[12:14] <ogra> no luck for me with the daily livecd on amd64 :`(
[12:18] <calc> anyone here happen to use SOAPpy?
[12:20] <calc> (python-soappy) seems not to work when return type is float, even the examples given on its homepage don't work :\
[12:58] <mdz> calc: I haven't, no
[01:18] <jdub> http://people.ubuntu.com/~jdub/planet/
[01:18] <jdub> el commento scorchio?
[01:21] <Riddell> jdub: why the white background?
[01:22] <jdub> so it doesn't look like someone weed on it
[01:22] <jdub> also, it generally looks better (people kinda assume)
[01:23] <Riddell> jdub: it might look better but it's inconsistent with all the other ubuntu/canonical websites
[01:23] <jdub> yep
[01:23] <jdub> i've left the rest of the chrome in the same colour
[01:24] <daniels> jdub: scorchio!
[01:24] <jdub> it's a simple change if people really deliriously object
[01:24] <daniels> jdub: not such a fan of the yellow up top, but shit, it's come a long way
[01:25] <jdub> website redesign contest announce going out soon ;)
[01:25] <mdz> jdub: what is up with the scaling in the multiload applet?
[01:25] <daniels> WORD
[01:25] <jdub> mdz: scaling?
[01:25] <mdz> what it appears to be doing is considering the highest sample in the current period to be the maximum
[01:25] <mdz> and scaling the graph relative to that
[01:25] <jdub> haha
[01:25] <mdz> which means that if there is any activity at all, it looks like it is maxed
[01:26] <jdub> for which graph?
[01:26] <mdz> all of them
[01:26] <jdub> doesn't seem to do that for cpu
[01:26] <jdub> arch?
[01:26] <mdz> maybe not for cpu, dunno
[01:26] <mdz> looking at network and I/O
[01:26] <mdz> i386
[01:26] <mdz> right now I am downloading a file at about 150kbyte/sec
[01:26] <mdz> over 100mbit ethernet
[01:26] <mdz> and the network graph is full
[01:27] <mdz> I think a smarter idea would be to remember the highest value ever seen
[01:27] <jdub> mdz: it has detected that you're using low quality cabling
[01:27] <mdz> and treat that as max, where no authoritative value is available
[01:27] <jdub> yeah
[01:28] <jdub> pinging davyd, who is probably not awake
[02:00] <jdub> http://weblog.zerokspot.com/posts/164/
[02:23] <robertj> are the autobuilders up yet?
[02:24] <lamont> robertj: up is such a relative term.
[02:24] <lamont> 3 of them were actually up and running (out of 12)...  I'm rebuilding chroots on all of them right now, spring cleaning and all that
[02:25] <robertj> ahh, do you know if there is anything in the que for xorg and friends
[02:25] <robertj> although I'm not unsure it's not pangos fault, I need to check on that
[02:26] <robertj> great results though, like every font in every gtk app was like 1 point  ;)
[02:30] <lamont> haven't looked - see people.ubuntu.com/~lamont/buildLogs/Lists/hoary.all.{i386,amd64,powerpc,ia64}
[02:30] <lamont> if it says 'Needs-Build', then it's queued
[02:36] <robertj> wow, that's quite a list
[02:56] <lamont> hrm.. wonder if thom is awake
[02:56] <jdub> o/~ all this time in litigation, so much for darl's happy ending o/!
[02:57] <mjg59> Is the canonical conference before or after LCA?
[02:57] <lamont> mjg59: I expect after, dunno
[02:57] <lamont> it's certainly after the release
[02:58] <jdub> after
[02:58] <jdub> mjg59: dates will be announced next week
[02:58] <jdub> mjg59: it'll basically be 25th-30th
[02:59] <mjg59> Hm. I might actually be able to make that, possibly.
[02:59] <mjg59> jdub: When are you getting married, then? :)
[03:00] <jdub> 17th
[03:00] <jdub> 17 18 19 20 21 22 23
[03:00] <jdub> 24 25 26 27 28 29 30
[03:00] <jdub> 17 == married
[03:00] <mjg59> Haha
[03:00] <jdub> 18-23 == linux.conf.au
[03:00] <jdub> 25-30 == ubuntudownunder
[03:00] <mjg59> I have to teach on the 2nd May
[03:00] <jdub> 1 == FUCK OFF
[03:00] <lamont> jdub: pia is gonna kill you, you know...
[03:01] <mjg59> But from the end of March until then, I'm free
[03:01] <jdub> lamont: i prepped her for it a long while back :)
[03:01] <mjg59> So I may actually be able to join you guys
[03:03] <jdub> that'd be rad
[03:03] <jdub> would love to see you at desktop.conf.au too :)
[03:03] <sladen> mjg59: you asking aswell?
[03:03] <mjg59> sladen: Mm?
[03:04] <mjg59> jdub: Well, when there's some better idea as to what's going on, I'll let you know what's possible
[03:04] <sladen> mjg59: same question I asked jdub, only a couple of hours ago :)
[03:05] <mjg59> sladen: Pff. You expect me to analyse history?
[03:05] <sivang> jdub: there are 2 gnome confs a year? the one in germany and the one in .au? and so close by date.. :)
[03:06] <jdub> sivang: there are many
[03:06] <jdub> there's GUADEC (which roams around europe, and is the biggest)
[03:06] <jdub> there's the summit, which is based in boston
[03:06] <jdub> there's a spanish one
[03:06] <jdub> there's gnome.conf.au
[03:06] <jdub> there's one in chile
[03:06] <sivang> jdub: which one is the best to attend ? :-) (I think I can predict the answer from you)
[03:06] <sladen> mjg59: it was a /privmsg to be fare 
[03:06] <mjg59> Guadec is probably the single most important
[03:07] <jdub> if ubuntu doesn't have more mini-confs than gnome within a year, i'm going to buy a jet and do them myself :)
[03:07] <jdub> sivang: guadec
[03:07] <sladen> GUADEC is 28th-30th May in Stuttgard
[03:07] <jdub> pretty much everyone goes to guadec
[03:07] <mjg59> The summit probably gets more done, but guadec shapes what's going to be done significantly more
[03:07] <jdub> and there's more fanboys
[03:07] <mjg59> Yeah, you get to meet and fanboy rml
[03:08] <sivang> mjg59: fanboy rml?
[03:08] <sladen> mjg59: and everyone gets to meet and fanboy you
[03:08] <sladen> sivang: you can rob him of his love
[03:09] <mjg59> sivang: rml deserves mad fanboying
[03:10] <sivang> who is rml? :-) 
[03:10] <jdub> rob love, kernel hackeur
[03:12] <kent> hmm,  is there some way of getting intouch with some one from Stuttgart who are going to GUADEC and could perhaps let some other, like me for example, live at his/her's place over the event? I have never been at such an event, but i think it would be fun..
[03:14] <sladen> kent: you have similar plans to me, but this is a good 4 months away yet...
[03:14] <sivang> jdub: ah yeah, just read about him in the speakers list :-) quite a guy
[03:15] <kent> sladen, yeah.. and i'm very good at making plans in the middle of the night, which i forget the next day. But i think i will write this down. I looked at the internet, and if i only have to pay for the tickets to get there, and some for the food, then i think i can afford it.
[03:15] <jdub> kent: jump on guadec-list and ask
[03:28] <Riddell> I like how this year's guadec is in much the same place as akademy was last year
[03:29] <sladen> Riddell: wouldn't want to risk holding a major conference somewhere without letting some minor, dispenable conference test the waters first
[03:30] <Riddell> sladen: you know that tone of voice in #ubuntu would get you a telling off from jdub  :)
[03:31] <jdub> akademy was bloody huge dude
[03:31] <jdub> and very long
[03:42] <Riddell> yes, too long one conference would have been enough, but good fun
[03:43] <Riddell> guadec misses a hacking session though which surprises me, you can't do that much in a weekend
[03:48] <sivang> is everything ok with bugzilla? won't let me in....
[03:49] <sivang> eh ok, cache refreshment did it.
[03:51] <sladen> sivang: that  /ubunturocks  spells  ubuntu  ubutu
[03:53] <sladen> sivang: might also be worth saying  ''This is what Ars said:'' as somebody has mentioned elsewhere as it could be seen as Ubuntu slagging off Debian by appearing to agree  ...which would be unlike Ubuntu
[04:03] <sladen> jdub: can I patch 'passwd root'  to say:   ''I'm sorry Dave...  I can't let you do that.  It's for your own safety you know''
[04:04] <jdub> :-)
[04:05] <mdz> jdub: is anyone working on adding the capabilty to configure which sound device should be primary in the desktop?
[04:05] <mdz> i.e., which one esd/polyp/etc. will use?
[04:05] <amu> hmmm is there's something wrong with katie? 
[04:06] <jdub> mdz: not atm
[04:06] <mdz> sladen: is the --rsyncable stuff really that simple?
[04:06] <jdub> amu: elmo said it was b0rked up a bit earlier
[04:06] <mdz> sladen: I looked at the code once, and it wasn't immediately obvious what it was doing
[04:06] <sladen> mdz:  uh huh
[04:07] <mdz> sladen: on what did you base the 0.07% penalty estimate?
[04:07] <sladen> mdz: the main bit is a counter that counts down to the next sync point and ensures and any string that is created in the output is capped at the next sync point and a new string started again
[04:08] <amu> jdub: yeah i thought it was fixed a bit later, looks not ... thx
[04:08] <sladen> mdz: somebody from Debian quotes somehwere that it works out as 0.7% on most packages and 1.2% on ''some of our core packages''
[04:09] <sladen> mdz: on the live CD, the sync point could be set every 1MB and it'd still probably make a major difference
[04:09] <mdz> sladen: hard to say, because the filesystem inside might allocate things differently from one run to the next
[04:09] <mdz> what I was thinking was that it could sync every block
[04:09] <sladen> mdz: I chatted to broonie earlier about what might get upstream and how to allow variable sync windows
[04:09] <mdz> since it's compressing blockwise anyway
[04:10] <mdz> and if we set fs block size = compression block size, and sync at every block, we should do very well
[04:11] <sladen> yup.  The other thing I forgot to mention, is running touch over the whole FS to force  ctime/atime to a known value
[04:12] <sladen> since those will change with each unpack/build
[04:12] <mdz> hmm
[04:12] <mdz> the entire inode would need to be identical for it to be a win
[04:13] <mdz> size, atime, ctime, mtime, permissions and ownership...
[04:13] <mdz> could be other filesystem-specific data in that block as well
[04:13] <mdz> reiser at least will pack the data for tiny files in the same block with the inode
[04:13] <mdz> though I don't think ext2 does anything like that
[04:14] <elmo> it doesn't
[04:14] <mdz> I think that's worthwhile to try, though
[04:14] <mdz> having an entire filesystem with the same inode times will look a bit weird, but I don't imagine it would break anything
[04:21] <mdz> elmo: did you find Riddell's bug?
[04:24] <elmo> actually, there's no bug in katie, the reject message is very scary, but all accurate.. 
[04:24] <jdub> elmo: calc mentioned the archive md5sums were wrong
[04:24] <elmo> jdub: err, seems unlikely?
[04:24] <jdub> dunno
[04:25] <elmo> got any details, I don't have a client logging at home anymore
[04:25] <jdub> calc: ping
[04:25] <calc> jdub: pong
[04:25] <calc> yea i ran apt-get update and it was failing
[04:26] <calc> it works now
[04:26] <calc> so not sure what changed, unless someone recently fixed it
[04:26] <calc> btw anyone know how to tell a python module to install under /usr/local ?
[04:26] <elmo> what's "failing", and where are your sources.list's pointing?
[04:28] <calc> it was hoary universe that had bad md5sum, but it works fine now
[04:28] <calc>  deb http://archive.ubuntu.com/ubuntu/ hoary universe
[04:28] <calc> that one afaik
[04:29] <elmo> well, please capture the error message next time - but nothing's changed on a.u.c, so I'm guessing it was a transient network thing
[04:30] <elmo> bah, silly fascist hotel starts dropping NEW ports if you don't make port 80 requests every so often
[04:31] <calc> ok
[04:39] <elmo> night all
[04:43] <calc> cool i found out why SOAPpy doesn't work for me
[04:43] <calc> its too darn old and needs to be updated
[04:43] <calc> 0.11.3 is in hoary and doesn't work but 0.11.6 works fine
[04:50] <robertj> I'm dangerously close to building my first .deb, kinda scary really
[04:51] <calc> robertj: read all the docs already? :)
[04:52] <tritium> robertj, awesome :)
[04:53] <robertj> calc: yeah
[04:53] <robertj> ignoring most of the stuff for now though ;)
[04:54] <tritium> robertj, new maintainer, I presume?  How long did the process take?
[04:55] <robertj> trit: nahh, just on my own
[04:57] <tritium> robertj, cool
[04:57] <robertj> pre and postinst dont seem to be firing yet though
[05:09] <lamont> Kamion: let me know if there were any issues with livecd/di (other than you needing to change the name of the host you fetch from...)
[05:13] <robertj> is there anything special that needs doing to get the scripts to run? Running by hand yields the expected results
[05:16] <robertj> set -e
[05:16] <robertj> gtk-query-immodules-2.0 > /etc/gtk-2.0/gtk.immodules
[05:16] <robertj> exit 0
[05:16] <robertj> that's the extent of it, but nothing hapening unless I run them by hand for preinst and postinst
[06:42] <thully_> Is there currently a recommended snapshot CD to be used for installing Hoary?
[06:47] <thully> I lost my connection - but I was just wondering if anyone knows a good Ubuntu Hoary snapshot to use for installation
[07:24] <bob2> Setting up debsums (2.0.13) ...
[07:24] <bob2> /var/lib/dpkg/info/debsums.config:5: local: not found
[07:39] <bob2> HAH
[07:39] <bob2> it *is* lamont's fault
[07:39] <bob2> I changed /bin/sh to be posh
[07:40] <bob2> and setting it back to bash fixes it
[08:05] <thully> hi - is Ubuntu going to handle the Mozilla trademark issue in the same way Debian eventually does?
[08:07] <froud> hi, does ubuntu support Ipv6?
[08:07] <bob2> yes
[08:07] <froud> thanks
[08:09] <bob2> hah, nevow is in supported
[08:11] <froud> bob2, ? what
[08:28] <bob2> is someone in charge of making the mono toolchain work on ubuntu atm?
[08:37] <fabbione> mornign
[08:42] <fabbione> lamont: ping
[09:26] <bob2> lamont: is it bad if I'm getting a ton of these: postdrop: warning: mail_queue_enter: create file maildrop/173496.8172: Permission denied
[09:32] <jdub> thom: ping
[09:36] <bob2> lamont: I found http://lists.debian.org/debian-user/2000/12/msg03524.html, but it doesn't seem to be the problem in this case
[09:57] <froud> is anyone maintaining a list of Release Notes for Hoary or knows where I can fin dinformation to compiled Release Notes?
[10:18] <bob2> lamont: erk, it seems was filesystem was slightly fucked
[11:17] <mdz> night
[11:24] <Kosai> Hrr.  Is it safe to remove esd?
[12:52] <T-None> Kamion: i'll have trouble giving you the content of /proc/cpuinfo if i'm unable to load a kernel ;)
[12:53] <T-Bone> Kamion: I can give you more detailed info about ia64, i have a scratch box, and ppc64 (same), but for the G4 it's gonna be more tricky: I'm not very inclined to wipe out the install I made :P
[12:56] <T-Bone> Kamion: i'll give a shot at elilo/efibootmgr package too.
[01:00] <Kamion> T-Bone: you can run through the install up to the beginning of the partitioning step; that does all the hardware detection but doesn't touch the box
[01:00] <T-Bone> good point, but the hotplug bug shows _after_ install
[01:00] <Kamion> T-Bone: also a tarball of /proc/device-tree on the G4 might be useful, along with what you needed to put in /etc/modules by hand
[01:00] <Kamion> T-Bone: even in hoary?
[01:01] <T-Bone> Kamion: i wouldn't have much to report tho: it's basically hanging
[01:01] <Kamion> hoary's installer uses hotplug
[01:01] <Kamion> warty's does not
[01:01] <T-Bone> Kamion: it doesn't happen during the install
[01:01] <T-Bone> it's after rebooting
[01:01] <Kamion> bizarre
[01:02] <T-Bone> i see "starting hotplug..." then the cursor moves 1 1/2 screen and then it hangs
[01:02] <Kamion> ok, now let's see what lamont did to debootstrap ... :)
[01:02] <T-Bone> hehe
[01:02] <T-Bone> Kamion: i'm off for breakfast, bbiab for more debug :)
[01:41] <Kamion> mdz: Argh. How do I convince apt that packages in /var/cache/apt/archives have already been authenticated?
[01:46] <daniels> Kamion: apt-ftparchive and gpg ;)
[01:51] <Kamion> daniels: ... uh-huh
[01:51] <Kamion> gpg with what key, exactly? one I just made up? I bet that'll be in ubuntu-keyring ;)
[01:52] <T-Bone> Kamion: oh and of course, it was daily from 20050115
[01:52] <ogra> hmm, daniels i got a new laptop with 1280x800 .... there seems to be no chance to convice xorg to do this res. except by specifying a modeline....
[01:52] <T-Bone> must have been on crack when writing the date
[01:53] <T-Bone> first bad new: I don't have /var/log/debian-installer
[01:53] <Kamion> lamont: uh, why does ia64 need libunwind7*-dev* in base?
[01:53] <Kamion> T-Bone: that's only there after the installation is complete
[01:53] <Kamion> during installation, it's /var/log/syslog
[01:54] <T-Bone> Kamion: i reiterate: there is no linux-image-* in /cdrom/pool/main/l/
[01:54] <Kamion> T-Bone: good.
[01:54] <Kamion> T-Bone: it's in linux-meta
[01:54] <Kamion> probably restricted
[01:54] <Kamion> find /cdrom -name linux\*
[01:54] <T-Bone> right, my bad
[01:54] <T-Bone> it's there ;)
[01:54] <Kamion> I reiterate, send me /var/log/syslog :)
[01:55] <Kamion> after base-installer has run and failed
[01:55] <T-Bone> Kamion: 'after installation' means after a successful installation, i suppose
[01:55] <Kamion> it means post-reboot after prebaseconfig has run
[01:55] <T-Bone> since it hasn't... ;)
[01:55] <Kamion> so it would probably have to be successful to some degree, yeah
[01:55] <T-Bone> ok, i'll scratch the box again and grab the files
[01:55] <Kamion> you have nc in the installer, so you can get files out
[01:55] <T-Bone> netcat, yummy ;)
[01:56] <Kamion> there's openssh-client-udeb too, but I don't want to exacerbate the anna problems ...
[01:56] <T-Bone> heh
[01:56] <daniels> Kamion: so, usefbdev on powerpc ... speak to me.
[01:56] <Kamion> daniels: I don't know what the generic answer is
[01:57] <daniels> Kamion: ok.  let's see how pinging overfiend goes :)
[01:57] <T-Bone> Kamion: i'll try the G5 with daily hoary too
[01:57] <Kamion> daniels: try Michel Daenzer
[01:57] <daniels> Kamion: good point
[01:57] <Kosai> Kamion: How's the Oqo doing?  :)
[01:57] <Kamion> T-Bone: yeah, will probably stand a better chance there; I think PPC970FX CPU support went in post-warty
[01:57] <Kamion> Kosai: installing today's daily, seeing how the new kernel goes
[01:58] <Kamion> I need support for my USB wireless thingy in order to develop on it sanely
[01:58] <T-Bone> Kamion: ok. FYI it's the top end G5 model. Dual 2.5GHz watercooled
[01:59] <T-Bone> (talk about a fine beast :)
[01:59] <Kosai> Kamion: Cool.
[02:00] <T-Bone> Kamion: before scratching the box i'll patch elilo and efibootmgr to the best of my abilities so that you can integrate it in next snapshots
[02:01] <lupus_> nick lupusBE
[02:01] <lupus_> crap :s sorry 
[02:02] <Kamion> T-Bone: ok, send me/lamont the patches
[02:02] <T-Bone> sure
[02:03] <Kamion> there's a documentation fix in the d-i manual needed as well, but I'll handle that
[02:03] <Kamion> brb, shuffling houses
[02:08] <daniels> rburton: hey dude :)
[02:08] <rburton> hi daniels 
[02:09] <trukulo> Mithrandir, u there?
[02:09] <Treenaks> trukulo: time for jeff_big ! :P
[02:09] <trukulo> :) ok
[02:09] <Mithrandir> trukulo: yes.
[02:09] <trukulo> so going to seed it
[02:09] <trukulo> it's the last one, isn't it?
[02:09] <Treenaks> yes
[02:10] <Mithrandir> blah, my bittorrent client had segfaulted. :/
[02:10] <trukulo> put it again before i seed it
[02:10] <Mithrandir> yeah, it's checking the files on disk now
[02:10] <trukulo> ok
[02:11] <Mithrandir> ok, just start seeding
[02:11] <trukulo> seeding
[02:11] <trukulo> it's checking
[02:12] <trukulo> seeding now
[02:13] <Mithrandir> yup, ETA 18-ish hours
[02:14] <ajmitch> ext2resize in hoary looks like quite an old version (same as sid) - I've got a locally updated version here that is meant to work with 2.6.9+ for online resizing
[02:14] <trukulo> perfect
[02:14] <trukulo> last one, cool
[02:14] <Mithrandir> yup, great.
[02:27] <daniels> THIS DOES NOT MAKE SENSE
[02:27] <daniels> die monolithic tree, die
[02:28] <Kamion> back
[02:28] <daniels> Kamion: represent
[02:28] <trulux> mdz: arpund?
[02:28] <trulux> around?
[02:34] <T-Bone> Kamion: who's gonna handle maintainership of Ubuntu-modified ia64 package? :)
[02:35] <bob2> it's cool when your ISP combines shaping of non-local traffic with an ubuntu mirror
[02:35] <bob2> er, "broken ubuntu mirror"
[02:36] <daniels> bob2: d'oh
[02:43] <T-Bone> Kamion: ok, on ia64, hotplug dies on /etc/hotplug/prc.rc start
[02:44] <T-Bone> Kamion: anything you want me try before wiping out the system and reinstalling from scratch?
[02:58] <jdub> hrm
[02:58] <jdub> the version of dovecot in warty is a bit pants
[03:02] <amu> daniels: you handle dbus, right?  
[03:02] <daniels> amu: yah
[03:03] <T-Bone> weeee! kernel panic during boot
[03:03] <amu> daniels: i need dbus-qt enabled :) could i send you a diff, or this is a problem cause qt isnt in main?  
[03:03] <T-Bone> in ext3
[03:04] <daniels> amu: problem 'cause qt isn't in main, but if we need to, i could do something similar to what i did for dbus-mono
[03:05] <daniels> amu: how urgently do you need it?  i can do it around wednesday if you like
[03:05] <mjg59> daniels: How's the crack?
[03:06] <Kamion> T-Bone: usually not necessary to change Maintainer:
[03:06] <Kamion> T-Bone: but, we hope, the ia64 port team ... i.e. you and others ... :)
[03:06] <T-Bone> lol
[03:06] <T-Bone> i can handle elilo i guess ;)
[03:06] <amu> daniels: that's fine for me, even better if you do it yourself ;) i started packaging kde3.4, and i really want this the dbus things enabled   
[03:07] <Kamion> T-Bone: try 'sh -x /etc/hotplug/pci.rc start'
[03:07] <bob2> daniels: how was dbus-mono handled without bringing mono into main?
[03:07] <Kamion> amu: Qt is in main
[03:07] <Kamion> libqt3-dev | 3:3.3.3-7ubuntu1 |    hoary/main | amd64, i386, ia64, powerpc
[03:07] <bob2> main = supported, right?
[03:08] <daniels> mjg59: it's crackful
[03:08] <T-Bone> Kamion: sure, if i can get that box to boot again i'll try :)
[03:08] <Kamion> bob2: yes, everything in main's supported
[03:08] <daniels> mjg59: (but it's also Sunday evening)
[03:08] <sladen> amu: riddell/haggai have been busy porting stuff.  KDE 3.4 should already be in there
[03:08] <daniels> bob2: separate source package
[03:08] <bob2> oh, that's crack
[03:08] <amu> Kamion: nice  
[03:08] <daniels> Kamion: oh, it's in main these days?  nice
[03:08] <daniels> bob2: yes, doubly so given it sets 'DESTDIR='
[03:08] <Kamion> daniels: libqt3-dev | 3:3.2.3-4ubuntu1 |    warty/main | amd64, i386, powerpc
[03:08] <daniels> which was omgwtfbbq
[03:08] <Kamion> was always in main
[03:08] <daniels> Kamion: yah
[03:08] <daniels> Kamion: really?
[03:08] <Kamion> yep
[03:09] <daniels> heh
[03:09] <daniels> jdub: was he rabbiting on about cargo bar? :)
[03:09] <Kamion> due to doxygen IIRC
[03:09] <amu> sladen: aeh?
[03:09] <bob2> hahaha
[03:09] <bob2> daniels: and scubar
[03:10] <daniels> bob2: apparently they have a machine with the getaway there, and there's a *massive* line of skanky english backpackers
[03:11] <bob2> oh man
[03:12] <bob2> that's doubly scary since some of my friends want to go there on saturday
[03:13] <bob2> mm, 308MB dist-upgrade
[03:15] <amu> is there someone alive, who could restart a sbuild on kdemultimedia-amd64 , it's a blocker, it's depends are just fixed 
[03:16] <daniels> Kamion: at least no-one's had the stupid idea of openoffice.org-amd64; that would be a horror
[03:16] <daniels> (phew!)
[03:17] <daniels> amu: did you just upload to fix its build-depends, or did it depend on broken stuff that's now been fixed?
[03:17] <Kamion> the auto-depwaiter often deals with the latter ...
[03:18] <daniels> and the former is done automatically
[03:19] <amu> daniels: it depended on broken stuff and now it is fixed, uploaded yesterday   
[03:19] <daniels> amu: so the stuff it depends on has been fixed?
[03:19] <amu> daniels: ack
[03:20] <daniels> ok, you may need to kick lamont, else it should just be taken care of
[03:21] <daniels> `kde_42-4.10ubp1_all.deb' -> `/var/apt/ubuntu/dists/warty-backports-staging/universe/binary-i386/kde_42-4.10ubp1_all.deb'
[03:21] <daniels> what sort of version is that?
[03:21] <daniels> what the hell?
[03:21] <amu> daniels: thx, dbus-qt is scheduled?  
[03:21] <daniels> _42?
[03:21] <daniels> amu: wednesday, if you don't need it earlier
[03:22] <amu> daniels: that's fine for me 
[03:23] <amu> daniels: guess this package never got some love ;) 
[03:24] <daniels> heh, not really
[03:25] <daniels> it was broken for *ages*
[03:25] <T-Bone> Kamion: side note. i don't think any ia64 system ships with PCMCIA (to be confirmed tho), so maybe PCMCIA isn't necessary in d-i
[03:27] <Kamion> just take out pcmcia-cs-udeb then
[03:28] <Kamion> it shouldn't be very important to take it out though?
[03:28] <amu> daniels: hehe
[03:28] <T-Bone> nah. It just asks a (imho) bogus question
[03:28] <daniels> (upstream)
[03:28] <Kamion> T-Bone: only in expert mode, which is itself bogus
[03:29] <Kamion> we must not ship a system that requires expert mode in order to install correctly :)
[03:30] <T-Bone> heh
[03:30] <T-Bone> Kamion: as lamont pointed, I suspect that one of the IDE modules is killing anna. I'll investigate that later, once I'll have answered all others points in your email :)
[03:31] <T-Bone> Kamion: i wonder why you need to pass the kernel boot arg to have rescue working. Selecting its udeb in the list doesn't seem to work
[03:32] <Kamion> T-Bone: long story
[03:32] <T-Bone> heh ok ;)
[03:32] <T-Bone> huh, interesting
[03:32] <T-Bone> i started in non expert mode by mistake, it's working just fine, the CDrom gets detected
[03:33] <T-Bone> it just red-screen on 'error while running 'modprobe -v amd74xx''
[03:33] <Kamion> T-Bone: I suppose I could db_input rescue/check in rescue mode, but I'm not sure I want to; the correct approach is to add bootloader config to have a separate rescue mode
[03:33] <Kamion> that's expert mode for you
[03:34] <T-Bone> Kamion: the anna issue hasn't been fixed, has it?
[03:34] <mjg59> Hm. There's definitely problems with swsusp and interrupt routing after resume at the moment.
[03:34] <Kamion> T-Bone: no
[03:35] <mjg59> Other than that, https://www.ubuntulinux.org/wiki/HoaryPMResults looks pretty promising
[03:35] <T-Bone> Kamion: then it doesn't affect my configuration
[03:35] <T-Bone> i bet that's a SCSI problem
[03:35] <T-Bone> i'll try on zx6000 in a while
[03:35] <Kamion> T-Bone: somebody needs to build a debugging version of anna, probably with extra printf() statements to figure out where it's segfaulting, put that debugging anna in an initrd, and iterate
[03:35] <Kamion> T-Bone: anna segfaulting is not a SCSI problem :-)
[03:36] <T-Bone> hmm
[03:36] <Kamion> something in anna or libdebian-installer is bogus, but the curious thing is why it doesn't affect Debian
[03:36] <T-Bone> Kamion: it's obviously hardware-dependent though
[03:36] <Kamion> or any of our other architectures
[03:36] <Kamion> T-Bone: yes but the set of udebs anna wants to install might be hardware-dependent
[03:36] <T-Bone> i wonder if i had that anna problem on my g4. I had so many problems I just can't recall what they were
[03:36] <Kamion> unlikely
[03:37] <Kamion> I've never ever seen that or heard of that on powerpc
[03:37] <Kamion> you were probably just missing a chunk in discover-mac-io
[03:37] <T-Bone> right
[03:37] <Kamion> if anna segfaulted, you'd know about it.
[03:37] <T-Bone> heh
[03:39] <T-Bone> ok so i finish debugging that hotplug issue and then get the syslogs, and then reply to your mail with a nice elilo patch
[03:40] <Kamion> some other ia64 people mentioned the kernel issue BTW
[03:41] <Kamion> unfortunately I don't quite have the base-installer tests working on Ubuntu yet
[03:43] <T-Bone> Kamion: does 'Ubuntu GNU/Linux' sounds like a good naming (I basically changed all '[Dd] ebian' strings with '[Uu] buntu')
[03:44] <Keybuk> should be just "Ubuntu"
[03:45] <rburton> risking the wrath of stallman?
[03:45] <Mithrandir> avoiding the issue altogether
[03:45] <rburton> true
[03:45] <T-Bone> Keybuk: 'Ubuntu Linux' at least, no?
[03:45] <Mithrandir> T-Bone: no, just Ubuntu
[03:45] <Keybuk> T-Bone: no, just Ubuntu
[03:46] <T-Bone> well then, so be it
[03:46] <Kamion> agreed, just Ubuntu
[03:46] <T-Bone> Mithrandir, Keybuk: same for the manpages, i suppose?
[03:46] <Kamion> T-Bone: all the rest of d-i is branded this way
[03:46] <T-Bone> Kamion: fyi, hoary iso signs itself as 'Debian Inst' in EFI
[03:46] <Kamion> T-Bone: yes, bug
[03:47] <Mithrandir> T-Bone: avoiding "Linux" means you don't get into the "Linux" vs "GNU/Linux" argument at all.
[03:47] <T-Bone> Mithrandir: ah ok
[03:47] <T-Bone> wasn't aware
[03:47] <Kamion> hm, although I can't find that "Debian Inst" string
[03:48] <Kamion> oh, there it is, whoops
[03:48] <T-Bone> Kamion: i guess that's some mkisofs configuration
[03:48] <Kamion> no, it's a mkfs.msdos parameter in debian-installer
[03:48] <T-Bone> ah ok
[03:49] <Kamion> T-Bone: fixed in my local tree, thanks
[03:50] <T-Bone> np
[03:52] <bob2> mjg59: where do all those variables ($SAVE_VBE_STATE, e.g.) get set for acpi/resume.sh?
[03:52] <mjg59> bob2: /etc/default/acpi-support
[03:53] <bob2> mjg59: it doesn't seem to source it
[03:54] <mjg59> bob2: Uh. sleep.sh and hibernate.sh source it.
[03:54] <mjg59> They source the other scripts.
[03:54] <bob2> ahhh, right, sorry
[03:55] <T-Bone> looks like that ext3 bug screwed my system, /etc/modules doesn't seem to be acknowledged
[03:55] <T-Bone> gonna reinstall and then debug hotplug
[03:56] <sladen> oooh.  SPCR ACPI table---serial port console redirection
[04:05] <T-Bone> Kamion: ok my bugreport about kernel package was incomplete
[04:05] <T-Bone> d-i looks for "kernel-image"
[04:05] <T-Bone> hence the bug
[04:06] <T-Bone> it says: "The current default kernel package is 'kernel-image'." Obviously remnant of Debian origins ;)
[04:06] <Kamion> that's only a fallback
[04:06] <Kamion> it shouldn't actually be a problem
[04:07] <Kamion> it's not a problem on other architectures, they all default to kernel-image too; and "kernel-image" without suffix generally doesn't exist on either Debian or Ubuntu anyway
[04:08] <T-Bone> true enough
[04:08] <Kamion> as I say, all the information I need to debug this is in syslog. :)
[04:08] <T-Bone> i'll send you the logs in a short while
[04:08] <Kamion> ok, thanks
[04:08] <T-Bone> heh, you'll get them :)
[04:09] <Kamion> I might need to get /proc/cpuinfo from you as well actually
[04:09] <Kamion> hm, I need new base-installer for PPC970FX support
[04:10] <T-Bone> hopefully dpkg is statically linked. Missing libunwind7 kills apt
[04:10] <T-Bone> Kamion: i'll send that as well
[04:10] <T-Bone> Kamion: oh, so basically i just downloaded 650M of worthless data? :^)
[04:11] <Mithrandir> T-Bone: no, dpkg is not statically linked.
[04:11] <T-Bone> Mithrandir: my bad, i meant it wasn't using libunwind
[04:12] <T-Bone> Kamion: i see that all kernel flavours available on the cd are smp ones. Could that be part of the problem? (I'm on a UP box)
[04:12] <Kamion> T-Bone: no, you may just need to pick the kernel by hand
[04:12] <Kamion> T-Bone: I was under the impression that SMP was fine on UP boxes; we can't afford space for all kernel flavours on the CD
[04:13] <Kamion> so if it'll work but is maybe a bit suboptimal, we include that and don't worry about the optimal ones
[04:13] <T-Bone> well iirc debian policy is the exact opposite: install up by default, right?
[04:13] <Kosai> Kamion: I've seen SMP/apic problems on UP, but they're rare.  It should be fine, AFAIK.
[04:13] <Kamion> mckinley's an exception because apparently itanium's really bad on mckinley boxes
[04:13] <Kamion> T-Bone: nope
[04:13] <Kamion> T-Bone: I'm most of the base-installer maintainer in Debian too ...
[04:13] <T-Bone> Kamion: yeah i know, but it's been a while since i've last tested Debian d-i ;)
[04:14] <Kamion> there isn't a consistent policy across architectures on this
[04:14] <T-Bone> Kamion: i'm used to a conservative approach WRT SMP, given the numerous bugs i've experienced on hppa and ia64 SMP boxes
[04:14] <T-Bone> (and recent kernels)
[04:15] <T-Bone> ah
[04:15] <Kamion> if real problems are demonstrated, we'll revisit it
[04:15] <T-Bone> another bug
[04:16] <T-Bone> if you do 'apt-get install linux-image-mckinley-smp', and then say 'n' to the "go further" question, you're left in an unstable state
[04:16] <Kamion> but the kernels are one of the few places where CD space is arch-dependent; if we blow out ia64 with kernels, then we have to reduce arch-independent stuff
[04:16] <T-Bone> apt wants you to run apt-get -f install
[04:16] <Kamion> sounds about right
[04:16] <Kamion> the preinst failed ...
[04:16] <T-Bone> no
[04:16] <T-Bone> the user answered 'no' to a given question
[04:16] <T-Bone> that's not a failure
[04:16] <T-Bone> imho
[04:17] <T-Bone> (i'm speaking from a user PoV)
[04:17] <T-Bone> it should cleanly abort the installation, and not require subsequent call to apt-get -f install or apt-get remove to be able to install any other package
[04:18] <Kamion> oh sure, just nobody's figured out how to do that. :)
[04:18] <T-Bone> lol
[04:18] <Kamion> I can't think of an interface in dpkg to abort a package installation and leave the package cleanly uninstalled
[04:19] <T-Bone> Kamion: simply silently run 'apt-get remove <package>' ?
[04:20] <Kamion> can't do that within a preinst
[04:20] <Kamion> dpkg is not reantrant
[04:20] <Kamion> er, reentrant
[04:20] <Kamion> (and besides you might not be using apt-get)
[04:20] <T-Bone> yeah sure, that's the general idea i was considering
[04:21] <Kamion> it basically only affects kernels; very few other packages have that initial "did you really want to install this?" question, and assume more sensibly that if you installed it then you meant to do so
[04:21] <T-Bone> yeah
[04:23] <bob2> 308MB dist-upgrade and not a single error
[04:23] <bob2> good work dudes
[04:25] <T-Bone> that usbkbd mess is a real pain :P
[04:26] <Mithrandir> bob2: yeah, we rock. :)
[04:26] <bob2> Unpacking replacement libwxgtk2.4-python ...
[04:26] <bob2> dpkg: error processing /var/cache/apt/archives/libwxgtk2.4-python_2.4.2.6ubuntu1_i386.deb (--unpack):
[04:26] <bob2>  trying to overwrite `/usr/bin/helpviewer', which is also in package wxpython2.5.3
[04:30] <Riddell> sladen: amu is doing KDE ubuntu as well
[04:32] <sladen> Riddell: ahhh
[04:34] <Kamion> ok, I think I've fixed the base-installer tests so that they can be used for Ubuntu in future
[04:38] <T-Bone> shit! That bastard didn't copied the logs when i asked to
[04:42] <Kamion> mdz: I'd like to merge base-installer 1.15 from experimental to fix a couple of issues with recent powerpc machines and to allow the kernel selection tests to be used on Ubuntu. Is that OK?
[04:45] <T-Bone> hmmm
[04:47] <T-Bone> i wonder if the kernel on the CD has all needed driver. Panic'd on VFS unable to mount root device 'hdb3'
[04:48] <Kamion> sounds more like a broken initrd
[04:48] <T-Bone> yup
[04:50] <T-Bone> kay i have the logs, the modified elilo. Need to test hotplug pci and then will mail you
[04:50] <Kamion> if the kernel on the CD didn't have that driver then you wouldn't have been able to create that partition or install to it in the first place
[04:50] <Kamion> thanks
[04:50] <Kamion> have to go to a concert now, later
[04:50] <T-Bone> Kamion: yeah sure, i misexpressed myself ;)
[04:50] <Kamion> ok
[04:50] <T-Bone> heh
[04:50] <T-Bone> have a nice day
[04:56] <T-Bone> bummer i didn't have the initrd line in my elilo.conf.
[04:59] <Keybuk> ripped the old wireless applet code and turned it into its own package
[04:59] <Keybuk> muahahahahaha
[05:01] <T-Bone> YAY! Getting there! The box boots
[05:01] <bob2> Keybuk: it's all about netapplet now anyway
[05:01] <Keybuk> bob2: yes, but netapplet depends on thom doing some work :p
[05:01] <bob2> haha
[05:04] <Treenaks> gm, no pitti
[05:04] <lamont> Kamion: because it's a poorly named package
[05:06] <lamont> Kamion: maybe it's just my memory that's faulty, but ISTR it got really mad at just libunwind7
[05:13] <bob2> lamont: found the postfix problem, it's because the scripts say they need /bin/sh but actually want bash
[05:14] <lamont> bob2: ah, and posh / dash doesn'
[05:14] <lamont> t like it, eh?
[05:14] <bob2> yeah, very much so
[05:14] <bob2> it fails in quite bizarre ways
[05:14] <lamont> will fix - please file a bug
[05:14] <bob2> debian or ubuntu?
[05:15] <lamont> (it's supposed to work with bin/sh)
[05:15] <lamont> I expect debian - go ahead and file it there
[05:15] <bob2> sure, thanks.
[05:15] <lamont> sev serious
[05:15] <bob2> hehe, cya :)
[05:30] <Riddell> elmo: can I try uploading a new amarok package with a fixed version number?
[05:30] <ogra> elmo ?
[05:47] <Simira> any Rosetta-people here?
[05:47] <HostingGeek> xorg i810 driver is broken or something
[05:48] <HostingGeek> i just --purge remove EVERYTHING including xorg
[05:48] <HostingGeek> and apt-get it all again and still am having the problem of the max pixel i can use is 800x600
[05:49] <crimsun> HostingGeek: let's take this back to #ubuntu.
[05:49] <HostingGeek> crimsun: lets unban me from #ubuntu first
[05:50] <tseng> maybe you could follow some simple rules
[05:50] <HostingGeek> thanks for unbanning me
[05:50] <tseng> like not proclaiming to developers that their system is broken, when you have no clue
[05:50] <T-Bone> HostingGeek: if you're not willing to RTFM and keep asking offtopic questions to everyone you're gonna be banned again
[06:02] <Keybuk> thank god, will that kid ever learn? :-/
[06:02] <Treenaks> oh wait
[06:09] <elmo> Riddell: yes
[06:09] <elmo> ogra: ?
[06:09] <ogra> my uploads seem to dissapear silently, is my key not in the ring yet ?
[06:09] <ogra> elmo ^^
[06:17] <T-Bone> geez, hoary install went fine, but i can't reboot my G5: open firmware freaks out when loading the kernel :P
[06:41] <Keybuk> T-Bone: yeah, it means that it doesn't matter exactly what we run or how we fuck up our own boxes
[06:42] <Keybuk> just how much LaMont fucks up the buildds :p
[06:42] <T-Bone> hehe
[06:42] <T-Bone> can't it lead to unbuildable package uploads?
[06:43] <Keybuk> yeah
[06:44] <Keybuk> that's the disadvantage
[06:47] <robertj> i'm working on my first .deb (just messing around), but I can't get the postinst and prerm to run
[06:48] <robertj> running by hand produces the desired result, any suggestions?
[06:50] <Keybuk> dpkg-deb -I blah...deb
[06:50] <Keybuk> does that show postinst in it?
[06:50] <robertj> one sec
[06:51] <Keybuk> it'll be in the list at the top with control (and probably md5sums)
[06:51] <robertj> no
[06:51] <Keybuk> then the reason your postinst and prerm aren't being run, is because they're not actually being put in the package
[06:51] <Keybuk> are you using debhelper?
[06:51] <robertj> yeah
[06:52] <Keybuk> and you're running dh_installdeb ?
[06:52] <robertj> hrmm, no
[06:52] <Keybuk> heh, that's quite a critical one
[06:52] <Keybuk> it copies files from debian/ into DEBIAN/ (which dpkg-deb --build turns into control.tar.gz)
[06:53] <robertj> thanks
[06:53] <robertj> building now
[06:53] <robertj> it's amazing the build difference between a 2.8 ghz celeron and a 1ghz duron
[06:54] <robertj> (whole machine was $350 from Dell)
[06:55] <robertj> I ran dh_installdeb and nothing changed after I built the package
[06:55] <robertj> do I need to create DEBIAN by hand?
[06:55] <Keybuk> no
[06:55] <Keybuk> try running "dh_installdeb -v" by hand
[06:55] <Keybuk> what does it do?
[06:55] <robertj> nothing
[06:56] <Keybuk> in debian/ you have a package.postinst file?
[06:56] <robertj> no, I have a postinst.ex
[06:56] <robertj> I guess that's probably it eh?
[06:56] <robertj> rename to mypackagename.postinst?
[06:56] <Keybuk> ah, rename that to <package>.postinst  (where <package> is the name of the package that should have it)
[06:56] <Keybuk> yup
[06:58] <robertj> btw, dh_installdeb must be automagic
[06:58] <Keybuk> why?
[06:59] <robertj> the files showed up without it
[06:59] <tritium> I have suspend to RAM mostly working, except for wireless lan card problems
[06:59] <robertj> well it's not working
[06:59] <robertj> hrmm
[06:59] <robertj> let me see if dh_installdeb is called from the control file
[07:00] <robertj> or rules
[07:00] <robertj> yeah
[07:00] <robertj> rules runs dh_installdeb
[07:00] <robertj>      991 bytes,    43 lines   *  postinst             #!/bin/sh
[07:00] <robertj>     1023 bytes,    41 lines   *  prerm                #!/bin/sh
[07:01] <Keybuk> there you go
[07:01] <robertj> the scripts still aren't working though, is there a way I can run them more verbosely?
[07:02] <robertj> set -e
[07:02] <robertj> gtk-query-immodules-2.0 > /etc/gtk-2.0/gtk.immodules
[07:02] <robertj> exit 0
[07:03] <robertj> that's the entirity of the script sans blank lines and comments
[07:03] <Keybuk> that's a dangerously incomplete script
[07:04] <robertj> yeah, but I figured it was best to get something working first
[07:04] <Keybuk> stick "set -x" in there? :)
[07:05] <Keybuk> (to complete it, add something like this to the top): [ "$1" = "configure" ]  || exit 0
[07:06] <robertj> I don't see why you wouldn't want it to run again
[07:07] <robertj> btw +x shows that  gtk-query-immodules-2.0 gets run but doesn't mention the > /etc/gtk-2.0/gtk.immodules bit
[07:10] <Keybuk> run it again?
[07:10] <robertj> yeah, all it does is check the locations in it's path and put them in the file
[07:10] <robertj> err all the modules it finds there
[07:11] <Keybuk> you're aware of all the situations that postinst can be run?
[07:11] <robertj> guess not
[07:11] <robertj> i thought only after a package was installed
[07:11] <robertj> or reconfigured 
[07:11] <Keybuk> no, it's also run if an upgrade to a newer version fails
[07:12] <robertj> oh
[07:12] <tritium> Does anybody know the difference between "platform" and "shutdown" for the hibernate_mode (caps) variable in acpi-support?
[07:12] <tritium> /etc/default/acpi-support says to use "platform" if you want acpi to shutdown, which seems backwards
[07:18] <robertj> Keybum: imhangul seems to run it in every case
[07:19] <robertj> whee
[07:19] <robertj> but now it's working
[07:29] <robertj> Keybuk: thinks for your help, I have alot of reading to do
[07:30] <Riddell> elmo: what is the status of my new unsermake package?
[07:40] <T-Bone> i'm not sure washing a cat is a good idea if you care about your clothes and skin ;)
[07:48] <T-Bone> sweet, after a fresh hoary install on ppc, the default user isn't in the sudoers
[07:49] <robertj> %s/washes/watches/g
[07:49] <robertj> T-Bone: can't trust those mac-using tards anyway ;)
[07:50] <T-Bone> lol
[07:50] <robertj> does your machine count include scanners that people carry to meetings thinking they are laptops?
[07:50] <T-Bone> btw, i was wondering what happens if the system falls back to single user mode at boot time... (maybe that's an already answered question tho)
[08:00] <mjg59> T-Bone: Straight to root shell
[08:00] <T-Bone> ?
[08:01] <T-Bone> ah ok
[08:01] <Keybuk> asking for a password is somewhat silly :)  given if they can type it, they can also boot with (e.g.) init=/bin/sh
[08:01] <T-Bone> wow nice work my bluetooth keyboard is supported, but not my mouse, afaict
[08:02] <T-Bone> Keybuk: damn true :)
[08:04] <T-Bone> ah yes, the mouse came back. Bluetooth neg lag i suppose
[08:13] <abelli> Kamion: ding
[08:22] <trulux> doko: hey
[08:22] <trulux> doko: more people interested in the meeting
[08:22] <trulux> i think we need to spread the word for all Ubuntu folks
[08:23] <thully> anybody know Ubuntu's position on the Mozilla trademark issues?
[08:24] <mjg59> fabbione: Hm, what happened to the patch that fixed interrupt handling after swsusp?
[08:35] <mjg59> fabbione: #5567 - I've attached a patch
[08:42] <no0tic> hi all! what's the acer's laptop support in hoary nowadays?
[08:43] <mdz> Kamion: apt doesn't authenticate packages; it doesn't even check the md5sum on packages in the cache (it's trusted)
[08:43] <mdz> Kamion: base-installer merge is fine with me
[08:47] <mjg59> no0tic: In what way?
[08:48] <no0tic> mjg59: suspend-to-disk, suspend to ram, acpi
[08:48] <mjg59> no0tic: On Hoary, you ought to get working suspend/resume
[08:48] <mjg59> suspend-to-disk is currently broken, but will be fixed in the next kernel upload
[08:48] <thully> mjg59: when will that be
[08:49] <thully> I need suspend-to-disk, and was about to try a Hoary install
[08:49] <mjg59> thully: Whenever fabbione builds it
[08:49] <mjg59> Ought to be tomorrow
[08:49] <no0tic> mjg59: ah, ok
[08:49] <mjg59> suspend and resume will work, but if you want PCI devices to work afterwards you need to pass the pci=routeirq parameter
[08:49] <mjg59> The new kernel will fix that so you don't need the parameter
[08:50] <thully> (i need this because my laptop uses 10% of it's power per hour in suspend-to-ram)
[08:51] <thully> however, apm suspend-to-ram works fine - weird - unfortunately speedstep doesn't work in APM :(
[08:51] <no0tic> mjg59: now in warty I have to close the lid after shutdown command to actually shutdown, otherwise it reboot... it's fixed?
[08:52] <mjg59> no0tic: Should be fixed
[08:52] <no0tic> mjg59: I often heard that ati fglrx drivers brake in some way suspend...
[08:52] <mjg59> no0tic: Yes, you may well have to use the stock ATI drivers
[08:53] <no0tic> mjg59: sorry, I didn't understand.. :)
[08:53] <mjg59> no0tic: the fglrx drivers don't support suspend and resume. You might be lucky and find that they work anyway, but otherwise you'll have to use x.org's radeon driver
[08:53] <mjg59> sladen: Around?
[08:53] <no0tic> mjg59: ok :(
[08:54] <mjg59> no0tic: Give it a try, anyway. If it doesn't work, contact ATI and ask them to support suspend and resume
[08:54] <no0tic> mjg59: I hope tomorrow ati's drivers release, will fix it, with the support at x.org 6.8.1
[08:55] <mjg59> no0tic: I haven't heard that they've improved that, but if they have done, that'd be good
[08:56] <mjg59> thully: But yeah, suspend to disk will work fine as long as you use the pci=routeirq thing
[08:56] <no0tic> ok, thanks for now!
[08:56] <thully> Oh, OK - I'll do that for now
[08:57] <sladen> mjg59: yup
[08:58] <mjg59> sladen: Know anything about apm?
[08:58] <mjg59> sladen: I've been thinking that we could possibly reverse engineer the T-series apm BIOS, which would give some indication as to what it's doing on apm suspend that isn't being done on acpi suspend
[08:59] <abelli> mjg59: ...
[08:59] <mjg59> abelli: Mm?
[08:59] <abelli> mjg59: im interested..
[08:59] <abelli> mjg59: please go on...
[08:59] <abelli> :)
[08:59] <thully> so is there a reasonably functional hoary live CD yet?  if so, what release?  is Gnoppix the same as hoary's live CD?
[09:00] <mjg59> Several Thinkpads use far too much power when in ACPI suspend, and so far we haven't been able to work out why
[09:00] <mjg59> So checking to see what the BIOS does in APM ought to give some indication
[09:00] <sladen> mjg59: yup.  the magic code is all in the 64Kb of SMM code.  This is outside of the normal 64GB of memory space but has to code from somewhere, so may be possible to pull it out of the BIOS image if you can get it out
[09:01] <mjg59> Right, so we need a full BIOS image?
[09:01] <mjg59> Did we ever work out the compression scheme IBM used?
[09:01] <abelli> i think so
[09:01] <sladen> mjg59: communication (same as ACPI) is normally through port 0xb1/0xb2 which aswell as taking the value given, cause SMI to be asserted  (From memory)
[09:02] <mjg59> sladen: Ok. So we need to find a full T-series BIOS image.
[09:02] <abelli> mjg59, sladen : i like it... how much is a Thinkpad that has the features youre working on?
[09:02] <mjg59> Oh. Unless it would be in the EC firmware?
[09:02] <sladen> mjg59: no, but some guy found me and was bugging me out it (yesterday of all coincidences).  flash2 is doing the ibm-specific decompression, patching in the system ID, bios version etc, producing  bios.wph  and then calling phlash16 which is the bog-standard Phoenix utility
[09:03] <thully> Can I extract it from my laptop somehow?
[09:03] <thully> I have a t42
[09:03] <mjg59> thully: With some difficulty
[09:03] <mjg59> sladen: Is this likely to be BIOS code or EC firmware code?
[09:03] <abelli> mjg59: is it, the bios, in the controller, or on a separate rom?
[09:03] <thully> if it involves doing anything hardware-based, I can't do it
[09:03] <sladen> mjg59: x86 code, it's run on the normal CPU, 
[09:03] <mjg59> abelli: That's what I'm trying to find out :)
[09:04] <mjg59> sladen: IBM provide separate EC and BIOS flashes
[09:04] <thully> meaning - I don't want to void my warranty and pull out some chip, etc.
[09:04] <abelli> because a teacher of mine, at the university.. helped me doing the same thing on the xbox..
[09:04] <thully> mjg59:  do the BIOS updates on IBM's site help you at all?
[09:04] <abelli> i think he could help us
[09:04] <mjg59> thully: They're compressed in an irritating format
[09:05] <mjg59> sladen: Ooh. Do you think we can get the flash2 stuff to work under dosemu?
[09:17] <Kamion> mdz: a current hoary install seems to refuse to install packages in the cache without turning on allowunauthenticated
[09:17] <Kamion> mdz: namely laptop-detect, mdetect, xresprobe
[09:18] <Kamion> abelli: pong
[09:18] <abelli> Kamion: resolved thank you.
[09:19] <abelli> Kamion: was about d-i.
[09:21] <mdz> Kamion: that means it can't authenticate the Packages file that it guesses they came from
[09:21] <T-Bone> Kamion: the default config for ppc kernel seems to lack a few things, while having worthless options enabled
[09:21] <sladen> mjg59: maybe, but it's that program (flash2) that is doing the checks ''do you have a thinkpad'', ''is it the right version'', ''do you have AC and battery plugged in and the battery is fully charged''
[09:23] <doko> Kamion: what is wrong with the libunwind library, anything I can test?
[09:24] <Kamion> T-Bone: I'm not the kernel maintainer; please file a bug with details
[09:24] <Kamion> doko: not AFAIK? I was just asking why lamont had put libunwind7-dev in base
[09:24] <doko> ahh, libgcc1 depends on it.
[09:25] <Kamion> lamont: if debootstrap works with libunwind7-dev removed from base, that would be far preferable, since then it will actually work when autogenerated from the seeds (it currently doesn't match what germinate says)
[09:25] <Kamion> doko: on libunwind7-dev? really?
[09:25] <doko> kamion: no, not the -dev.
[09:25] <Kamion> ok, good
[09:26] <mjg59> sladen: Yeah. I'm wondering whether there's a command-line switch to disable that stuff.
[09:26] <T-Bone> Kamion: is reportbug to be used for that kind of bugs or should i use the bugzilla web interface instead?
[09:26] <mjg59> Under dosemu, the checks thing returns 21
[09:26] <Kamion> T-Bone: bugzilla please
[09:26] <T-Bone> gah :(
[09:26] <Kamion> reportbug just mails -users
[09:27] <Kamion> cheesy hack until launchpad works properly :-/
[09:27] <T-Bone> yummy
[09:27] <lamont> Kamion: I'm happy to be found 100% wrong about libunwind7-dev
[09:27] <T-Bone> i just don't like bugzilla
[09:27] <Kamion> T-Bone: nor do I
[09:28] <Kamion> lamont: ok, so you won't be too annoyed if I do another debootstrap upload that matches the seeds? :) I need to do one soon for authentication support
[09:28] <T-Bone> heh
[09:28] <T-Bone> Kamion: could you get something useful of my few mails?
[09:29] <mjg59> sladen: Ok, there's no command line option. Ought to be a simple hack though, right?
[09:30] <Kamion> T-Bone: I only have your initial mail plus the followup about anna/zx2000/etc.
[09:30] <Kamion> if there was more it hasn't reached me
[09:30] <lamont> Kamion: way cool with me.
[09:31] <lamont> Kamion: and I believe that the ln -s vs sf bug is already in the debian bts
[09:31] <lamont> with me as the submitter
[09:31] <sladen> mjg59: there is a /batt or something which I guess tells it to ignore the battery state
[09:31] <crimsun> could someone please do a simple recompile of python-ecasound in hoary (it just needs to be updated for python2.4)
[09:31] <T-Bone> Kamion: i sent you a reply to the mail you sent me with attachements, and some followups about the G5 install
[09:32] <Kamion> T-Bone: oops, it got caught as spam. d'oh
[09:32] <thom> lamont: you were looking for me?
[09:32] <Kamion> my rules are a bit overenthusiastic about killing attachments sometimes
[09:32] <mjg59> sladen: it looks like the only times it prints anything on screen are when it's doing the checks, which ought to make it easy enough to drag out
[09:32] <sladen> mjg59: strings on it was somewhat interesting;  I'm puzzled by updtflsh*  the bigger one (39kB) apperas to contain very similar code to flash2---and the some block of parameters to pass to phlash
[09:32] <mjg59> Is the only way to print stuff in DOS to do bios interrupts?
[09:32] <ogra> crimsun: i would love to, but obviously elmo hasnt given me access yet....(dunno if i miss anything required thiugh)
[09:33] <lamont> thom: uh... not still looking, at any rate
[09:33] <sladen> mjg59: it's the only way they're probably using.
[09:33] <crimsun> ogra: aye, I understand.
[09:33] <thom> oh well
[09:33] <lamont> Kamion: I figured I was missing a step or 3 in my debootstrap upload... sorry about that
[09:33] <T-Bone> Kamion: the G5-related mail had attachements as well (it's stated in both mails anyway)
[09:33] <mjg59> sladen: uptdflsh is a menu-driven thing
[09:33] <mjg59> It seems to do the same otherwise
[09:35] <Kamion> T-Bone: yup, got both
[09:35] <mjg59> sladen: So we find out which code prints stuff to screen and then work back from there?
[09:35] <Kamion> lamont: no, it's ok, the auto-germinator bit isn't in the debootstrap source package yet :(
[09:36] <Kamion> Jan 16 14:59:57 base-installer: warning: apt update failed: 127
[09:36] <Kamion> that would explain why it can't find any kernels ...
[09:36] <sladen> mjg59: I think int 0x15 is the BIOS video stuff.  There's also a DOS (0x21) call that many people used which I think copes with redirection, 
[09:36] <Kamion> is VERY confused
[09:37] <T-Bone> heh
[09:37] <mjg59> sladen: Ok, there's a stack of both of them
[09:41] <sladen> mjg59: other thing is just to disassemble and grep for a shortish loop (<100-200 instructions) with a right  lodsb/stosb combination.  The compression is LZ-something based (you can see the original ASCII strings in the file), so it's a copy/marker setup
[09:41] <sladen> s/right/tight
[09:43] <mjg59> sladen: Heh. Easier to skip the checks if possible, I think.
[09:43] <sladen> mjg59: the BIOS this guy gave me yesterday was 1UET63WW 
[09:44] <mjg59> I'm not really in the mood to rip out and reimplement someone's compression algorithm :)
[09:44] <mjg59> What model's that from?
[09:45] <sladen> mjg59: dunno
[09:46] <mjg59> Ah! This is the section that shuts it down.
[09:46] <sladen> mjg59: I also got a mail from the ibm-acpi guy today saying he's done a -10 with my changes in it, but he doesn't seem too keen on stick the tpb nvram stuff into the driver
[09:46] <sladen> mjg59: where did you find the code?
[09:47] <mjg59> sladen: I'm just doing ndisasm over flash2
[09:48] <thully> This is very interesting - you're reverse engineering the IBM T-series BIOS
[09:48] <thully> I hope it leads to a fix to this issue
[09:48] <mjg59> thully: At the moment, I'm reverse engineering the code that decompresses the T-series BIOS
[09:48] <mjg59> The BIOS itself is going to be harder :)
[09:48] <thully> ah
[09:50] <sladen> thully: not the first time http://www.paul.sladen.org/thinkpad-r31/x40-bios-dis.txt ; and I was trying to get LinuxBIOS on my R31 too
[09:55] <Kamion> T-Gone: ah-ha! the entire kernel thing is due to libunwind7 not being there
[09:55] <Kamion> apt-get: error while loading shared libraries: libunwind.so.7: cannot open shared object file: No such file or directory
[09:55] <Kamion> that's what happens when base-installer tries to call apt-get update
[09:55] <Kamion> so it can't find any kernels
[09:59] <sladen> mjg59: if I had a copy of the same file, both compressed and uncompressed.  Do you still have that for your machine back from the minipci patching?
[10:01] <T-Gone> Kamion: i suspected that yeap
[10:03] <mjg59> sladen: Ought to, yes
[10:03] <mjg59> Hang on...
[10:03] <Kamion> T-Bone: fixing your powerpc kernel issue now by means of a merge from Debian
[10:04] <T-Bone> the 'wrong kernel gets installed' one?
[10:07] <mjg59> sladen: Hrm. I have an uncompressed one, but can't find the compressed one to go with it
[10:07] <mjg59> Oh, I'm looking on the wrong machine. Duh.
[10:09] <Kamion> T-Bone: yes, it'll install linux-power4 now
[10:09] <Kamion> (uploaded)
[10:09] <T-Bone> cool
[10:11] <T-Bone> Kamion: i've tested modified elilo as much as i could, it seems to behave as we'd like
[10:15] <Kamion> ok, will look tomorrow, not going to be around much more this evening
[10:15] <sladen> mjg59: if you have the image, can probably grep the 1UET63WW style identifer from it and find the install dis
[10:24] <mjg59> sladen: http://www.codon.org.uk/~mjg59/bios/
[10:24] <mjg59> I believe those are both the same BIOS
[10:26] <mdz> Kamion: has apt-get update been run by that point?
[10:31] <sladen> mjg59: interesting coincidence, that is the same (md5sum) one I was playing with yesterday.  
[10:31] <Kamion> mdz: I think it should have been run by apt-setup, yes
[10:31] <mdz> Kamion: are there .gpg files in /var/lib/apt/lists?
[10:31] <mdz> those are apt's "I authenticated this" markers
[10:32] <Kamion> mdz: I can check later; my test install is on the OQO and non-networked
[10:32] <Kamion> I'm away from home at the moment
[10:32] <mdz> ah
[10:33] <mdz> oh, we have source CDs now
[10:33] <mdz> ...he said as his disk filled overnight
[10:35] <Kamion> :-)
[10:37] <jdub> haha
[10:37] <jdub> morning all!
[10:37] <sivang> morning jdub 
[10:37] <jdub> where's this elmo character i hear about?
[10:38] <jdub> probably at church, being such a baby jesus fearing young man
[10:39] <thom> morning duder
[10:39] <jdub> dude
[10:39] <jdub> i had a funny night last night
[10:39] <jdub> was setting up my server, testing new imap setup for pipka
[10:40] <jdub> decided to try imapping on my extensive mail archives
[10:40] <thom> uhuh
[10:40] <jdub> and found all sorts of funny stuff
[10:40] <jdub> excie :-)
[10:40] <thom> old table/slug stuff?
[10:40] <thom> hehe!
[10:40] <thom> gods thats old school
[10:40] <jdub> i think i have the first "meet you on irc" mail too
[10:41] <thom> heh
[10:42] <sladen> jdub: this is #gagging-for-it ?
[10:42] <thom> i think all my mail from that long ago is dead
[10:43] <jdub> it is very weird reading old mail
[10:43] <jdub> but i can't believe how much has been packed into the last four years
[10:43] <Keybuk> ever tried reading old blog posts?
[10:44] <jdub> some of my advogato ones would be amusing, i'm sure
[10:44] <rburton> jdub: "decided to help out with the gnome 2 release. should be over shortly"
[10:45] <jdub> rburton: yeah, those are funny
[10:45] <jdub> rburton: garnome complaints from a young ross burton were funny too ;)
[10:46] <rburton> doh
[10:46] <sivang> rburton: hehe
[10:47] <jdub> 15 april 2002 might be interesting though
[10:48] <sivang> ehrm people, don't want to sound like a #users issue, but how do i get my gnome back running after the upgrade? :)
[10:48] <sivang> (this is first time nothing I do helps)
[10:49] <jdub> you have blank panels?
[10:49] <sivang> none
[10:49] <sivang> I have X.
[10:49] <sivang> period
[10:49] <sivang> :)
[10:49] <sivang> oh, and gdm that fires up the gnome-session.
[10:50] <jdub> what happens after this so called... "period"?
[10:50] <sivang> hmmmm....nothing? :)
[10:50] <sivang> I just get the X screen and that's it.
[10:50] <tseng> just brown?
[10:51] <sivang> tseng: now, plain X offwhite texture, like when running X without parameters or a client app.
[10:51] <sivang> *no
[10:51] <tseng> the checker patter thingy?
[10:51] <jdub> that's bong
[10:51] <tseng> nasty++
[10:52] <mjg59> sladen: Are you /sure/ it writes a bios.wph file?
[10:52] <sivang> tseng: exactly.
[10:52] <sivang> ergghgh
[10:52] <sivang> suddenly came back
[10:53] <sivang> I don't get it..
[10:53] <sivang> I just apt-get up/upg again, nothing was changed, startxed and it works.
[10:53] <sivang> garrr
[10:53] <tseng> well that pattern is patched out in gdm.conf, its supposed to set it to a slightly nicer brown
[10:53] <tseng> so something is pretty well off.
[10:53] <sivang> weird
[10:54] <sivang> hmm, fonts are smaller....well, could have been worse..
[10:56] <sivang> like all went non-bold
[11:04] <thom> hurrah, my desktop has tomboy love again. life is complete
[11:04] <sivang> thom: you reintalled your desktop?
[11:05] <thom> sivang: no?
[11:05] <thom> sivang: just reinstalled mono
[11:05] <sivang> oh
[11:05] <sivang> thom: anyway, night :)
[11:05] <thom> night
[11:07] <sladen> mjg59: now that I've looked, I'm puzzled.  It might not be
[11:08] <sladen> mjg59: I suspect .wph == win phlash
[11:09] <sladen> mjg59: I'm also wondering whether ''flash over LAN'' actually means ''PXE load a floppy image and execute it''
[11:11] <mjg59> sladen: I thought I'd flashed a compressed image with phlash.exe
[11:12] <mjg59> On floppy, there's no room to write an uncompressed file anywhere, which makes me suspect that it's phlash that does the decompression
[11:15] <sladen> mjg59: didn't you get it unpacked the previous time by using the windowsy bios-editor tools?
[11:16] <sladen> mjg59: the supposed "demo" version;  however the guy yesterday claims that he used that and succesfully claimed the BIOS boot image in his vmware session (?)
[11:18] <sladen> mjg59: grep 'FL' *.EXE -> PHLASH16.EXE  I think you win
[11:18] <mjg59> sladen: I don't /think/ I could open the compressed ones there
[11:20] <mjg59> Are old ubuntu debs available from anywhere?
[11:21] <mjg59> sladen: I unpacked it last time by flashing it and making a backup copy...
[11:30] <Riddell> are the buildd's down?
[11:30] <azeem> elmo said something about that, yes
[11:30] <azeem> dunno if that is still the case, it has been a while
[11:33] <mjg59> Hurrah. It's freeze_processes()'s fault.
[11:33] <buga> Riddell: I'm also waiting for your new amarok binary packages. :)
[11:38] <mjg59> Hurrah. Got it, I think.
[11:41] <kent> Just a little notice. Using Hoary, if i start Program -> system tools -> ubuntu update manager.  Then press the "settings" button, i get a window called "software preferences" where it says that Hoary is officially supported. But, is this true? I mean, since its a development-version..?
[11:42] <mjg59> Woo!
[11:43] <sladen> kent: I think that's saying the packages are produced by Ubuntu, rather than say 'random' mariallet archives
[11:43] <sladen> mjg59: news?
[11:44] <Kamion> mjg59: morgue.ubuntu.com
[11:45] <Kamion> kent: sounds like the sort of thing where if it matters to the code whether it's "officially supported" or not, then it's better to have that in well before release so that we can test it
[11:45] <mjg59> sladen: With a swsusp bug, not the BIOS (sadly)
[11:45] <mjg59> Kamion: Ah, thanks
[11:45] <mjg59> bugzilla is very, *very* slow
[11:46] <Kamion> go bugzilla
[11:46] <Kamion> we love you with rapturous flaming death
[11:47] <sladen> ''oooh, it's Canonical-sponsored international snail racing with Wiki on the Right and Bugzilla on the left.  
[11:49] <mjg59> I fucking rock
[11:49] <Kamion> mjg59: you can have my installer bug list if you like
[11:50] <mjg59> Kamion: Only when I start being given your salary as well
[11:52] <Kamion> heh
[11:52] <Kamion> a fair point :-)