[12:03] <jbailey> It needs to be loaded in the initrd on IDE root systems, though.
[12:03] <jbailey> I had to add ide-generic to get my laptop to boot, anyway.
[12:03] <mdz> oh, you aren't using hotplug in your initrds?
[12:04] <jbailey> No, the plan had been to have hotplug in userspace give me a list of the drivers, and then I loaded all of those manually.
[12:04] <jbailey> That way hotplug at bootup got the coldplugging events.
[12:04] <mdz> hmm
[12:04] <jbailey> Or rather at sysvinit time.
[12:04] <mdz> proper hotplug in the initrd has many benefits
[12:04] <jbailey> (I tend to think of boot up being after I've chained to the proper init)
[12:05] <mdz> like not having to mangle the initrd if your root device moves
[12:05] <jbailey> Yup.  The downside is that it means all the boot drivers and stuff need to be in the initrd.
[12:05] <jbailey> Doable, though.
[12:05] <jbailey> That would actually make my task so far way easier.
[12:05] <mdz> storage drivers and network drivers should be there
[12:05] <mdz> (network to support NFS-root)
[12:05] <jbailey> and usb keyboard.
[12:05] <mdz> yeah
[12:05] <mdz> that would address another class of bugs too
[12:05] <jbailey> Yup
[12:06] <mdz> it'd do two coldplugging rounds, one with the initrd modules only, and a second one with all modules
[12:06] <jbailey> Cool.  Id' been chatting with dilinger about various ways to convince hotplug to give me a list of the devices that it would need for any given kernel version.  This saves that hassle at the cost of a larger initramfs.
[12:07] <mdz> this is what I thought you were talking about earlier
[12:07] <mdz> when you said you were doing it so as to replace initrd-tools
[12:07] <jbailey> mdz: Nope.  I wanted to keep the initramfs small so do it at generate time.
[12:08] <jbailey> Doing it at boot time is a simpler version of it.
[12:08] <mdz> so either way, using the hotplug infrastructure rather than mkinitrd's hacks
[12:08] <jbailey> Yes. =)
[12:08] <mdz> but we were thinking of two different approaches
[12:08] <mdz> we're already using hotplug in d-i
[12:08] <mdz> so we've got its dependencies down to a manageable level
[12:09] <jbailey> Yeah, saw that on the hoary goals.
[12:09] <mdz> it basically works with busybox-cvs + module-init-tools, I think
[12:09] <jbailey> *sigh* hotplug doesn't apply all of its patches cleanly on a fresh apt-get source.
[12:09] <mdz> oh?
[12:10] <jbailey> I've been using the tools that come with klibc, but I've seen that other people have gotten hotplug running with klibc.
[12:10] <mdz> unstable or hoary?
[12:10] <jbailey> hoary
[12:10] <mdz> they all apply for me in hoary
[12:10] <jbailey> ppc box.  The only non-ubuntu thing on here is dvdcss and the debian ppc kernel (Hoary's doesn't boot on the pegasos box.  The grub2 update on Monday should fix)
[12:10] <jbailey> Applying patch debian/patches/./blacklist2-eepro100 failed!
[12:11] <mdz> how weird
[12:11] <mdz> it's possible that it's never been built on powerpc before
[12:11] <mdz> but I see no reason why that should matter
[12:11] <mdz> Applying patch debian/patches/./blacklist2-eepro100 successful.
[12:12] <jbailey> Nothing in the hotplug Makefile or the spec file about klibc, but hopefully it's not too much elbow grease to beat it into shape.
[12:13] <mdz> hotplug doesn't have any binaries
[12:13] <mdz> it's all scripts
[12:13] <jbailey> Oh, is it all shell?
[12:13] <mjt> btw, as of 2.6.10, there's no /sys/bus/pci/devices/xx/driver symlink - helps with coldplug
[12:13] <jbailey> Nice. =)
[12:13] <mjt> s/no //
[12:13] <mdz> the only C bit is grepmap, which is 1) in a separate package, 2) extremely simple (should be fine with klibc), and 3) optional :-)
[12:14] <mdz> it really helps performance, though, so I recommend bringing it into the initrd
[12:14] <mjt> when you have only few modules in initramfs, performance does not matter
[12:14] <mjt> (or initrd, whatever)
[12:15] <mjt> and, using modules.alias instead of modules.*map for most things speed things up further
[12:15] <mdz> I'm not sure how few we'll end up with
[12:16] <mdz> hotplug doesn't seem to have any support for modules.alias whatsoever
[12:16] <mjt> mdz: i think we talked with you about this very stuff a while back, no? ;)
[12:16] <mdz> yes
[12:17] <mjt> modules.alias works for pci - the most important thing. inputmap (where modules.alias does not work) is small
[12:17] <mdz> but you're arguing a nonexistent solution against an existent one, and the existent one always wins :-)
[12:17] <mjt> non-existing?
[12:17] <mdz> hotplug does not use modules.alias
[12:17] <mdz> it uses modules.*map only
[12:18] <mdz> the version we're using, anyway
[12:18] <mjt> while read alias module; do case $deviceid in $alias) echo $module;; done < modules.alias -- that's all that is needed
[12:18] <mjt> dash rocks, btw
[12:19] <mjt> or, rather, while read junk alias module; do ...
[12:19] <mdz> that, getting deviceid into the right format, and testing it for a few thousand user=months, yeah, that's all :-P
[12:19] <mdz> s/=/-/
[12:19] <mjt> eh?
[12:20] <mjt> printf "pci:v%08Xd%08Xsv%08Xsd%08Xbc%02Xsc%02Xi%02X" $vendor $device $s_vendor $s_device $baseclass $subclass $if
[12:20] <mjt> that's the $deviceid
[12:21] <mdz> for PCI devices
[12:21] <mjt> vendor etc are from /sys/bus/pci/device/xx/vendor etc directly
[12:21] <mjt> yes
[12:21] <mjt> the main case
[12:21] <mdz> I see no incentive whatsoever for us to switch to using modules.alias, honestly
[12:21] <mjt> usb is even simpler
[12:21] <mjt> that switch makes grepmap even more optional
[12:22] <mdz> I'm not entirely convinced of that
[12:22] <mdz> just doing a while/read loop over the pcimap took longer than running grepmap
[12:22] <mjt> note modprobe handles wildcards for you, so that while.. loop is not needed
[12:22] <jbailey> mdz: At 200k uncompressed, I don't want to use more than I have to. =)
[12:22] <mjt> but that loop is fast - much faster than current cruft in pci.agent etc
[12:23] <mdz> jbailey: I don't see much choice about adding busybox to that, if we're going to use hotplug
[12:23] <mdz> that's ~128k
[12:23] <mjt> btw, module stuff in busybox is sorta broken 
[12:24] <mdz> yes, that's why we don't use it
[12:24] <mdz> not even in the installer or initrd
[12:24] <mjt> and busybox is a very good thing to have in an initrd
[12:24] <mjt> helps alot if something goes wrong
[12:25] <jbailey> mdz: There are reports on the hoplug list that maked it look like udev, hotplug, klibc, and modprobe should be enough.
[12:25] <mdz> we need busybox for a shell toolbox, not for module utilities
[12:25] <mjt> (saved my ass several times with broken root raid stuff ;)
[12:25] <mdz> jbailey: the hotplug scripts use cut, all sorts of stuff that's not built into dash
[12:26] <mdz> some of them even use awk, but those are not needed in initrd
[12:26] <mdz> grep
[12:26] <mjt> cut - where? in tape.agent?
[12:27] <mjt> grep - in net.agent?
[12:27] <mdz> grep is used _everywhere_
[12:27] <jbailey> Yeah.  The utilities in klibc doesn't have grep or cut.
[12:27] <mdz> sed is used _everywhere_
[12:27] <jbailey> cat false insmod ln mkfifo nuke run-init true uname chroot fstype ipconfig minips mount pivot_root sh udev dd gzip kinit mkdir nfsmount printf sleep umount
[12:27] <mjt> in hotplug.functions, really
[12:28] <mdz> mjt: in functions in hotplug.functions which are called from _everywhere_
[12:28] <mjt> yeah
[12:28] <jbailey> http://sourceforge.net/mailarchive/forum.php?thread_id=6319817&forum_id=3157 is the email that I'm looking at.
[12:29] <mdz> busybox gives you everything needed to run hotplug at a small cost
[12:29] <mdz> though if you're doing initramfs rather than initrd, that would make busybox a build-dep of the kernel, no?
[12:29] <mjt> grep $MODULE /proc/modules - ugh, isn't it simpler to use [ -d /sys/module/$MODULE ]  ?  Or just let modprobe to do its work? ;)
[12:29] <mdz> the former is portable across 2.4 and 2.6
[12:30] <mdz> hell, 2.2
[12:30] <mjt> is 2.4 supported in ubuntu?
[12:30] <jbailey> mdz: No.  initramfs can be generated anytime after.  You can just make a cpio image and hand it to the kernel the same way you hand it an initrd.
[12:30] <mdz> no
[12:30] <mdz> but it's supported by hotplug
[12:30] <mdz> hotplug is not specific to ubuntu :-)
[12:31] <jbailey> mdz: If we *did* make it a dependancy of the kernel, we could just ship an already configured initrd.  Apparently (I haven't tested) you can hand in multiple cpio files and it'll unpack them all.
[12:31] <mjt> initramfs works the same way as any other sort of initrd. except of the way you do final pivot_root/whatever stuff - for initramfs, there's special utility in klibc for that.
[12:31] <mdz> a preconfigured initrd would be pretty fly
[12:31] <jbailey> mdz: So that way the only bit to be done is just light config files (mostly for sane net booting), but those can also be handed in on the command line.
[12:31] <sladen> jbailey: is cpio like tar;  you can just concatrenate stuff?
[12:32] <mdz> jbailey: there's so much potential for weird bugs generating the initrd on the fly
[12:32] <mjt> neither tar nor cpio can be concatenated ;)
[12:32] <jbailey> sladen: Eh?
[12:32] <mdz> you can concatenate new files to a tar
[12:32] <mdz> s/concatenate/append/
[12:32] <sladen> mjt: okay, remove the 2kB of zeros from the end :)
[12:32] <jbailey> New files, yes.  Not multple tars.
[12:32] <mjt> after removing the trailer, yes
[12:33] <jbailey> Yeah.
[12:33] <mjt> both tar and cpio have a trailer record
[12:33] <jbailey> mdz: Ayup
[12:33] <mjt> btw, generating initramfs (the cpio part) is umm.. not easy
[12:33] <jbailey> mdz: So...  I'll need to do a version of busybox that's built with klibc.
[12:33] <jbailey> mjt: http://people.ubuntu.com/~jbailey =)
[12:33] <mdz> jbailey: we would need to build-dep on non-kernel stuff anyway, I guess; klibc seems to lack a shell ;-)
[12:33] <jbailey> mdz: klibc has ash.
[12:34] <mjt> i've written something in shell to do that (cpio) -- pretty fun stuff
[12:34] <mdz> oh?
[12:34] <mdz> in how many places do we need to duplicate this stuff? :-)
[12:34] <jbailey> BTW, you really want a newer version of the mkinitramfs than is there.  I found a bug a minute or so ago. =)
[12:34] <jbailey> mdz: As few as possible, ideally. =)  But we don't live in an ideal worls.
[12:34] <mjt> jbailey: can yours be used as non-root?
[12:34] <mdz> we already fork loads of stuff into busybox
[12:35] <mjt> btw, why klibc?
[12:35] <jbailey> mjt: Yes.  (cd ${TMPDIR} && find . | cpio --quiet --dereference -o -H newc | gzip -9 >${outfile})
[12:35] <mjt> there's nothing that stops using other stuff (dietlibc, uclibc, even glibc but it's large) in initramfs
[12:36] <jbailey> mjt: The difference is that klibc is eventually being targetted to be integrated in with the kernel build anyway.
[12:36] <jbailey> mjt: udev is already prewired for udev, as is the replacement for pivot_root
[12:37] <mdz> what does the pivot_root replacement do differently?
[12:37] <mjt> jbailey: hmm. that cpio example - 2 probs. first of all, it packs ./ instead of / - last time i checked kernel chocked on that (that's why i wrote "my own cpio" in shell - pretty easy really); and 2), that way requires mknod, which is root-only
[12:38] <jbailey> mdz: It cd's into the mounted root directory, deletes everything on the parent filesystem, overmounts the mounted directory as /, unmounts the previous one.
[12:38] <jbailey> mdz: So it's all done in userspace, no in kernel magic.
[12:38] <mjt> mdz: take a look at kinit in klibc
[12:38] <jbailey> mjt: It's working fine on two systems here with the ./, and there are no devices in there.
[12:38] <mdz> jbailey: hmm, I hope that doesn't mean pivot_root(2) is going away
[12:38] <jbailey> mjt: The kernel provides /dev/console automatiically.  Everything else is done with udev.
[12:39] <jbailey> mdz: I beleive that it does.
[12:39] <mjt> pivot_root isn't needed with initramfs
[12:39] <mdz> what about kernel threads?
[12:39] <mdz> they find their way to the new root ok?
[12:39] <jbailey> mdz: I'm using run-init.  kinit isn't quite useful yet.
[12:40] <mjt> er.. how it's not useful?
[12:40] <mjt> oh /me bad
[12:41] <mdz> kinit in klibc 0.198 seems to use pivot_root
[12:41] <mjt> i really meant run-init, not kinit ;)
[12:41] <mdz> oh, run-init is the example
[12:41] <jbailey> mjt: It's a conglomerate tool that basically just combined run-init and nfsmount.  It doesn't have anything in the way of extra magic.  I think I had another concern about it, but it's not coming to mind.
[12:41] <mjt> utils/run-init.c
[12:41] <mjt> yeah - sorry for the noise
[12:41] <mdz> I would not have expected mount(".", "/", NULL, MS_MOVE, NULL) to work :-)
[12:42] <jbailey> mjt: No, all good. =)  I'm still learning the new initramfs stuff.  I went to a workshop at OLS on it and come out with more questions than I went it with.
[12:42] <mdz> but I can certainly adapt casper to that when pivot_root goes away, assuming kernel threads dtrt
[12:42] <mjt> what's wrong with kernel threads and that initramfs/run-init stuff?
[12:42] <jbailey> Oh!  Something that occured to me.  If we build the initramfs right into the kernel, would could probably still do thge DSDT update by handing it in using the initrd= method.
[12:43] <jbailey> mdz: Casper?
[12:43] <mdz> jbailey: the live CD magic
[12:43] <mdz> it uses pivot_root(8)
[12:44] <mjt> so just replace pivot_root(8) with run-init(to-be-8) ;)
[12:44] <mjt> pretty easy to integrate into busybox, btw
[12:44] <mdz> more likely replace pivot_root(8) with similar mount/chroot stuff
[12:45] <mdz> it needs to do a few more things before execing init
[12:45] <mdz> and it doesn't exec init, even, it lets init do that
[12:45] <mjt> yes - deleting old stuff
[12:45] <mdz> no, I meant casper needs to do more things
[12:45] <jbailey> mdz: I need to bug off to a dinner engagement...  Anything else you need before I run off?
[12:46] <dholbach> sleep tight guys... i'm off
[12:46] <mjt> i once tried to pivot_root and umount /initrd with initramfs - the kernel freezes... ;)
[12:46] <mdz> jbailey: nah, I need to run myself
[12:46] <mdz> jbailey: you want to take on #1763 for Hoary?
[12:46] <mdz> it would clean up a lot of mess
[12:48] <jbailey> mdz: Yeah.  I've just assigned it to me.
[12:50] <mdz> ok, thanks
[12:50] <mdz> I've been pretending to get around to it for too long
[12:50] <sladen> jbailey: the DSDT updates can be done at the moment by appending it onto the initrd
[12:51] <jbailey> sladen: If we're talking about shipping a pre-generated config, though, it would be nice to not muck with the shipped kernel at all, and have grub hand it in.
[12:51] <sladen> jbailey: from what I remember the initramfs includes a more formal way of doing that
[12:51] <jbailey> sladen: That way it's completely independant with no magic.
[12:52] <sladen> jbailey: I was originally thinking of hacking grub to do the appending at load time, but I think other people were keener on doing it in mkinitrd
[12:52] <jbailey> sladen: I don't know enough about it yet.  I'm lucky.  Aside from being fubar with 2.6.10, my laptop's acpi Just Works(tm)
[12:53] <mdz> Kamion: how much of a headache would it be to rename /install to /boot (or some such) on the CDs?
[12:53] <jbailey> sladen: I gotta run, sorry.
[12:53] <jbailey> *poof*
[12:59] <mdz> jdub: when the about ubuntu page moves into yelp, will we be able to translate it then?
[01:20] <YokoZar> Hey, can someone help me with a packaging problem I'm having.  It seems like it's a simple thing I'm overlooking (a make error) but, strangely, the .deb file builds and works fine (although it bugs out before it can ask for my password)
[01:22] <mjt> .deb file should not ask for any passwords...
[01:22] <YokoZar> Err yeah it gets there and then bugs out before signing the dsc
[01:23] <mjt> if there's a "make error", it should not build.  error in signing (as dpkg-buildpackage performs) is non-fatal
[01:26] <lupus_> would enabling extended attributes on the home directory by default
[01:26] <lupus_> inpact performance
[01:26] <lupus_> impact
[01:28] <mjt> just enabling EAs makes no difference
[01:28] <lupus_> beagle needs it :)
[01:28] <lupus_> so I wonder if it be worth while for ubuntu to enable it
[01:28] <mjt> enabling and *using* ACLs (which are built on top of EAs) will impact performance a little
[01:29] <lupus_> does ubuntu create a seperate partition for /home if you select default install?
[01:29] <crimsun> lupus_: no
[01:29] <mjt> your question was like  "if i put this dir there, will it impact on peformance?" -- the answer is "yes, if you will hit the dir hard" ;)
[01:30] <sivang> rehi all
[01:30] <lupus_> isn't it safer for the user if ubuntu would create a seperate partition for /home and maybe /var
[01:31] <lupus_> I remember my slackware box not working anymore after 8gig of logs :) in /var hehe 
[01:31] <lupus_> gig = GB
[01:36] <thully> I can officially say that the timezone bug is GONE!  Just tested w/system clock on local time (and Windows installed) and everything seems to be working fine
[01:37] <lupus_> nice :)
[01:42] <mjt> which $TZ bug it was?
[01:58] <jdub> mdz: when the about ubuntu page is docbook based, the doc team can set up i18n infrastructure for it
[01:58] <jdub> mdz: yelp will handle i18n/docbook
[02:12] <mako> mdz: you still need me?
[02:13] <mdz> mako: no, I was looking for a copy of the keyring from Mataro, but I ended up finding them all on keyservers anyway
[02:13] <mjt> . o O { don't throw away please }
[02:13] <sivang> I am posting this link again :) if anyone has remarks or other stuff, mail me and I'd appriciate it : the new g-s-t pkg should have ubunut's 2 default profiles for creating users, default (unprivileged) and Desktop , link:
[02:13] <sivang> http://muse.19inch.net/~sivan/g-s-t/
[02:14] <mako> mdz: i think it's at people.u.c/~mako/ksp-mataro/
[02:14] <mdz> mako: that page lies and says that it will be there by the time of the keysigning
[02:14] <mdz> but there's no link
[02:14] <mako> ok.. i forgot to turn on the link
[02:14] <mako> i did upload it
[02:14] <mako> let me fix that
[02:15] <mjt> speaking of repeating "postings".. what to do with videocard driver problem in xorg package? Just file a bugreport?
[02:16] <mako> mdz: fixed
[02:16] <daniels> mjt: bug report on xserver-xorg
[02:16] <mjt> (i've a prob with trident_drv - xv extension does not work; while using that driver from xfree-4.3 helps - surprizingly it works with xserver-xorg)
[02:17] <mjt> er
[02:17] <mjt> so the next question is - how to file a bugreport against ubuntu package?  I've a mix of debian/ubuntu right now, and don't even know where's the ubuntu bug tracking system... ;)
[02:18] <mdz> if you have a mix of Debian and Ubuntu, try reproducing your problem with only Ubuntu first
[02:18] <mdz> that is not a supported configuration
[02:19] <mjt> it does not really matter in this case
[02:19] <mdz> booting a Hoary daily live CD would be a good test
[02:19] <mdz> don't be so sure
[02:20] <mjt> eh-heh.. I't be nice to have live CD here... ;)
[02:20] <mjt> for some reason it isn't available in russian stores ;)
[02:21] <mdz> it is available on russian Internet
[02:22] <mdz> http://cdimage.ubuntu.com/daily-live/current/
[02:22] <mjt> sure it is
[02:23] <mjt> 497Mb / 3Kb/s = 165666 sec = 46 hours ;)
[02:23] <mdz> gah, openGL apps are still hanging my laptop
[02:25] <mjt> speaking of benig sure -- what else, except of the xserver package (well, all packages from xorg source, really) can be a problem in this case?
[02:26] <mjt> i see no other possibility
[02:31] <mdz> it is inappropriate to file bugs based on a mixed system
[02:31] <mdz> it is documented extensively that we neither recommend nor support this
[02:33] <daniels> mjt: you have no idea the amount of weird problems that can spring from Debian with half the Ubuntu X.Org packages on them, trust me
[02:34] <mjt> i think i do have that idea... ;)
[02:34] <mjt> been there myself ;)
[02:36] <mjt> this very prob is driver-vs-hardware, not software-vs-software issue
[02:38] <schweeb> mjt: your request would be better received if you partitioned, ran a pure ubuntu hoary system, and THEN checked for the problem, and filed the bugreport
[02:39] <mjt> ok
[02:39] <mdz> you should choose whether you want to run Debian or Ubuntu; neither project will want to receive bug reports from a system with both sets of packages
[02:40] <mjt> i really just wanted to know whenever i can solve the prob directly, with a help from someone more expirienced with that area 
[02:40] <mjt> for now it works for me, after "downgrading" the driver to xfree-4.3, so i'm all set
[02:41] <mjt> no "broken" bugreports (from mixed install/whatever), that is.
[02:43] <mjt> maybe someday i will have some time to debug/fix it myself, who knows..
[02:45] <mdz> thanks for understanding
[02:48] <mjt> oh well
[02:50] <YokoZar> mdz: email sent to ubuntu-devel
[02:51] <mdz> YokoZar: I hope that you don't intend to notify me of each message sent to ubuntu-devel ;-)
[02:51] <YokoZar> ehh, nah.  Just the first big one we were talking about last night
[02:51] <mdz> there are quite a few of them, and I am subscribed to the list, so they are sent to me automatically
[02:52] <robertj> mdz: is there going to be an install updates automatically option on update-manager?
[02:52] <robertj> i see the download option all there by it's lonely self
[02:53] <jdub> robertj: very tentative maybe once some other bits fall into place
[02:53] <mdz> no, I don't expect there will be
[02:53] <mdz> it isn't wise to install Ubuntu updates unattended
[02:54] <jdub> robertj: but vveeeerrrryyy tentative
[02:54] <mdz> especially not on a development branch
[02:54] <robertj> mdz: well on the dev branch I understand, but for stable it seems like a good idea
[02:54] <jdub> mdz: once we've got "you need to re-login" and "you need to reboot" flags for upgrades, letting users turn automagic installs on stable systems would be fairly sane
[02:55] <robertj> jdub: might I add that XP does not behave smartly in this regard
[02:55] <mdz> I disagree; packages are not designed with that in mind
[02:55] <robertj> if you don't respond to the dialog it will automatically reboot your system, which is just great when you are playing a game
[02:55] <HrdwrBoB> haha
[02:55] <jdub> nice one.
[02:56] <mdz> the user takes an explicit action, saying "it is OK if things get a bit weird until I am finished with this operation"
[02:56] <mdz> and things do get weird
[02:56] <HrdwrBoB> auto download is a nice (and already implemented in apt) option
[02:56] <robertj> mdz: security updates should minimize wierdness though
[02:56] <YokoZar> Perhaps we should incorporate more XP annoyances like that into Wine, to make it more compatible.
[02:56] <jdub> firefox upgrade == weird!
[02:56] <jdub> YokoZar! DUDE!
[02:56] <jdub> man, i've been trying to catch you :)
[02:56] <YokoZar> Ah
[02:56] <YokoZar> hey
[02:57] <robertj> installing during an idle time might be nice
[02:58] <HrdwrBoB> what's an idle time?
[02:59] <jdub> mdz: you can answer the autopackage question :)
[02:59] <jdub> HrdwrBoB: when does your screensaver start? :)
[02:59] <robertj> yeah, it's good enough for most users
[02:59] <HrdwrBoB> jdub: yes, but when I come back to my desk and firefox is all bizarre because it updated while I was at lunch
[02:59] <HrdwrBoB> or all the windows have closed
[03:00] <jdub> or you get those wonderful xul errors
[03:00] <jdub> "you can't quit because <!--- OSIUDFOIjdjjjdjdjj cccccxssssshhhhhh..."
[03:00] <robertj> jdub: I just know none of my users run their softwareupdates
[03:01] <robertj> they go out of their way dragging the dialog to the bottom of the screen because they have forgotten their passwords
[03:01] <HrdwrBoB> that's an inherent user problem though
[03:01] <robertj> it is, but there is no good way to remedy it I'm afraid
[03:02] <jdub> lobotomy
[03:02] <HrdwrBoB> big stick
[03:02] <robertj> I can't crond softwareupdate because quicktime has graphical prompts
[03:02] <jdub> electric shock therapy
[03:02] <HrdwrBoB> graphical prompts suck
[03:02] <jdub> there was some dude
[03:03] <robertj> it's great, it even does it on quicktime
[03:03] <jdub> who turned up on the gnome lists a few years ago
[03:03] <jdub> who thought that tamagotchis were the future of user interfaces
[03:03] <robertj> err even does it on 10.3 server
[03:03] <HrdwrBoB> but the problem there is up to the sysadmin locally, not really ubuntu
[03:03] <HrdwrBoB> wtf?
[03:03] <jdub> and encouraged us to use emotional blackmailing as a user interface device
[03:03] <robertj> jdub: haha
[03:03] <robertj> "Do your updates or we will sodomize this hackergotchi"
[03:04] <HrdwrBoB> hahahaha
[03:04] <mdz> jdub: it solves all problems, who can argue with that?
[03:04] <jdub> "i'm sad, you should upgrade me!"
[03:04] <jdub> "i made a mess! fsck this!"
[03:04] <robertj> I swear, I wish there were viruses specially designed to mess with stupid people
[03:05] <HrdwrBoB> there are
[03:05] <robertj> not this lame stuff like we have now, but stuff that really ruined careers by editing word documents andspreadsheets
[03:05] <HrdwrBoB> except smart people have to fix it
[03:05] <robertj> If it was good they could fix it when they took their job ;)
[03:25] <robertj> has pam_keyring been given a oneover?
[03:35] <lupus_> hmm gtkmm 2.5.5 seems to be the only thing missing for using coaster in hoary
[03:44] <YokoZar> How long does it take the wiki to email me after I ask for an account?
[03:59] <jdub> YokoZar: shouldn't take long
[03:59] <YokoZar> jdub: thanks
[03:59] <YokoZar> Ah, it's in there now
[03:59] <jdub> Kamion, lamont, elmo: is there anything useful a normal human can do to help sarge's mipsel woes?
[05:00] <srbaker> are there any prerelease hoary cds?
[05:03] <jdub> cdimage.ubuntu.com
[05:04] <srbaker> excellent.  thanks
[05:05] <srbaker> going to install from the current cd
[07:22] <thully> just closed that annoying timezone bug for good
[07:25] <thully> also, btw - if anyone before remembers my troubles with conflicts between snd-intel8x0m and hsfmodem drivers from linuxant - the latest versions of these resolve the problem without blacklisting anything!
[08:23] <sivang> thully: cool, migh solve the problem for me and my dell lappi ?
[08:24] <aj> daniels: around? (or any other X types?)
[08:27] <fabbione> aj: ?
[08:27] <aj> 915g chipset support on warty?
[08:27] <sivang> anyone have an abiword installation? I have a light emeregency, I need to save an abiword doc (which is in XML) to a .doc or .rtf (shush, god forbid) so it would be printable in a uni's lab...
[08:27] <aj> daniels indicated there was 2d support, but i810 doesn't seem to see anything?
[08:27] <fabbione> iirc no, you need Xorg and 2.6.10 for dri
[08:28] <fabbione> probably the flgrx driver can manage it
[08:29] <aj> don't seem to have such a driver?
[08:29] <aj> do i need to switch to hoary, or?
[08:29] <fabbione> aj: it's in l-r-m
[08:30] <aj> l-r-m?
[08:30] <fabbione> aj: fglrx is the ati binary driver (restricted -> linux-restricted-modules)
[08:30] <fabbione> otherwise you can grab X.org from hoary
[08:30] <sivang> all that I need is for someone to open the file, save it as RTF and that's it ;-)
[08:31] <fabbione> sivang: i can do it
[08:31] <fabbione> sivang: put the file somewhere
[08:32] <sivang> fabbione: Can I send you it using email? they block anything but smtp and pop3 and http here :-(
[08:33] <fabbione> sure
[08:33] <sivang> fabbione: fabbione@ubuntu.com ?
[08:33] <fabbione> yup
[08:33] <aj> modprobe fglrx just says "No such device"
[08:34] <fabbione> aj: i am not 100% sure.. i hate binary crap, but iirc you need to rmmod r128 or something like that
[08:36] <aj> r128's not loaded  
[08:36] <sivang> fabbione: sent
[08:37] <fabbione> aj: according to https://www.ubuntulinux.org/wiki/BinaryDriverHowto there is nothing special you need to do...
[08:37] <fabbione> aj: i am afraid you need more recent stuff
[08:37] <aj> so xserver-xorg from hoary?
[08:37] <fabbione> aj: you need a bit more than the server
[08:38] <aj> ?
[08:38] <aj> beyond what apt pulls in with it?
[08:38] <fabbione> clearly you need the libs too
[08:38] <fabbione> just upgrade the x-window-core virtual package
[08:38] <fabbione> (or something like that)
[08:39] <fabbione> it will pull in everything you need
[08:39] <fabbione> x-window-system-core
[08:39] <fabbione> this one should be enough
[08:42] <sivang> fabbione: lemme know when you've sent , sorry for the troulbe
[08:42] <fabbione> sivang: done.. mail is coming back
[08:42] <sivang> fabbione: thank you!
[08:42] <fabbione> no problem
[08:43] <fabbione> i am back to bed
[08:43] <aj> fabbione: yay, 130M download :(
[08:43] <fabbione> fever is going down but not that much yet
[08:44] <fabbione> aj: only?
[08:44] <fabbione> put the fonts on hold
[08:44] <fabbione> i am pretty sure they get updated too
[08:44] <aj> fabbione: it's mostly python, i think
[08:44] <fabbione> it might reduce a few MB
[08:44] <fabbione> are you dist-upgrading?
[08:44] <sivang> fabbione: get well soon!
[08:44] <aj> fabbione: nah, that's a 400MB download
[08:44] <fabbione> Xorg does
[08:44] <aj> fabbione: easier to wait 10mins than futz more
[08:44] <fabbione> Xorg doesn't need python updated
[08:45] <aj> *shrug* seems to
[08:45] <fabbione> at least... NOT when i packaged it
[08:45] <aj> probably have some python-gnome module installed that wants the old X and the old python or the new X and the new python
[08:45] <fabbione> right
[08:45] <fabbione> that's possible due to the lib split
[08:46] <fabbione> good catch
[08:46] <aj> well, hopefully this'll work; thanks and get well :)
[08:46] <fabbione> aj: good luck.. :)
[08:46] <fabbione> welcome and cya around
[08:46] <sivang> fabbione: c'ya soon, and well :)
[08:46] <fabbione> thanks sivang 
[08:58] <aj> err, that almost worked, except the 1024x768 modeline doesn't seem to exist after a dpkg-reconfigure xserver-xorg?
[09:13] <syn-ack> Man, I thought seb128 was going to be here.
[09:14] <syn-ack> I wanted to thank him for getting that bug fixed in Rhythmbox so quick. I thought that since I filed it as "trivial" it would be on the bottom of the list.
[09:39] <dholbach> morning
[09:39] <syn-ack> good morning.
[09:40] <sivang> dholbach: hey daniel, morning :)
[09:40] <dholbach> sivang: hello sivan!
[09:40] <dholbach> sivang: how did g-s-t go?
[09:42] <sivang> dholbach: I have a source pkg for you ;-) http://muse.19inch.net/~sivan/g-s-t
[09:42] <sivang> dholbach: I have to go now, be online again later, let me know if you managed to use it :)
[09:42] <dholbach> sivang: right, i'll give it a go
[09:51] <sladen>  /whois aj
[09:52] <sladen> it is that aj
[10:17] <Riddell> sladen: no it isn't, it's the other aj
[10:31] <Mithrandir> depends on which aj is the other aj for you.
[10:52] <Kaloz> morning
[11:05] <Kamion> mdz: I don't really see a good reason to rename /install to /boot, and on some architectures it would cause me considerable confusion in the CD scripts, so I'd rather not
[11:06] <Kamion> mdz: I'd rather not conflate directory names on the CD with directory names on an installed system that have a sort-of-similar but not-quite-identical meaning
[12:27] <dholbach> hello seb128
[12:28] <seb128> hi
[01:13] <Kamion> mdz: 12:13 < svenl> Kamion: can you ask Matt to look at the mail titled "Re: Ok to commit ubd patch ?" from december 19 ?
[01:13] <Kamion> mdz: (parted)
[01:22] <thom> jdub: how is beagle, slacker
[01:22] <thom> and where is my netapplet list?
[01:23] <jdub> are you sitting down?
[01:24] <thom> yes
[01:25] <jdub> i attempted to relax this weekend :)
[01:25] <thom> HFSNW!
[01:26] <jdub> it was not all out proper relaxation
[01:26] <jdub> tony put star wars battlefield in pipka's handbag before we left
[01:26] <jdub> oh yes
[01:26] <jdub> btw
[01:26] <jdub> pipka has a handbag
[01:27] <mjg59> Bugger. All the LCA accommodation has gone.
[01:28] <jdub> YokoZar: :-)
[01:29] <thom> sorry, don't believe you
[01:29] <jdub> i'm serious!
[01:29] <jdub> dude, i can get photos
[01:29] <jdub> hold on
[01:29] <thom> you'll have to
[01:29] <mjg59> Serious as cancer?
[01:30] <mjg59> pub
[01:30] <mjg59> Argh
[01:30] <thom> mjg59: dude, you've obviously been drinkin' already ;p
[01:30] <dholbach> *grrrr* why does  gnome_program_init()  crash *cry*
[01:31] <jdub> nah, it would just not be the same if it weren't pipka+handbag
[01:32] <jdub> (she's asleep now)
[01:32] <thom> ahr
[01:32] <thom> oh well; i'll continue to disbelieve you then
[01:43] <dholbach> hai ogra
[01:43] <Kamion> elmo: can http://people.ubuntulinux.org/patches/ be made to work again, please? It used to work, and I mailed some patches to Debian on that basis; now I'm getting grief because they don't work any more.
[01:44] <ogra> dholbach
[01:44] <ogra> hi
[01:44] <Kamion> or for that matter
[01:44] <Kamion> thom: ^---
[01:44] <thom> i'll take a look post breakfast
[01:45] <Kamion> ta
[01:45] <thom> (and yes, i know it's getting on for 1pm :-) )
[01:45] <ogra> hehe...
[01:45] <Kamion> not as if I've had breakfast yet ...
[01:46] <dholbach> anyone who would like to test if coaster-0.1.4 just crashes on my box (amd64) or anyone would like to provide a manpage, while i battle gdb? ;-)
[01:55] <T-Bone> Kamion: hi, got my last message?
[01:56] <dholbach> brb
[01:56] <ogra> i have very bad news....
[01:57] <ogra> to make this work: http://www.grawert.net/hal_cpu_patch.png http://www.grawert.net/hal_acpi1.png http://www.grawert.net/hal_acpi2.png .....
[01:57] <ogra> acpi has to be stared before hal.....which extends the bootime by a second :(
[01:58] <ogra> mdz ^^^
[01:59] <Kamion> T-Bone: which one?
[01:59] <T-Bone> Kamion: the one i sent yesterday, summarizing what's up with ia64 and the 'hal bug'
[02:00] <Kamion> T-Bone: did you get my mail saying that I think it's architecture-independent?
[02:00] <thom> ogra: yeah, that's total crack, i've been following the threads on the hal list
[02:00] <ogra> thom: i backported it to our hal ;)
[02:00] <thom> ogra: i suspect we'll be taking a different approach :-)
[02:01] <T-Bone> Kamion: nop
[02:01] <ogra> thom: before hoary ?
[02:01] <thom> ogra: we'll see
[02:01] <T-Bone> Kamion: ah bummer. You sent it to @parisc-linux.org ?
[02:02] <dholbach> re
[02:03] <Kamion> T-Bone: yeah, think so
[02:03] <T-Bone> Kamion: FtC (*@parisc-linux.org and a few debian machines) is down for the week end. Can you resend it to varenet@esiee.fr please?
[02:03] <Kamion> T-Bone: bounced
[02:06] <T-Bone> Kamion: got it, thx
[02:06] <Mitario> hi everyone
[02:06] <T-Bone> Kamion: i'm definitely _NOT_ running in expert mode
[02:07] <T-Bone> the amd error occurs in non expert mode 3 times
[02:08] <T-Bone> and networking was up. It downloaded a few Releases/Packages files but took an error at the end
[02:08] <T-Bone> and i have NFC what needs to be done to get a driver hotplug-capable :P
[02:08] <T-Bone> for the rest I'll look at it post-breakfast
[02:08] <T-Bone> (and yes it's 2PM) ;}
[02:12] <Kamion> T-Bone: huh, weird
[02:13] <T-Bone> Kamion: was mostly looking like it couldn't fetch some release/package file actually
[02:14] <T-Bone> i'll try to find out in a bit
[02:14] <T-Gone> Kamion: the problem being that aptitude starts immediately, thus making it impossible to read the message
[02:15] <T-Gone> (that's imho a bad thing (tm))
[02:30] <Kamion>   rhythmbox: Depends: gstreamer0.8-mad which is a virtual package.
[02:30] <Kamion>   libopenh323-1.13.2: Depends: libpt-1.6.3 but it is not installable
[02:30] <Kamion>   libpt-plugins-v4l: Depends: libpt-1.6.3 (= 1.6.6.4-5.1) but it is not installable
[02:30] <Kamion> anyone working on that lot?
[02:31] <Kamion> T-Gone: the message is logged in /var/log/base-config.log
[02:31] <zopi> Hi
[02:32] <zopi> is it plan to add gst-ffmpeg on Ubuntu ?
[02:32] <zopi> http://freshmeat.net/projects/gstreamer/?branch_id=55564&release_id=183840
[02:32] <zopi> there is a package for Debian already on alioth
[02:32] <tseng> zopi: its a legality issue, I believe
[02:32] <zopi> pkg-gnome.alioth.debian.org/debian/pool/g/gst-ffmpeg/
[02:33] <zopi> huh ?
[02:35] <dholbach> does anyone know, who Jakob Schurdak is? or if he's on IRC?
[02:36] <Kamion> ah, finally, working networking on the OQO
[02:36] <Kamion> we need linux-wlan-ng in main
[02:37] <jdub> Kamion: woo :)
[02:38] <Kamion> I don't get to use the CD-ROM and networking at the same time, but hey ...
[02:38] <jdub> oh?
[02:38] <Kamion> well, DVD-ROM
[02:38] <Kamion> USB DVD-ROM and USB network adaptor; only one USB slot.
[02:39] <jdub> ahr
[02:39] <Kamion> don't think I have a hub either
[02:39] <jdub> how's everything else? mouse?
[02:40] <Kamion> haven't got the mouse working yet, but one of the guys mailed me an xorg.conf which I'll try in a moment
[02:40] <Kamion> just installing ubuntu-desktop over ssh
[02:40] <jdub> rad!
[02:40] <Kamion> until I had networking working, it was too painful to do any significant text-mode development
[02:40] <Kamion> the keyboard is usable for short periods, but hacking on it sucks
[02:41] <Kamion> just fixed up base-installer for it, since it didn't know about transmeta CPUs
[02:42] <jdub> whoa
[02:42] <Kamion> (wasn't a showstopper, it fell back to 386 anyway, so didn't make a difference on cdrom installs)
[02:45] <trulux> mdz: ping
[02:49] <Kamion> so, are we removing -mad support from rhythmbox, or what?
[02:50] <Kamion> it seems a shame not to have it able to play MP3s if gstreamer0.8-mad is installed; does it have the ability to use plugins?
[02:51] <Kamion> it doesn't seem to actually link against the gst backends ...
[02:51] <tseng> right now muine uses the libs directly to read metadata, and gst/xine for playback
[02:51] <tseng> i think the other gst players are in the same boat
[02:52] <rubenv> Kamion: heh?
[02:52] <rubenv> still plays mp3 here :)
[02:52] <Kamion> I'm really just wondering if rhythmbox can be made to work without gstreamer0.8-mad installed, yet still have the added functionality if gstreamer0.8-mad is installed
[02:52] <Kamion> rubenv: yes, it still depends on gstreamer0.8-mad
[02:52] <Kamion> rubenv: the problem is that rhythmbox is in main while gstreamer0.8-mad is in universe, so it can't stay that way
[02:53] <rubenv> seems like the whole file formats thing should be made modular
[02:53] <tseng> rubenv: thatd be fine, when everyone switches to using gst for reading metadata
[02:54] <rubenv> ain't the xine metadata horribly broken on RB?
[02:54] <jdub> Kamion: rhythmbox doesn't need the gstreamer plugin installed
[02:55] <jdub> Kamion: without it, it doesn't do mp3; with it, it does
[02:55] <rubenv> so basically there's nor problem at all
[02:55] <Kamion> jdub: in that case I'll fix our package to stop depending on it
[02:55] <tseng> oh, i thought he wanted to remove mp3 support more
[02:55] <jdub> Kamion: (this is exactly what we did for warty)
[02:56] <jdub> Kamion: hrm, bad merge?
[02:56] <tseng> jdub: tomboy today? sorry to sound like a broken record
[02:56] <jdub> tseng: ahr!
[02:56] <jdub> tomorrow
[02:56] <tseng> its nearly freeze
[02:56] <jdub> it is very near to bed time here
[02:57] <tseng> alright dude, have a good one.
[02:57] <Kamion> jdub: could be, looking
[02:58] <daniels> fabbione: er, i915 is intel, dude
[03:01] <aj> hoary's xorg with a manually added horizsync/vertrefresh worked fwiw
[03:02] <Kamion> jdub: yeah, looks like a plain merge glitch, fixing
[03:06] <thom> gar, I want mono on amd64
[03:06] <jdub> thom: mint doesn't work?
[03:06] <thom> jdub: bootstrap hell
[03:06] <jdub> heh
[03:06] <jdub> eek
[03:06] <jdub> ah, bedtime.
[03:06] <thom> jdub: 1.1.3 would be the right solution ;-) 
[03:07] <thom> g'night
[03:07] <daniels> aj: bong.  please send over xorg.conf and Xorg.0.log.
[03:07] <daniels> jdub: yes
[03:07] <jdub> thom: my desktop is 2002:dad6:43e3:0:230:1bff:feb3:cd1 :-)
[03:07] <jdub> but my gateway is not routing to it atm
[03:08] <thom> that's something of a show stopper ;-)
[03:08] <thom> my ISP has native v6 support on it's dsl, but i need a new router to cope
[03:08] <thom> uh, its
[03:08] <jdub> shorewall is a showstopper :|
[03:09] <thom> shorewall needs to burn in the utmost fires of hell ;-)
[03:10] <jdub> got a suggestion for a replacement?
[03:10] <thom> unfortunately not
[03:10] <jdub> (i don't write firewall assembly, btw)
[03:10] <jdub> oh
[03:11] <jdub> same problem here whenever i go looking for something
[03:11] <thom> i write firewall assembly when necessary
[03:11] <T-Bone> Kamion: my guess is that there's little chance that making mptscsih hotplugable is a good idead actually: you need it to boot the box no matter what after the installation, and some machines ship with SCSI CD drives
[03:11] <T-Bone> Kamion: imho the right thing to do is to have it preloaded at initrd time
[03:11] <Kamion> T-Bone: we are using hotplug for hardware detection in general
[03:11] <T-Bone> as mptbase is already
[03:11] <Kamion> T-Bone: there's a project in the works to have hotplug running in the initrd to figure out what to load
[03:12] <Kamion> so mptscsih should be made hotpluggable
[03:12] <T-Bone> Kamion: well then you have to tell me how to do that
[03:12] <Kamion> there are others around here far more qualified to do that than I
[03:12] <T-Bone> Kamion: are we trying to increase boot time on ubuntu?
[03:12] <Kamion> mdz knows the basics
[03:12] <Kamion> at least
[03:12] <Kamion> T-Bone: er, no
[03:13] <Kamion> we're trying to make hardware detection not a screaming nightmare
[03:13] <T-Bone> Kamion: well then even tho i can imagine why you want to hotplug everyhting, you'll end up increasing it anyway I'm afraid ;P
[03:13] <Kamion> not significantly
[03:13] <lupus_> is gaim default download folder pointing to Desktop?
[03:13] <Kamion> the number of modules in the initrd is not large, and grepmap speeds it up enough
[03:13] <T-Bone> depends on the hardware
[03:13] <lupus_> like it is ubuntu policy
[03:13] <thom> Kamion: p.u.c/patches/ works now
[03:13] <Kamion> thom: thanks
[03:13] <T-Bone> Kamion: loading the initrd is already a slow operation on some machines
[03:14] <Kamion> look, we're committed to hotplug; none of the other hacks were sufficiently maintainable
[03:14] <Kamion> so the solution to any gripes with hotplug is to fix hotplug. :)
[03:15] <Kamion> I am *certainly* not going back through the installer and undoing all the changes I made to make it use hotplug
[03:15] <Kamion> even if it weren't a hoary goal
[03:15] <T-Bone> heh
[03:16] <T-Bone> that's fine by me. I know how to disable initrd/hotplug when i want to :)
[03:16] <Kamion> look for MODULE_DEVICE_TABLE sections in other drivers; that should be a good starting point
[03:16] <T-Bone> (and i usually do)
[03:16] <Kamion> T-Bone: jbailey is the one working on hotplug in initramfs
[03:17] <T-Bone> Kamion: the base-install.log is absolutely cripled. Is that a dump of what's actually displayed on the screen? It's barely human readable
[03:17] <Kamion> it's generated by 'script'. Yeah, it's hard to read
[03:17] <T-Bone> yeah i'll ask jeff asap
[03:17] <Kamion> doesn't matter though 'cos you're generally only looking at it for debugging :)
[03:18] <T-Bone> i'm surprised since I'm finding error messages there that I haven't even noticed during base install
[03:19] <Kamion> T-Bone: anyhow, hoary's boot with hotplug almost throughout is already faster than most other distributions
[03:19] <T-Bone> There was a prm installing the selected software  One or more packages failed to install. This may be due to bugs in the  packages, or you may be ouof disk space or experiencing some other  problem.
[03:19] <T-Bone> this is what it says
[03:19] <Kamion> right, probably uninstallable ubuntu-desktop
[03:19] <Kamion> this happens a fair bit I'm afraid :(
[03:19] <Kamion> I'm fixing up one of the problems right now, that's the rhythmbox thing
[03:20] <T-Bone> doh
[03:20] <T-Bone> found the culprit
[03:20] <T-Bone>   openoffice.org-debian-files: Depends: openoffice.org-bin (> 1.1.2+1.1.3) which is a virtual package.
[03:20] <T-Bone>   openoffice.org: Depends: openoffice.org-bin (> 1.1.2+1.1.3) which is a virtual package.
[03:20] <T-Bone>   ia32-libs-openoffice.org: Depends: lib32gcc1 (>= 3.4.2-2ubuntu3) which is a virtual package.
[03:20] <T-Bone> now that suprises me alot; i ran apt-get install ubuntu-desktop by hand and it worked just fine
[03:20] <Kamion> oh, um, argh, that's going to be complicated
[03:21] <Kamion> elmo: around?
[03:21] <T-Bone> lol
[03:21] <Kamion> elmo: does your stuff that generates Task: ubuntu-desktop lines use current germinate / otherwise take account of architecture-specific seed entries?
[03:22] <Kamion> elmo: openoffice.org* shouldn't have Task: ubuntu-desktop on ia64
[03:22] <Kamion> T-Bone: if you used 'aptitude install \~tubuntu-desktop', you'd see the problem
[03:23] <T-Bone> correct
[03:23] <T-Bone> just reproduced it
[03:23] <Kamion> T-Bone: we have 90% of the infrastructure for fixing that though, in germinate (colin.watson@canonical.com--2004/germinate--mainline--0 if you want to look); looks like it just needs to be pushed out to one more place
[03:24] <T-Bone> what's that URL?
[03:25] <T-Bone> oh
[03:25] <T-Bone> that's an arch thing?
[03:25] <Kamion> yeah; install bazaar
[03:25] <Kamion> then 'baz register-archive http://people.ubuntu.com/~cjwatson/archives/colin.watson@canonical.com--2004'
[03:26] <Kamion> then 'baz get colin.watson@canonical.com--2004/germinate--mainline--0'
[03:26] <T-Bone> gah
[03:26] <Kamion> (or tla works too)
[03:26] <T-Bone> yeah but it's "arch" you know
[03:26] <T-Bone> ;P
[03:27] <aj> daniels: ?
[03:28] <aj> daniels: what do you want to conf/log for? (is the one that worked useful?)
[03:33] <daniels> aj: if you have to think about touching xorg.conf, it's a bug -- non-working (default) conf/log is useful to me so i can hopefully fix it
[03:34] <thom> seb128: nautilus/gamin is hosed
[03:34] <daniels> night kiddie-winks
[03:34] <dholbach> thom: even with jdub's new upload?
[03:34] <thom> dholbach: how new is new
[03:34] <thom> night daniel
[03:34] <dholbach> thom: 2-3 hours
[03:35] <dholbach> thom: 0.0.21-0ubuntu2
[03:35] <Kamion> T-Bone: (of course, fixing OOo on ia64 would be good, too ... :-) but that may be beyond the resources you have, OOo being the monster it is)
[03:36] <T-Bone> you damn betcha :)
[03:36] <T-Bone> and i hate java
[03:36] <T-Bone> ;)
[03:36] <thom> dholbach: lets try that
[03:38] <seb128> thom: what ?
[03:40] <tseng> morn mxpxpod 
[03:40] <dholbach> hi mxpxpod
[03:40] <mxpxpod> tseng: mornin'... what's up?
[03:40] <mxpxpod> hey dholbach!
[03:41] <tseng> mxpxpod: nm dude, see priv
[03:41] <mxpxpod> tseng: got it :)
[03:41] <Kamion> haggai: seems to be a missing include of <stdio.h> or <cstdio> or whatever in src680-m66/desktop/source/m
[03:41] <Kamion> igration/services/jvmfwk.cxx
[03:41] <Kamion> haggai: (sorry for broken line), judging from the powerpc build log
[03:41] <Kamion> haggai: (openoffice.org2)
[03:42] <mxpxpod> tseng: what has improved in tomboy 0.3.1?
[03:42] <tseng> mxpxpod: see .changes
[03:42] <mxpxpod> tseng: ah, ok
[03:43] <mxpxpod> dholbach: thanks for the translation
[03:43] <mxpxpod> dholbach: I'm going to make a 0.1.4.1 release today
[03:44] <dholbach> mxpxpod: cool! :-)
[03:44] <dholbach> mxpxpod: just wrote a manpage for coaster and got a fully-compiling package without any lintian-warnings - bad thing is: still a segfault in gnome_program_init() :-(
[03:44] <mxpxpod> dholbach: I have a fix for the egg libraries from bugzilla
[03:45] <mxpxpod> dholbach: that's strange... I'd like to see a backtrace
[03:45] <dholbach> mxpxpod: it's very short, i'll send it in a query
[03:45] <rubenv> dholbach: when trying to build coaster pkgs a while ago, i got the same probs
[03:45] <mxpxpod> dholbach: ok
[03:45] <mxpxpod> that's really really strange... I'm on ubuntu hoary and it works fine for me compiling from source
[03:45] <thom> seb128: nautilus doesn't appear to be updating the Desktop when files land in ~/Desktop; I have to open Desktop as a folder and reload it
[03:45] <dholbach> rubenv: were you able to fix it?
[03:46] <rubenv> dholbach: because of upcoming exams i didn't start to look into the source :)
[03:46] <dholbach> rubenv: hmmm, i'll have an exam next week monday too (in fact my last ;-))
[03:46] <mxpxpod> dholbach: could you file a bug?
[03:46] <mxpxpod> in coaster's bugzilla with that trace?
[03:47] <mxpxpod> rubenv: what arch are you on?
[03:47] <rubenv> thom: click on desktop, ctrl-r, saves you time when gamin doesn't do it ;)
[03:47] <rubenv> i386
[03:47] <mxpxpod> hmm
[03:47] <seb128> thom: blame gamin/inotify :)
[03:47] <mxpxpod> powerpc works fine ;)
[03:47] <dholbach> mxpxpod: of course... just wasnt sure, if i broke the libraries or something
[03:47] <thom> rubenv: shrug, i don't really care, i'm just reporting the bug :-)
[03:47] <thom> seb128: heh
[03:47] <thom> seb128: not gtk+, then?
[03:47] <rubenv> dholbach: me tuesday, also last
[03:47] <dholbach> mxpxpod: but inkscape and regexxer worked fine with those g*mm packages
[03:47] <seb128> thom: gnome-session-remove nautilus && GAM_DEBUG=1 nautilus
[03:47] <dholbach> rubenv: i'll have my fingers crossed
[03:48] <mxpxpod> dholbach: hmm
[03:48] <rubenv> dholbach: thx, will do the same
[03:48] <mxpxpod> dholbach: file a bug and I'll take a look at it
[03:48] <mxpxpod> dholbach: that really confuses me
[03:48] <dholbach> mxpxpod: i can easily imagine
[03:48] <mxpxpod> because it's usually powerpc that has problems :)
[03:49] <dholbach> mxpxpod: just the same with amd64 ;-)
[03:49] <mxpxpod> dholbach: what's your repo's address again?
[03:49] <thom> *sigh*; once i g-s-r'd nautilus and reran it, it works. 
[03:49] <dholbach> mxpxpod: http://ubuntu.stufenseite.de
[03:49] <thom> killall nautilus however hadn't cured it
[03:50] <mxpxpod> is gamin replacing fam finally?
[03:50] <dholbach> mxpxpod: but i'll move another coaster package to it later (when i figured that *grmbl* manpage out)
[03:50] <dholbach> mxpxpod: yes, since 0.0.21-0ubuntu2
[03:50] <mxpxpod> dholbach: you should really set up a repo that people can apt-get from
[03:51] <dholbach> mxpxpod: if you tell me how...
[03:51] <tseng> dholbach: second
[03:51] <mxpxpod> dholbach: we can work on it tomorrow
[03:51] <dholbach> mxpxpod: i dont want it to be a lasting institution
[03:51] <tseng> dholbach: ill msg you a quick script
[03:52] <mxpxpod> rubenv: did you build coaster from source or did you use dholbach's pkgs?
[03:53] <rubenv> from source
[03:53] <rubenv> but it's quite some time ago
[03:53] <mxpxpod> rubenv: did you build the *mm libs from source too?
[03:53] <rubenv> (the coaster that still had bakery)
[03:53] <rubenv> mxpxpod: no
[03:53] <mxpxpod> rubenv: ohhh
[03:54] <mxpxpod> rubenv: the reason that crash didn't work was because bakery 2.3.11 had a different virtual method API than 2.3.10
[03:54] <mxpxpod> s/crash didn't work/crashed/
[03:54] <mxpxpod> dholbach: so, rubenv's crash is different from yours
[03:54] <mxpxpod> yours still puzzles me
[03:56] <dholbach> mxpxpod: one guy in the gnome BTS said, he had 0.1.4 working on hoary
[03:56] <mxpxpod> dholbach: yup
[03:56] <dholbach> mxpxpod: but he didnt have the libraries
[03:57] <mxpxpod> dholbach: try building the *mm stuff from source and then building coaster and tell me if you still get that crash
[03:57] <mxpxpod> if you don't, it's something with your libs
[03:57] <mxpxpod> :)
[03:57] <dholbach> strangely enough, it's /usr/lib/libgnome-2.so.0
[04:01] <dholbach> nothing i covered in the packages at all
[04:03] <mxpxpod> bbiab
[04:20] <dholbach> well guys... see you later (in an hour or two)
[04:50] <kent> on the hardwaresupport section of the wiki, I added a page for webcameras. Can some one check it and tell me if i did the right thing? I think I did the right thing, but I can remove it if it was wrong of me.  :)
[04:51] <rubenv> looks fine to me
[04:59] <marcin_ant> hello - short question - is Beagle installable/buildable (from cvs) on hoary?
[05:04] <mx|gone> tseng: has tomboy changed their notification area icon yet?
[05:07] <mjg59> thom: Oh, hi
[05:09] <rubenv> marcin_ant: i heared that jdub was making pkgs somewhere
[05:19] <dholbach> re
[05:21] <srbaker> anyone here doing the ubuntu thing on an ibook?
[05:21] <srbaker> i'm thinking about trading these two notebooks for an ibook g3
[05:21] <marcin_ant> rubenv: ok - thanks - I just was trying to follow procedure described on this page: http://www.ubuntulinux.org/wiki/BeagleInstallHowto
[05:22] <marcin_ant> rubenv: but it doesn't work
[05:22] <sladen> srbaker: tonnes of people
[05:22] <tritium> seb128, did rhythmbox used to say "Paused" when playback is indeed paused?  It looks like fix for #5926 prevents "Paused" from ever being indicated.
[05:23] <srbaker> sladen, what's the verdict on speed?
[05:23] <srbaker> i'm thinking of going from a p3-750/192M to a g3 500/640M
[05:23] <sladen> srbaker: only you can be a judge of that.
[05:23] <mjg59> thom: Yeah, that looks good
[05:24] <srbaker> sladen, that doesn't help much.
[05:24] <seb128> tritium: dunno, the fix has been to drop the broken patch
[05:24] <thom> mjg59: anything to add?
[05:24] <sladen> srbaker: this is probably a question that belongs on #ubuntu (this is a development related channel)
[05:25] <tritium> seb128, I don't recall the behavior from before either
[05:25] <mjg59> thom: Nothing that springs to mind
[05:25] <mjg59> I'll play with the apm stuff later today
[05:25] <seb128> tritium: the current one, since the package has been patched and now the patch has been dropped
[05:25] <seb128> tritium: so it's back to the previous one
[05:26] <tritium> okay, thanks.
[05:28] <seb128> np
[05:28] <srbaker> sladen, i'm interested in hearing it from developers.  i know the difference.
[05:35] <titoo> Hello, any IBM laptop users? Ubuntu is not supporting my internal modem (t41p), I can use my laptop for testing if you have an idea but not the same hardware.
[05:37] <Mithrandir> titoo: I use an IBM laptop, but I couldn't care less about the internal modem. :P
[05:37] <rubenv> laptop now 2grams lighter :)
[05:38] <Mithrandir> ooh, I should do that. ;P
[05:38] <Mithrandir> (as I have an x40, so anything to get it lighter)
[05:38] <rubenv> dell here
[05:38] <Mithrandir> ew.
[05:39] <Mithrandir> I _seriously_ don't like dell due to personal experiences.
[05:39] <rubenv> Mithrandir: what happened?
[05:39] <Mithrandir> rubenv: I wore out two in less than a year, got fucked over a couple of times and wasted a lot of time on trying to fix it.
[05:40] <Mithrandir> (basically)
[05:40] <rubenv> :)
[05:40] <rubenv> (just in case ;))
[05:41] <titoo> Mithrandir: :( I am only needing 1 week a year... forced to use Windows 
[05:42] <Mithrandir> titoo: buy (or just get) a real modem?
[05:43] <titoo> Mithrandir: I am looking at what is done on the Debian side, I may find
[05:45] <Mithrandir> titoo: there are supposed to be some solutions for those stuck with mwave modems, but I've never looked at it myself.
[05:47] <titoo> Mithrandir: True that I can get my old USRobotics out of the dust :)
[05:48] <Mithrandir> titoo: I'd do that if I were you -- probably less work. :)
[05:48] <mjg59> rubenv: It's an hsf modem
[05:48] <mjg59> There's no open driver
[05:49] <mjg59> Uh, s/rubenv/titoo/
[05:51] <titoo> mjg59: Thanks for the info, I understand the support pb...
[05:54] <dholbach> WOOOOOOHOOOOOOOO!!!!!
[05:55] <dholbach> just a sec
[05:55] <dholbach> have to upload it
[05:56] <rubenv> *poke* *poke* *whip* faster i tell ya
[05:57] <dholbach> rubenv: start building the g*mm-libraries unless you have amd64 :-)
[05:57] <rubenv> source debs anywhere? :)
[05:58] <dholbach> http://ubuntu.stufenseite.de 
[05:58] <rubenv> jummy :)
[05:59] <dholbach> i'll try to make  tseng 's script work later; so you'd at least be able to  deb-src  them
[05:59] <rubenv> wget -r for now ;)
[06:01] <dholbach> rubenv: you'd just need the .dsc and .gz files :-)
[06:01] <smurfix> dholbach: isn't the volume name used by ubuntu these days, too?
[06:02] <dholbach> smurfix: ask mx|gone - it's his code :-)
[06:02] <smurfix> That's not code, that's a string in a .po file. ;-)
[06:03] <dholbach> smurfix: you're right, tomorrow there'll be 0.1.4.1 which has german translation :-)
[06:04] <dholbach> i'm so happy that it works :-)
[06:04] <dholbach> s/that//
[06:05] <dholbach> rubenv: you won't need the timer-applet :-)
[06:05] <rubenv> yeah, i noticed it slipping in ;)
[06:07] <dholbach> rubenv: i'll do an 0.1.4-2 upload, which is proper
[06:08] <rubenv> take your time
[06:08] <rubenv> mm stuff first ;)
[06:13] <dholbach> rubenv: nice to have you testing :-)
[06:14] <rubenv> with pleasure :)
[06:14] <rubenv> bleh, my puilder chroot is hopelessly outdated :)
[06:15] <dholbach> rubenv: i never got mine working actually :-(((
[06:15] <rubenv> dholbach: what's the problem?
[06:15] <dholbach> rubenv: it complains about gcc not being able to compile executables
[06:16] <rubenv> eh?
[06:16] <rubenv> i'll tell you what mine does in a sec :)
[06:16] <rubenv> it's been a while since i used it
[06:16] <dholbach> rubenv: but it's the same with each and every .dsc i feed it :-(
[06:17] <thom> bet you don't got build-essential installed 
[06:17] <dholbach> thom: but i wonder why
[06:18] <dholbach> thom: i even told it    EXTRAPACKAGES=build-essential
[06:18] <rubenv> ah i hate ide :)
[06:19] <dholbach> rubenv: ide?
[06:19] <ubernoob> i upgraded to horax, and now x-screen wont start :( i just get a blank screen. anyone know what the problem might be?
[06:19] <ubernoob> *horay
[06:20] <rubenv> dholbach: the beloved hard disk interface we all like ;)
[06:20] <thom> ubernoob: users questions in #ubuntu, please :-)
[06:21] <ubernoob> thom: ok. sorry
[06:22] <dholbach> rubenv: be sure to grab  coaster_0.1.4-2*   - i'm cooking something
[06:22] <rubenv> glibmm is building
[06:25] <rubenv> glibmm builds perfectly :)
[06:26] <Nigelenki> pitti:  ping
[06:28] <rubenv> hmmm, i wonder how to make pbuilder pick up my freshly built pkgs
[06:31] <dholbach> rubenv: pick up?
[06:32] <rubenv> yeah, how to get pbuilder to use my freshly built glibmm
[06:32] <rubenv> must be first time i'm building a fresh dep tree
[06:32] <dholbach> rubenv: i use  debuild && sudo debi
[06:33] <rubenv> normally my libs were in universe
[06:33] <dholbach> well   dpkg-source -x <bla>.dsc; cd <bla>; debuild && sudo debi
[06:34] <rubenv> yeah, but i'm trying to build em in chroot
[06:35] <rubenv> normal building then :)
[06:44] <dholbach> rubenv: you were on i386?
[07:07] <abelli> sladen: ping
[07:07] <abelli> or ding
[07:09] <rubenv> dholbach: yes
[07:10] <dholbach> rubenv: so we'd need someone on powerpc to test it too ;-)
[07:24] <sladen> abelli: is 'ding' the sound a bicycle makes?
[07:24] <abelli> sladen: could be yeah
[08:08] <thom> yay, elmo is doing NEW
[09:00] <mdz> jdub: ping?
[09:31] <YokoZar> jdub: You asked me something earlier (roughly 8 hours ago) ?
[09:34] <sid77> hi
[10:55] <Mithrandir> Kamion: http://err.no/patches/pam_env_per_user.diff ; what do you think of something along those lines to have an ~/.pam_environment which could be touched by the utf8 migration tool (and anything else that wants to handle such a file with a somewhat-predictable format)
[10:57] <Kamion> Mithrandir: going out soon so I haven't reviewed it in detail, but the spirit looks fine; I assume it has the same format as /etc/environment, that an assignment in ~/.pam_environment supersedes any assignment of the same name in /etc/environment, and that if a name isn't mentioned in ~/.pam_environment then the value from /etc/environment is used?
[10:58] <Kamion> I think that'd basically be my spec
[11:01] <mdz> Kamion: what would be the consequences of skipping the location question on the live CD?
[11:01] <mdz> bad keyboard layout defaults?
[11:04] <Mithrandir> Kamion: yeah, that's the idea at least.
[11:04] <Kamion> mdz: you wouldn't get a full locale. I don't want en_US ...
[11:04] <Mithrandir> Kamion: I had to whack pam_env a fair bit to make it suck in different ways -- the code should be completely rewritten, but that's for another day.
[11:05] <Kamion> mdz: language and location are a single udeb now, partly to emphasise the fact that they really can't be split up
[11:29] <mdz> Mithrandir: I don't remember if I heard from you regarding evms-udeb; did you have a chance to look into it?
[11:31] <Mithrandir> mdz: I've looked at it a bit and if we want to support all of what evms is capable of, it's a fair chunk of work (and frankly, I'm not sure we should do it, as the UI will be horrible).  If we just want to do what partman-lvm is doing (and which should be enough for a lot of the use cases we're interested in, it's a couple of days of work.
[11:32] <Mithrandir> mdz: so the answer is "it depends".
[11:33] <mdz> Mithrandir: I think trying to do everything EVMS does through a cdebconf UI would be horrific
[11:33] <Mithrandir> mdz: yes, I agree.
[11:33] <robertj_> hey mdz: do you know anything about gecko#?
[11:33] <mdz> we should provide the basic types of configurations that people want to use on their root filesystem
[11:33] <mdz> robertj_: nope
[11:33] <mdz> e.g., RAID, LVM and RAID+LVM
[11:33] <robertj_> the dashboard guys said jdub was giving it some TLC to work with beagle .5+
[11:33] <Mithrandir> mdz: that ought to be doable with, somewhere around a week of work, I think.
[11:35] <mdz> Mithrandir: could you email me a quote based on this basic spec?
[11:35] <mdz> Mithrandir: how is the utf8 migration tool going?
[11:36] <Mithrandir> mdz: I've found out how to do it and begun hacking the surrounding stuff (like, we probably want http://err.no/patches/pam_env_per_user.diff to have pam_env support ~/.pam_environment.  Else, we'd have to migrate the whole system at once.
[11:36] <mdz> Mithrandir: migrating the whole system at once is fine for the desktop, really
[11:37] <Mithrandir> mdz: I'm designing it to work in two modes -- per user and whole-system.  Whole system will convert all users' home directories, change locale in /etc/environment, stuff like that, while per-user will just do it for one user.
[11:38] <Mithrandir> most of the code will be shared, so that shouldn't be a problem.
[11:45] <Kamion> mdz: (oh, note that if there's only one sane country to pick for that language, then the country question will already be skipped; try e.g. Greek)
[11:46] <mdz> Kamion: oh, ok
[11:47] <Kamion> well, s/skipped/asked at medium priority/
[11:51] <mdz> Kamion: do we use 'high' now, or still 'critical' by default?
[11:51] <Kamion> high
[11:51] <Kamion> fixed some automatic install issues that way
[11:52] <Kamion> mdz: oh, did you see http://people.ubuntu.com/~cjwatson/kickstart.png?
[11:52] <mdz> I was thinking that casper's needs probably correspond fairly closely to what I imagine 'critical' ought to be
[11:52] <Kamion> mdz: yes, very possibly indeed
[11:52] <mdz> re: kickstart.png, no, I hadn't
[11:52] <Kamion> although critical is kind of intended to be used in conjunction with a tuned preseed file, at the moment
[11:53] <mdz> is that a modified version of RH's kickstart tool?
[11:53] <Kamion> yep
[11:53] <Kamion> getting there; should have *something* in by feature freeze
[11:53] <Kamion> though not everything will work yet in that timeframe
[11:53] <mdz> speaking of which, could you make a quick pass over HoaryGoals and make any appropriate status updates for your items?
[11:54] <Kamion> then I have a piece that sits in the initrd, reads kickstart files, and sets the corresponding debconf questions
[11:54] <Kamion> ah, yes, been meaning to do that
[11:54] <mdz> so the output from that tool is a kickstart file, which will be read by some kickstart-aware d-i component and used to provide answers to debconf questions?
[11:54] <Kamion> right
[11:55] <mdz> neat
[11:55] <mdz> what's the normal way to provide the file?
[11:55] <Kamion> should I take myself off the UTF-8 goal, or leave myself there as documentation?
[11:55] <mdz> (the latter)
[11:55] <Kamion> various, you boot with ks=<stuff>
[11:55] <mdz> the documentation team guys will be going through that list at some point to write up the release notes
[11:56] <Kamion> pretty much the same modes of operation as preseed
[11:56] <mdz> and they'll want to contact people for more information
[11:56] <Kamion> there will be a small amount of suckage with regard to things that run before hardware detection; I'm not sure if I'll be able to fix those or whether we'll just have to tell people to add some extra bootloader options for hoary
[11:57] <mdz> ah, hmm
[11:57] <mdz> language/location/keyboard fall into that category, right?
[11:57] <Kamion> yeah
[11:58] <Kamion> preseed has the same problem, so it's not new
[11:58] <Kamion> anaconda basically runs appropriate hardware detection early in kickstart mode
[11:58] <Kamion> which is not necessarily infeasible in d-i, if you accept some messages being untranslated; I haven't tried it out though
[12:00] <Nuak> hi all
[12:00] <Nuak> i want to collaborate in the rosetta project
[12:00] <Nuak> translating into spanish
[12:01] <Kamion> mdz: HoaryGoals updated
[12:01] <Nuak> do you know where i can register myself into the rosetta project?
[12:01] <mdz> Nuak: there is a channel specifically for Rosetta questions/discussion, #rosetta