[11:47] <xivulon> evand, slanga required some testing for 217348 if possible (requires vm with multiple hd)
[11:48] <xivulon> also would it be possible to update wubi on the iso?
[11:48] <xivulon> s/slanga/slangasek/
[13:08] <CIA-48> installation-guide: cjwatson * r424 ubuntu/ (13 files in 2 dirs): Bump kernelversion to 2.6.26 (main architectures) and 2.6.25 (ports).
[13:09] <CIA-48> installation-guide: cjwatson * r425 ubuntu/ (debian/changelog en/welcome/what-is-ubuntu.xml): Bump GNOME version to 2.24.
[13:11] <CIA-48> installation-guide: cjwatson * r426 ubuntu/ (build/entities/common.ent debian/changelog): Bump release version and names for Intrepid.
[13:16] <CIA-48> installation-guide: cjwatson * r427 ubuntu/debian/changelog: releasing version 20080520ubuntu1
[14:28] <evand> xivulon: I'm pulling down the latest ISO to test wubi now, then I'll fill out a SRU for it and ping slangasek when his core hours start.
[14:31] <xivulon> evand, great
[14:32] <davmor2> xivulon: what logs do you need?
[14:33] <xivulon> davmor2: if the error is on first reboot after windows, I need /casper.log
[14:34] <davmor2> cool np's I'll paste in a second along with menu.lst
[14:34] <xivulon> if it is a post-install boot issue you have to boot with the "debug" option and check the log in /tmp
[14:34] <xivulon> as mentioned on testing, please use paste.ubuntu.com for the time being as I cannot access gmail
[14:54] <davmor2> xivulon: http://paste.ubuntu.com/22345/ for menu.lst
[14:55] <davmor2> having issues finding the other log files though :(  not under /var/log/
[14:56] <davmor2> no casper and no installer and nothing under /dist-upgrade
[15:05] <davmor2> xivulon: I'm going to delete the install and try again
[15:07] <xivulon> davmor2: menu.lst looks good
[15:07] <xivulon> as mentioned in a normal installation you have to boot with "debug" then a log will appear under /tmp
[15:08] <xivulon> ps shouldn't debug be a default option in recovery mode? that is certainly handy
[15:09] <xivulon> davmor2: boot with the following kernel line in menu.lst
[15:10] <davmor2> xivulon: need to re-install first
[15:10] <xivulon> no that shouldn't be required
[15:10] <xivulon> good to have a broken setup
[15:10] <xivulon>  /boot/vmlinuz-2.6.24-19-generic root=UUID=78983F43983EFEE6 loop=/ubuntu/disks/root.disk
[15:10] <xivulon>  /boot/vmlinuz-2.6.24-19-generic root=UUID=78983F43983EFEE6 loop=/ubuntu/disks/root.disk debug
[15:10] <davmor2> no I deleted the install so I need to reinstall first 2 ticks
[15:11] <xivulon> ah
[15:11] <xivulon> ^ see the kernel line in menu.lst to get a log under /tmp
[15:12] <davmor2> can I run wubi with --debug?
[15:12] <xivulon> no, that will only stop the frontend every 2 secs
[15:12] <xivulon> it only helps to debug windows side issues, and nowdays does not add much over the wubi log
[15:13] <davmor2> okay np's
[15:13] <xivulon> you seem to have a post-installation issue, so I need the initramfs logs
[15:13] <xivulon> and those are produced if you have debug as one of the kernel boot args
[15:38] <davmor2> xivulon: ha ha I shall add the C:ubuntuinstallation-logs.zip to my website the install fails
[15:39] <evand> argh, wubi install failed.
[15:44] <evand> http://evalicious.com/tmp/wubi-20080623-syslog.txt for the curious
[15:48] <xivulon> davmor2 do you get the same error as evand?
[15:48] <xivulon> also your previous log indicates that installation was completed
[15:48] <xivulon> evand, that is a puls error, can you check why that fails?
[15:49] <davmor2> could be
[15:49] <xivulon> davmor2, so first try the installation went through up to reboot and second try it stopped, did you do anything different?
[15:50] <xivulon> like installing of an ISO as opposed of installing of a CD
[15:50] <xivulon> off
[15:50] <davmor2> no the first time it must of failed but not flagged the fact
[15:51] <evand> xivulon: the error isn't in ubiquity, it just bubbles up to it.
[15:52] <evand> looks like grub-installer and, if I had to guess, the lack of a root device.
[15:52] <evand> I could be wrong about the latter.
[15:54] <davmor2> http://www.davmor2.co.uk/installation-logs.zip
[15:55] <davmor2> evand: xivulon: they're my logs knock yourselves out :)
[16:01] <xivulon> is this using a CD or ISO, i.e. do you have a directory /ubuntu/install/cdrom?
[16:02] <xivulon> the issue seems to be the same as evand
[16:07] <davmor2> xivulon: cd
[16:14] <xivulon> evand can you run grub-installer with set -x?
[16:14] <xivulon> davmor2: can your un on an ISO?
[16:15] <davmor2> I'm just going to try the cd incase it's actually the cd at fault but I can try after that if you want
[16:15] <xivulon> davmor2: unlikely since evand has the same issue
[16:15] <evand> xivulon: already working on it.  The problem appears to be in is_removable.
[16:16] <xivulon> ah
[16:16] <xivulon> removabledevice="$(mount | grep "on $ROOT${1%/} " | cut -d' ' -f1)"
[16:16] <xivulon> removabledevice="$(mount | grep "on $ROOT${1%/} " | cut -d' ' -f1)"
[16:16] <xivulon> probably needs head as well...
[16:16] <xivulon> try to add | head -n 1
[16:17] <xivulon> in my vm test I killed it at copying files and did not get that far... arg...
[16:20] <xivulon> davmor2: please try with an ISO, as I think this is due to grub-installer being confused in the presence of bindmounts
[16:20] <xivulon> and bindmounts are only used when installing from CD
[16:21] <xivulon> evand if I am right findfs and findfstype are also affected
[16:21] <evand> hrm, actually I stand corrected, it might not be there after all.
[16:22] <evand> one moment, I'm going to try an install again with set -x in grub-installer.
[16:22] <xivulon> hmm it might be, see the fix in partman-auto-loop
[16:22] <xivulon> this looks very similar issue
[16:23] <xivulon> see how many lines we get for: mount | grep "on $ROOT${1%/} "
[16:27] <xivulon> hmm nope
[16:28] <xivulon> we are grepping for the mountpoint here and not for the device
[16:29]  * xivulon waits for set -x
[17:53] <CIA-48> debian-installer-utils: cjwatson * r657 ubuntu/ (17 files in 4 dirs): merge from Debian 1.58
[17:59] <xivulon> evand, any progress?
[18:02] <CIA-48> debian-installer-utils: cjwatson * r658 ubuntu/debian/ (changelog rules): merge from Debian 1.59
[18:03] <evand> xivulon: ah, there you are.  I might be very wrong about grub-installer: http://evalicious.com/tmp/wubi-20080623-syslog-2.txt
[18:03] <evand> I'm looking into it.
[18:04] <CIA-48> debian-installer-utils: cjwatson * r659 ubuntu/debian/changelog: releasing version 1.59ubuntu1
[18:04] <xivulon> hmm, davmor2 did you try with an ISO by any chance?
[18:05] <xivulon> ISO worked for me since I did a full test, but in CD mode I only stopped at file copying...
[18:06] <xivulon> so I would tend to think it is related to the bindmount of /cdrom
[18:06] <xivulon> which is the main difference between the 2 setups
[18:21] <davmor2> shall test shortly just had some other stuff to do :)
[18:23] <hardwire> compbrain: hows life?
[18:24]  * hardwire notes that the latest hardy kernels support openvz sans bugs, finally.
[18:49] <evand> hrm, the install succeeded that time.
[18:59] <davmor2> xivulon: installing sp3 I thought all the updates had finished :(
[19:04] <xivulon> evand ? did it work? anything different?
[19:05] <xivulon> did you use ISO or CD?
[19:05] <evand> it worked that time, I had enabled debugging and added some print statements.  Running another install in just debug mode to see if that's enough to get it to succeed.
[19:06] <evand> I'm wondering if it's an apt bug.
[19:12] <davmor2> xivulon: evand: would you believe it xp seems to have a bug ;) It's downloading SP3 and the dl bar vanished :)
[19:18] <stgraber> davmor2: not a problem, last time I gave it a try I never saw a progress bar during the whole install process :)
[19:18] <stgraber> and it actually downloads it when "Preparing for download" ...
[19:19] <davmor2> still a bug ;)
[19:20] <stgraber> report it :)
[19:22] <davmor2> I don't like windows enough to want to improve it ;)
[19:32] <evand> xivulon: very odd, it's not the use of --debug (still reproduceable, not sure why it worked the second time), and it's not an issue in regular installs as far as I can tell.
[19:35] <xivulon> evand could it be that you used an ISO in the second case?
[19:36] <xivulon> as opposed to using a CD? (wubi will eject the CD whether virtual or not)
[19:36] <evand> nope, this is all in vmware
[19:36] <evand> using an emulated CD device
[19:37] <xivulon> do we know we chunk of code generates it?
[19:37] <xivulon> I'll be home in a couple of hours and start playing with it
[19:37] <evand> apt
[19:37] <evand> but as far as what ubiquity code, scripts/install.py
[19:38] <xivulon> is it when chrooting?
[19:39] <xivulon> I'd guess so, since I imagine apt runs in chroot, I mean is the chrooting operation per se' or is it the command run in chroot mode
[19:43] <evand> I don't follow what you're asking.  We tell apt that the root is /target.  I haven't stared at the python-apt code long enough to if it is in fact chrooting.  I don't believe it does though.
[19:44] <evand> a grep for chroot in the source tree seems to prove that.
[19:48] <xivulon> ignore that, have no code under my nose, just blind guesses, I seemed to remember that apt on the target was working in chroot, probably wrong
[21:06] <evand> oh fun, as soon as I try to print self.percent in pulse() the problem disappears.
=== joejaxx_ is now known as joejaxx
[21:10] <davmor2> xivulon: is there a specific folder that the iso needs to be in?
[21:24] <davmor2> evand: does the iso need to be in a specific diretory do you know?
[21:28] <evand> according to the wiki it needs to be in the same folder as the executable.
[21:30] <davmor2> evand: ta :)
[21:33] <cjwatson> apt doesn't actually chroot itself. (Proof: grep for chroot in apt's source code; no matches.) We do tell apt to tell dpkg to chroot, using --root=/target, and we point apt's administrative file locations at /target (though may have missed some).
[21:33] <cjwatson> ah, I see Evan already did the grep proof
[21:46] <davmor2> xivulon: iso has gone further than the cd install current 94% should finish shortly
[21:51] <davmor2> xivulon: iso seems to of worked without fault