pitti> ogra: nice :.-)
[09:00] <pitti> whoops, EBROKENSMILEY
[09:00] <ogra> heh
[09:00] <mdz> elmo: the one that henrik sent?
[09:01] <mdz> elmo: just little, yes
[09:01] <pitti> sivang: can I wait with uploading g-s-t until you fixed the sudo issue, too?
[09:01] <Treenaks> pitti: Invasion of the three-eyed smileys!
[09:01] <pitti> sivang: then we can do it in one shot
[09:01] <elmo> mdz: yeah, the reference to maitri has to go
[09:01] <pitti> mdz: thanks for fixing the hpoj script
[09:02] <pitti> mdz: nice to see it working :-)
[09:02] <mdz> pitti: with that patch, and a non-interactive postinst, it can go into supported
[09:03] <mdz> pitti: but I don't see much to be gained by putting it in desktop, since it doesn't work until you run the configuration tool :-/
[09:03] <pitti> mdz: right
[09:03] <mdz> pitti: the autodetection that it does is quite smart, it might be possible to use it non-interactively
[09:03] <sivang> pitti: sure
[09:09] <sivang> pitti_dinner: regarding commenting the code of the change, I am still checking but it seems it would also break the assigning of uid when creating a new user...so if we can do without it would probably be better.
[09:10] <ogra> pitti_dinner: if i change: if (0==fork ()) to if (fork ()) it works....
[09:17] <Kamion> hooray, a little bit more sanity inflicted upon cdimage
[09:17] <Kamion> now it actually uses germinate to generate the installer task rather than grep-dctrl to pick it out of Packages files
[09:23] <carlos> pitti_dinner: I'm going to rename the pmount's template at Rosetta from hoary-template to hoary-pmount
[09:26] <pitti_dinner> ogra: that's wrong
[09:26] <tritium> trulux, I finally got it!
[09:26] <pitti_dinner> ogra: that reverses the sense of it and the previous version was right
[09:26] <ogra> pitti_dinner: so i will resort to switch(),feof didnt work either....
[09:27] <mdz> pitti_dinner: so many firefox vulnerabilities :-(
[09:28] <tritium> trulux, /msg me for the fix
[09:29] <jdub> pants off!
[09:34] <pitti_dinner> mdz: indeed, there were three new ones; fortunately with patches
[09:34] <pitti_dinner> pitti_dinner: not switch, select :-)
[09:37] <Menaherann> holle
[09:37] <Menaherann> could you guys help me with a problem that i have with ubuntu in my laptop?
[09:37] <Menaherann> if possible/
[09:37] <Menaherann> ?
[09:38] <zul> have you tried on #ubuntu first?
[09:39] <trulux> jdub: I dislike such jokes ;)
[09:39] <Menaherann> yep but they haven't deliverd any results
[09:42] <Menaherann> i don't have to much time, i just wanted t see if i could get some help here, so i can come bak letr and then deal with it....
[09:42] <Menaherann> [gotta work] 
[09:43] <mdz> lamont: ping
[09:46] <lamont> ack
[09:49] <mdz> lamont: cloop build changes
[09:50] <mdz> lamont: there is an ubuntu-live metapackage in the queue now, headed for NEW
[09:50] <Kamion> is the live seed actually sane?
[09:50] <lamont> mdz: and livecd should have base+desktop+live?
[09:50] <mdz> Kamion: jdub and I disagree a bit on this particular point :-)
[09:50] <mdz> lamont: yes, note that the live seed will depend on language-support-en
[09:50] <mdz> lamont: so you should be able to remove the locale.gen hack
[09:51] <jdub> we don't disagree
[09:51] <jdub> but it is up for discussion :)
[09:51] <lamont> mdz: ok - want me to remove that in this round?
[09:51] <mdz> so far, language-support-en is the only thing which makes obvious sense to me for the live seed
[09:51] <mdz> lamont: once ubuntu-live is up, yes
[09:52] <mdz> lamont: i.e., pthread_cond_wait(&elmo)
[09:52] <Kamion> haha
[09:53] <Kamion> gah, why is baz ignoring the cacherev in seeds--hoary--0--patch-10 and merrily patching away from base-0
[09:54] <Kamion> oh, germinate totally ignores the live seed at the moment
[09:54] <Kamion> I should fix that
[09:54] <jdub> when are we going to start stripping mo files?
[09:56] <zul> kablooie
 the current live seed is fairly huge
 gnumeric and abiword are over 10M each
 jdub: isn't the "show off abiword and gnumeric" use case better satisfied by luis' live CD?
 I'm inclined to remove everything but language-support-en for now, and have a serious discussion about what extra stuff should go on the live CD
[09:58] <amu> elmo: could you give the kubuntu-meta your go into
[09:59] <jdub> mdz: ok
[09:59] <mdz> amu: elmo already gets mail about NEW automatically
[09:59] <jdub> mdz: btw, can normal humans who usually just play in the cloop pool change the isolinux splash easily?
[10:00] <mdz> jdub: yes
[10:00] <mdz> jdub: isolinux/splash.rle
[10:00] <jdub> boh
[10:00] <jdub> i even saw taht the other day
[10:00] <jdub> *bong*
[10:00] <jdub> thanks
[10:01] <amu> mdz: right, NEW have an average lag of 1-2 days 
[10:01] <jdub> mdz: bmp in rle format?
[10:01] <mdz> /mnt/isolinux/splash.rle: Syslinux SLL16 image data, 639 x 320
[10:02] <jdub> oh ok, i gather there'll be a converter in syslinux
[10:02] <Kamion> see the debian-installer source package
[10:02] <mdz> jdub: if you do this experiment, please add instructions to the wiki howto
[10:02] <Kamion> build/boot/x86/pics/
[10:02] <jdub> mdz: i'll encourage luis to ;)
[10:02] <Kamion> there are various README files there with conversion rules from .png
[10:03] <Kamion> (etc.)
[10:03] <jdub> Kamion: ah, ok
[10:03] <luis_> mdz: I will, thanks
[10:03] <mdz> luis_: excellent, thanks
[10:03] <luis_> mdz: are you who i'd talk to about the language selections?
[10:04] <mdz> luis_: which language selections?
[10:04] <luis_> mdz: someone asked today about a dutch-language liveCD
[10:04] <jdub> mdz: back to above; aren't we going to put more than just -en on the livecd?
[10:04] <mdz> jdub: probably
[10:05] <mdz> jdub: we're going to hit space limitations very soon though
[10:05] <luis_> mdz: I have no answer for them, but I was curious if you or anyone else had given thought to fleshing out the language TODO in the wiki
[10:05] <mdz> depending on how much win-foss we add
[10:05] <jdub> mdz: surely that's because we're not stripping mo files from the original packages?
[10:05] <amu> luis_: see http://sweb.cz/Frantisek.Rysanek/splash/isolinux-splash-HOWTO.html
[10:05] <jdub>   1. convert klowner.png klowner.bmp
[10:05] <jdub>   2. bmptoppm < klowner.bmp | ppmtolss16 #FBFDFA=7 > klowner.rle
[10:05] <mdz> jdub: partly
[10:05] <luis_> amu: ooh, awesome, thanks
[10:05] <jdub> ^ the # colour there is the background
[10:06] <mdz> jdub: remember that language packs have all of main in them, and we also would need to install language-support-XX
[10:06] <jdub> mdz: that's going to require a full rebuild...
[10:07] <Kamion> d'oh. apparently the answer is "just totally ignores it"
[10:07] <jdub> mdz: i'd hazard a guess that BigTen of main would be smaller than mo files in all livecd packages
[10:07] <mdz> Kamion: is there much point in germinating the live seed, since the live image builder doesn't use germinate?
[10:07] <Kamion> mdz: curiosity :-)
[10:07] <Kamion> mdz: also it's useful to give you size calculations
[10:08] <jdub> it's a good check to make sure live is a subset of supported
[10:08] <Kamion> that too
[10:08] <Kamion> I thought I did most of the work required to support this earlier this afternoon, but apparently I need to do some more to support disjoint seeds
[10:09] <Kamion> or, well, two seeds neither of which inherit from the other but which may expand to contain the same packages
[10:09] <Kamion> but I want to do this anyway because it makes germinate more general, which is a good thing
[10:10] <jdub> yes, the launchpad guys can look over your shoulder when they do it ;)
[10:10] <mdz> hmm, forgot to add ubuntu-live to the live seed, adding now
[10:11] <mdz> jdub: that's an easy hypothesis to test
[10:11] <Kamion> jdub: ... might have been a result of reviewing something the launchpad guys were doing, yes ...
[10:12] <mdz> 74M     usr/share/locale
[10:12] <mdz> that's in the live environment
[10:12] <jdub> Kamion: *blink*
[10:12] <Kamion> # Keep this list topologically sorted with respect to SEEDINHERIT.
[10:12] <mdz> jdub: add up BigTen Installed-Size and compare to that
[10:13] <Kamion> I love writing comments like that; it gives me wonderful schadenfreude
[10:13] <amu> :)
[10:14] <jdub> mdz: hrm, ends up being around 96
[10:14] <jdub> d'oh
[10:14] <mdz> jdub: the language packs have, e.g. gcc translations in them
[10:15] <mdz> speaking of which...build-essential in the live seed?
[10:15] <jdub> yeah!
[10:15] <Kamion> make germinate CRY
[10:15] <jdub> one problem with having bigten installed by default
[10:15] <jdub> (install, not live)
[10:15] <jdub> is that locales regeneration is *insane*
[10:15] <jdub> language-pack-en is bad enough ;)
[10:16] <Kamion> -en only has about forty-seven locales, you wuss
[10:18] <mdz> Kamion: does that make germinate cry any more than language-pack-en?
[10:18] <mdz> er, s/pack/support/
[10:18] <mdz> both l-s-en and build-essential are in ship already
[10:19] <Kamion> mdz: no (and don't worry about it, anyway, the bug needs to be fixed and good test cases won't do any harm)
[10:19] <luis_> hrm
[10:20] <luis_> jdub: any opinions on the advisability of modifying system-wide defaults v. putting things (like a .gconf dir) into /etc/skel/ for the livecd?
[10:20] <mdz> Kamion: does germinate-output/hoary obsolete germinate-hoary-output?
[10:20] <jdub> luis_: system wide defaults is cleaner and easier to understand later
[10:21] <luis_> that was my initial thought
[10:21] <luis_> but it seems like maintenance-wise, it'll be a bigger hassle
[10:21] <mdz> luis_: also consider that things in /etc/skel cost you memory and boot time when they are copied into the user's homedir, while system-wide defaults don't
[10:22] <Kamion> mdz: yes, germinate-hoary-output is a compatibility symlink to germinate-output/hoary
[10:22] <Kamion> I got fed up of the germinate-%s-output naming scheme
[10:22] <luis_> mdz: ah, good point
[10:22] <mdz> Kamion: so it would be safe to update ubuntu-meta, debzilla, and whatever other places use that
[10:22] <luis_> mdz: though at this point I've got like 20M of data files in /etc/skel/ so defaults probably don't matter too much
[10:22] <mdz> Kamion: (if you know others, tell me, I'm making a list)
[10:23] <mdz> luis_: eek!
[10:23] <jdub> luis_: (you should so symlink those)
[10:23] <luis_> jdub: ooh
[10:23] <Kamion> mdz: er, I didn't know anything automatic used germinate-output!
[10:23] <Kamion> mdz: nothing should, since that output is only for i386
[10:23] <luis_> jdub: you are wise beyond your years
[10:23] <Kamion> mdz: ubuntu-meta only seems to use the seeds
[10:23] <jdub> luis_: s/years/ears/g
[10:23] <luis_> (though really I haven't noticed that much startup penalty)
[10:23] <mdz> luis_: it is also perfectly valid to create /home/ubuntu on the cloop image
[10:24] <luis_> jdub: also, mdz is wiser than you are
[10:24] <mdz> that whole /etc/skel thing is pretty crack, actuaally, and I had been meaning to remove it from the howto
[10:24] <amu> luis_: heh
[10:24] <mdz> Kamion: ah, right; debzilla does use it though
[10:24] <mdz> Kamion: and so its ideas about non-i386 are probably skewed
[10:24] <luis_> amu- so, I've been really, really busy and totally focused on just getting this done for LWE
[10:25] <luis_> amu: after I've burned it I'll write you an email describing everything I've done and we'll see how we can go about getting it merged into gnoppix, or whether or not that makes sense even
[10:25] <Kamion> mdz: eek, OK, should I give you multi-architecture output?
[10:25] <mdz> luis_: Kamion is the person to talk to about setting the default language
[10:25] <Kamion> mdz: if so, leave the link alone for now and I'll make it keep pointing to i386
[10:25] <mdz> Kamion: maybe; doesn't seem particularly urgent, though
[10:26] <Kamion> booting with preseed/locale=<locale> should set the language
[10:26] <Kamion> assuming I haven't broken localechooser recently :)
[10:26] <mdz> is there a way to read preseed stuff from a file on the CD as well?
[10:26] <mdz> "create /path/to/foo" is preferable to "edit isolinux/isolinux.cfg, add this parameter"
[10:27] <mdz> for cut-and-pasteability
[10:27] <seb128> elmo: have you synced evince this afternoon ?
[10:27] <Kamion> yes, put the file on the CD and boot with preseed/file=/cdrom/foo/bar/baz
[10:27] <jdub> luis_: in future, our configuration changes will be defaults files
[10:27] <jdub> luis_: so that'll make life easier
[10:27] <mdz> Kamion: bleah ;-)
[10:27] <amu> luis_: cool, if we can merge, it will make the things easier ;)  
[10:28] <Kamion> the file should look like 'd-i preseed/locale string en_GB.UTF-8' or similar
[10:28] <amu> luis_: guess you didnt build debs?
[10:28] <Kamion> mdz: consider that you can put preseed/file in isolinux.cfg once and never change it again
[10:28] <Kamion> as in, *we* can put preseed/file in isolinux.cfg
[10:28] <mdz> Kamion: does it DTRT if the file is missing?
[10:28] <sri> hi all
[10:29] <Kamion> mdz: think so; if not we can always make the file be present but blank
[10:29] <luis_> amu: no, I've never built a deb in my life and this week didn't seem a good time to start :)
[10:29] <sri> following up on my problem I've been having with my X cursor that I started having last week.
[10:29] <mdz> Kamion: does kickstart by any chance have a parameter which specifies the default system runlevel?
[10:29] <mdz> I implemented that in casper recently, and it deserves to be more generic if something else can use it
[10:29] <sri> so currently, I kind have gotten xcursor somewhat working with ximian artwork
[10:29] <amu> luis_: feel free and build rpm's alien convert them into deb *ducks*  
[10:29] <sri> however, some stuff like the icon for drag doesn't show up
[10:30] <luis_> haha
[10:30] <Kamion> mdz: s/kickstart/preseed/
[10:30] <luis_> never built an rpm either :)
[10:30] <sri> and if I set it up such that I use default it's really awful (box wtih black and transparent lines)
[10:30] <Kamion> mdz: oh, you actually do mean kickstart, sorry :)
[10:30] <sri> so I'm kinda mystified now. :/
[10:30] <Kamion> mdz: no, it doesn't
[10:30] <mdz> Kamion: I actually meant kickstart in context, as in: would you have to implement this anyway in order to support compatibility with kickstart
[10:30] <sri> any ideas?  I got the latest update from today
[10:30] <Kamion> mdz: yeah, that just occurred to me, sorry
[10:31] <mdz> ok, just another lonely casper hack then :-)
[10:31] <amu> luis_: ;-)  
[10:31] <sri> one other question, how do I get xorg to just use the built in default?
[10:32] <mdz> doko: python2.4-minimal should not be essential
[10:32] <mdz> only python-minimal
[10:32] <mdz> doko: please fix
[10:34] <mdz> hmm, probably away/asleep, will file bug
[10:34] <Kamion> should be fixed immediately, I don't think dpkg forgets the essential flag once it's seen it once?
[10:34] <Kamion> could be wrong, but
[10:35] <sivang> pitti: If I am to comment the code of the uid change, I would take out the ability to set it up even when creating a new user, there is a way in the code to check if A user is new or old, so I will use that instead, ok?
[10:36] <pitti> sivang: sure :-)
[10:37] <doko> mdz: ok
[10:38] <mdz> doko: oh, good, you're here
[10:38] <jdub> dear mister zimmerman
[10:38] <mdz> doko: I filed #6362
[10:38] <jdub> please choke esound like a baby kitten
[10:38] <Kamion> -bazaar                                    | bazaar                          | Ship seed                                | Robert Collins <robert.collins@canonical.com>                             |          275892 |            1176
[10:38] <Kamion> hmm, wonder what I broke
[10:38] <pitti> yay
[10:38] <jdub> and shove it into the garbage compactor
[10:38] <Kamion> +bazaar                                    | bazaar                          | Live seed                                | Robert Collins <robert.collins@canonical.com>                             |          275892 |            1176
[10:38] <jdub> like the vermin that it is
[10:38] <jdub> love,
[10:38] <Kamion> +bazaar                                    | bazaar                          | Live seed                                | Robert Collins <robert.collins@canonical.com>                             |          275892 |            1176
[10:38] <jdub> jdub
[10:38] <mdz> jdub: today, seriously?
[10:38] <jdub> I AM SERIOUS
[10:38] <mdz> POLYPAUDIO ARE GO?
[10:39] <jdub> yeah
[10:39] <jdub> can i add polypaudio-alsa and polyaudio-x11 to the seed?
[10:39] <mdz> Kamion: I hope that isn't true about dpkg/essential
[10:39] <jdub> and polypaudio-clients to supported?
[10:39] <mdz> jdub: wtf is polypaudio-x11?
[10:39] <jdub> load-module module-x11-bell sample=x11-bell sink=output
[10:39] <jdub> :-)
[10:39] <jdub> the machine that goes (nice) bing!
[10:40] <mdz> its slacker maintainer didn't give it a proper description :-)
[10:40] <mdz> but yeah, sure
[10:40] <sivang> jdub: yay!
[10:40] <jdub> heh
[10:40] <jdub> WHERE WE'RE GOING, WE DON'T NEED ROADS
[10:40] <sivang> jdub: heheh
[10:41] <sivang> esound is dead , long live poly audio!!
[10:41] <Kamion> jdub: yay back to the future
[10:42] <jdub> Kamion: i often use it as test material
[10:43] <Kamion> testing for what exactly?
[10:43] <jdub> Kamion: recently for testing polypaudio/alsa synchronisation and then 5.1 output :)
[10:43] <Kamion> ah :)
[10:43] <jdub> also to see if there is still a heart beating in my chest
[10:43] <pitti> Mithrandir: ping
[10:44] <jdub> mdz: (note: polypaudio-alsa by default)
[10:44] <sri> ha, fixed my prboelm.
[10:44] <sri> I rock.
[10:44] <pitti> Mithrandir: http://www.ubuntulinux.org/support/documentation/usn/usn-78-1
[10:45] <sivang> jdub: what hardware did/are you going to test 5.1 output on?
[10:45] <jdub> have tested
[10:45] <jdub> nforce2
[10:46] <sivang> cool 
[10:49] <pitti> night, guys
[10:51] <Kamion> gar, no python 2.4 in warty => I don't get to use the sets module in germinate :(
[10:52] <mvo_> mdz, jdub: can we have the isdn stuff in ShipSeedProposals in the ship seed plesae? at least pppdcapiplugin and capiutils is essential for isdn support.
[10:53] <doko> mvo_: yes please, I think they are marked as such in the Wiki
[10:53] <sivang> jdub: should probably work fine on emu10k right?
[10:53] <jdub> haven't tried
[10:54] <mjt> ok, so it seems quite alot of ppl are having probs with MD5Sum mismatches when apt-get update'ing hoary 
[10:56] <mdz> mjt: that usually means that they got Release from one mirror and Packages from another
[10:56] <mjt> "usually" does not apply in this case ;)
[10:57] <mdz> mjt: ?
[10:57] <mjt> take a look into #ubuntu
[10:57] <mjt> i *know* i specified all that stuff correctly
[10:57] <mjt> on my machine anyway
[10:57] <mdz> I didn't say that you specified it incorrectly
[10:58] <Kamion> mjt: think round-robin DNS
[10:58] <seb128> mdz: that happens here too, and I've only archive in my sources.list
[10:58] <mjt> host archive.ubuntu.com
[10:58] <mjt> archive.ubuntu.com has address 82.211.81.138
[10:58] <mjt> no round-robin DNS
[10:58] <Kamion> oh, that was round-robin until recently ...
[10:58] <mdz> yes, i thought so too
[10:59] <mdz> anyway, it's happening to me, too
[10:59] <mdz> perhaps auckland is itself out of sync
[10:59] <seb128> Kamion: I think there is an issue somewhere ... does it work for you ?
[11:00] <Kamion> seb128: I haven't noticed any problems, but I update irregularly
[11:00] <Kamion> it worked for me half an hour ago or so
[11:00] <mjt> ditto here
[11:00] <mjt> but does not work now
[11:00] <mvo_> I get md5sum mismatch here now too
[11:00] <seb128> I've the issue for like 1 hour
[11:00] <seb128> and I've updated a bunch of times
[11:01] <mjt> does it really mismatch, anyone tried to verify? ;)
[11:01] <jamin> i get errors when i run apt-get update too.  and then it tells me i should run apt-get update to fix the problems. :)
[11:01] <dilinger> thom: ooh, do it again!
[11:02] <sivang> phe, /me getting the same
[11:02] <mjt> hmm.. only 5min TTL for archive.ubuntu.com A record -- it it normal?
[11:02] <mdz> mizar:[/tmp]  grep main/binary-i386/Packages.gz Release
[11:02] <mdz>  ea462f666061853e4e2397429f6a9768           562705 main/binary-i386/Packages.gz
[11:03] <mjt> and the md5 sum of the file is? ;)
[11:03] <Kamion> yeah, fails on amd64 now too
[11:04] <mjt> btw, what's the point checksumming .gz file instead of the original?  There are many ways to gzip the file (-0..-9 etc), but original content does not change
[11:04] <mdz> 67d8d283a8c9e5e25ad5d43f06b7e36e  Packages.gz
[11:04] <mdz> mjt: if it says the md5sum doesn't match, it doesn't match
[11:04] <Kamion> mjt: .gz file is what you download, it's convenient to have that in Release
[11:04] <thom> dilinger: with pleasure
[11:05] <mdz> mjt: Release has checksums for uncompressed, gzip, and bzip2
[11:05] <dholbach> thom: how do you do that? roundhouse kicking is new to me :-)
[11:05] <Kamion> dholbach: knee up to the side, twist while extending foot
[11:05] <mjt> re checksumming .gz: imagine that same apt-proxy which gets the file using rsync and gzips it after - it will not work
[11:06] <mjt> s/work/match/; s/it/checksum/
[11:06] <dholbach> Kamion, thom: wow... you two are doing martial arts, right? ;-)
[11:06] <Kamion> roundhouse kicking is not a concept entirely exclusive to martial arts :)
[11:06] <thom> only on firefox
[11:06] <mdz> mjt: that would be an example of a broken proxy
[11:07] <jdub> dholbach: Kamion is irish, he learned this at the pub when only a bairn.
[11:07] <dholbach> Kamion: well it's no known concept where i live :-)
[11:07] <Kamion> anyway I think of a roundhouse kick as mawashigeri ;)
[11:08] <doko> elmo: any chance to get libmpr-dev into main? gcc-4.0 builds are currently broken.
[11:08] <dholbach> Kamion: you do karate! :-)
[11:08] <Kamion> dholbach: yeah
[11:08] <Kamion> although nowhere remotely close to, say, lamont's standard
[11:09] <seb128> new control-center coming which handles xmodmap files, I think it'll make some people happy :p
[11:09] <jdub> seb128: ooh
[11:09] <dholbach> you guys seem very destructive to me in general
[11:09] <jdub> seb128: more crazy X keyboard bugs to chomp on!
[11:09] <seb128> :)
[11:09] <dholbach> roundhouse-drop-whatever--kicking, choking-kittens - that's not nice at all
[11:10] <dholbach> seb128: go ahead :-)
[11:10] <seb128> uploaded it this afternoon to debian and asked for a sync to elmo, but no sync for the moment
[11:10] <sivang> seb128: GO! GO! GO! ;-)
[11:10] <Kamion> dholbach: broken software does this to you
[11:11] <dholbach> Kamion: i must have found other ... compensation
[11:11] <dholbach> s
[11:11] <Kamion> argh, extra complication in germinate
[11:11] <seb128> thom: stop kicking firefox and fix the cursor :)
[11:11] <Kamion> so, when it finds a package needed in, say, the ship seed that's currently in, say, supported, it promotes it
[11:12] <Kamion> but if that package is needed in the live seed too, now there's nowhere from which to promote it
[11:12] <Kamion> I think I need to have it leave a shadow or something
[11:12] <lamont> Kamion: how about supported->ship->live->desktop->base?
[11:13] <lamont> that is, force ship to be a superset of live
[11:13] <Kamion> lamont: could do, but it's semantically wrong and I want the flexibility anyway ...
[11:13] <lamont> ah,ok
[11:14] <lamont> so supported->{live,ship}->desktop->base?
[11:14] <lamont> to continue my sloppy syntax
[11:15] <Kamion> lamont: yeah
[11:15] <Kamion> SEEDINHERIT = {
[11:15] <Kamion>     'desktop':          ['base'] ,
[11:15] <Kamion>     'ship':             ['base', 'desktop'] ,
[11:15] <Kamion>     'live':             ['base', 'desktop'] ,
[11:15] <Kamion>     'supported':        ['base', 'desktop', 'ship', 'live'] ,
[11:15] <Kamion> }
[11:22] <sivang> hrm, nice => http://nat.org/2005/february/#9-February-2005
[11:23] <lamont> Kamion: cool
[11:28] <sladen_> lamont: do you know if the kexec stuff is packaged?
[11:29] <jdub> amu: around?
[11:31] <amu> jdub: yep
[11:32] <sivang> jdub: so gnomeui is considered as dead ;-) http://developer.gnome.org/doc/API/2.0/libgnomeui/gnomeapp.html ?
[11:33] <ajmitch_> morning
[11:33] <jdub> sivang: it'll be deprecated soon; why?
[11:33] <sivang> jdub: ah, that link is broken is it intentional or just an error?
[11:34] <jdub> oh, dunno
[11:34] <jdub> 404?
[11:34] <jdub> probably error
[11:34] <jdub> http://developer.gnome.org/doc/API/2.0/libgnomeui/
[11:34] <sivang> yeah, ok never mind
[11:35] <shaya> uh, anyone else get MD5 errors when they try to apt-get update?
[11:35] <shaya> Failed to fetch http://archive.ubuntu.com/ubuntu/dists/hoary/main/binary-i386/Packages.gz  MD5Sum mismatch
[11:35] <shaya> Failed to fetch http://archive.ubuntu.com/ubuntu/dists/hoary/universe/binary-i386/Packages.gz  MD5Sum mismatch
[11:35] <shaya> Failed to fetch http://archive.ubuntu.com/ubuntu/dists/hoary/main/source/Sources.gz  MD5Sum mismatch
[11:37] <jaco> shaya it don't works this evening i think
[11:39] <lup|gnometogtk> jdub, then this http://bugzilla.gnome.org/show_bug.cgi?id=165804 will have to be integrated in gtk ?
[11:41] <jdub> lup|gnometogtk: lots of stuff will be
[11:41] <jdub> the migration has already begun :)
[11:42] <shaya> jaco: isn't it a bit beyond evening for you?
[11:42] <seb128> jdub: should we drop the network icon from computer:/// (#2302) ?
[11:43] <jdub> yes :)
[11:43] <seb128> right
[11:43] <seb128> thanks :)
[11:43] <Mithrandir> seb128: new gaim uploaded
[11:43] <seb128> cool
[11:43] <jdub> Mithrandir: dude
[11:43] <shaya> seb128: evolution is usable again :)
[11:43] <jdub> Mithrandir: so i was thinking
[11:43] <seb128> shaya: amd64 ?
[11:43] <shaya> imap on any platform
[11:43] <seb128> oh that
[11:43] <jdub> Mithrandir: how hard would it be to add an "accept mails signed by keys in this keyring" option to mailman?
[11:44] <shaya> I was getting used to thunderbird
[11:44] <jdub> Mithrandir: that could handle mime and inline sigs?
[11:44] <shaya> evolution needs an rss reader integrated with it
[11:44] <Mithrandir> jdub: easy.
[11:44] <jdub> shaya: blam
[11:44] <seb128> shaya: what bug was that ?
[11:44] <jdub> Mithrandir: i would love that feature :-)
[11:44] <Mithrandir> jdub: give me some beer and a couple of hours and you'll have it. :P
[11:44] <jdub> Mithrandir: you coming to .au? ;)
[11:44] <shaya> jdub: I actually like thunderbird's better.  blam only has the feed, it doesn't remember all entries
[11:45] <Mithrandir> jdub: 'course.
[11:45] <jdub> shaya: oh? sucky
[11:45] <jdub> shaya: use Planet :-)
[11:45] <jdub> Mithrandir: rock, beer's on me :-)
[11:45] <shaya> planet doesn't remember what you read or want to read again
[11:45] <shaya> i like the read/unread interface
[11:45] <jdub> heh
[11:45] <jdub> PICKY!
[11:45] <jdub> tried straw?
[11:45] <shaya> no
[11:46] <shaya> doesn't seem to be in ubuntu
[11:46] <Mithrandir> straw scrolls the item window to the top when the feed is refreshed.
[11:46] <thom> i like the "not sucking" part of planet, which is somewhat lacking from every other linux rss reader
[11:46] <mdz> Kamion: mind if I make the apt-setup template change I just suggested on the list, or would you rather do it?
[11:46] <shaya> thom: thunderbird is the best one I've used
[11:46] <Mithrandir> thom: doesn't have any feed priority thingy
[11:46] <jdub> shaya: make a hep account and use rss by imap in evolution :)
[11:46] <thom> shaya: i think you're making my point for me :-)
[11:46] <Mithrandir> which means news sites which get 60-ish stories a day would dominate.
[11:47] <Mithrandir> shaya: chunderbird is kinda crackful.
[11:47] <shaya> hep account?
[11:47] <jdub> google for hep
[11:48] <shaya> hmm
[11:48] <shaya> hep is interesting
[11:48] <shaya> packaged for ubuntu?
[11:48] <jdub> no
[11:48] <jdub> you can just create a hep accountv
[11:48] <jdub> on the demo site
[11:48] <jdub> but it shouldn't be hard to use on ubuntu
[11:48] <jdub> if you want to run it locally
[11:49] <shaya> demo site?
[11:49] <Mithrandir> rss2email too; I'm considering setting it up and just dropping it in some maildir
[11:49] <shaya> anyways
[11:49] <shaya> time for a 90 minute showing of new stargate sg1 episode
[11:53] <enrico> Hello.  I did cat /etc/debian-version on Hoary and it says "3.1".  Is that intended?  And is there a way to detect what is the ubuntu version?
[11:54] <dholbach> enrico: cat /etc/lsb-release
[11:54] <enrico> dholbach: thanks!
[11:55] <dholbach> enrico:    grep DISTRIB_RELEASE /etc/lsb-release | sed 's/DISTRIB_RELEASE=//g'   :-)
[11:55] <enrico> oh.  Debian doesn't have /etc/lsb-release...
[11:59] <jdub> jordi: when they ask "who is your daddy?" the correct answer is "mako"
[11:59] <jordi> jdub: :)