[11:54] <cjwatson> jibel: Has bug 1059619 been reproducible with any image since 2012-10-08?  I have a suspicion that an image-building change may have avoided it.
[11:54] <ubot2> Launchpad bug 1059619 in grub-installer "Installation hangs on: grub-mount /dev/sdb1 /var/lib/os-prober/mount" [High,Confirmed] https://launchpad.net/bugs/1059619
[11:55] <cjwatson> Which would be nice.
[12:14] <jibel> cjwatson, I haven't reproduced it with yesterdays and todays images on i386 and amd64 and the 2 machines I had this problem. I'll close it.
[12:15] <cjwatson> Yay
[12:16] <cjwatson> xnox: Is the udisks2 port of usb-creator still happening?
[12:17] <xnox> cjwatson: I didn't start it. Did read a bit about udisks2. Something needs to happen: e.g. wrap usb-creator to startup udisks1 bus... or port to udisks2.
[12:17] <xnox> as far as I can see, usb-creator is not usable in quantal right now.
[15:59] <bdmurray> xnox: barry and I were able to use usb-creator in quantal
[16:00] <cjwatson> I think it depends a bit on whether udisks-daemon manages to start, doesn't it?
[16:00] <xnox> bdmurray: interesting. I wasn't. I got "error communicating" with something over dbus...
[16:00] <stgraber> cjwatson: well, I was really hoping not to have to write that preseed generator in perl, but it turned out to be the easiest way to do it (based on debconf-get-selections). I'm getting something that looks reasonable generated now. Will do some tests this afternoon.
[16:01] <xnox> stgraber: \0/
[16:01] <stgraber> cjwatson: I'm basically skipping any entry where the current value is identical to the default value and I'm skipping all partman-* keys
[16:01] <cjwatson> Neat.  I suspect it will need a bit more fine-tuning than that, though - you'll probably want to go over it rather carefully against the installation-guide
[16:01] <stgraber> that gives me a 50 lines long preseed file for a standard server install with just ssh selected in tasksel, so that seems reasonable
[16:02] <xnox> stgraber: curious about skipping partman-* keys. can we not select some partman-auto key there?
[16:02] <stgraber> cjwatson: yeah, the blacklist should be flexible enough to block any other yet we want. Already spotted a few more.
[16:03] <stgraber> xnox: the plan for the initial implementation was not to touch partman, for the final implementation we'll probably want at least partman-auto to work
[16:03] <stgraber> xnox: not sure how easily we can reconstruct a reasonable recipe post-install though, so manual may remain unsupported
[16:03]  * xnox was pondering about "partman" mini-settings: default to ext4/xfs/btrfs/crypto/lvm/raid.
[16:05] <xnox> cause apparently partman is the hardest bit in pre-seeding, especially all the various confirm questions partman asks much depending on what happens to be there on the installed system.
[16:05] <cjwatson> partman/default_filesystem does the fs default
[16:06] <cjwatson> kickseed just slams in a bunch of confirm preseeds
[16:06] <cjwatson> you could check what it does
[17:20] <bdmurray> cjwatson: so I found 2 ubiquity bugs with the fixed version of apt and using ubiquity 2.12.5 where there was an error message from pluginstall regarding 'Problem with MergeList'
[17:22] <bdmurray> I was wondering if there were any changes in ubiquity that might have dealt with apt lists
[17:23] <bdmurray> one of them is bug 1063261
[17:23] <ubot2> Launchpad bug 1063261 in ubiquity "Ubuntu12.10 Installer crashed - "/usr/share/ubiquity/plugininstall.py", line 1705, in <module>" [Undecided,Confirmed] https://launchpad.net/bugs/1063261
[17:24] <cjwatson> I don't think it touches apt lists directly
[17:24] <cjwatson> Fully booked right now though