[11:09] <saispo> hi all
[11:09] <saispo> anyone know why i get this error in log file when i generate a dapper CD : Missing debootstrap-required hfsplus ?
[02:41] <saispo> tepsipakki: have you open the bug report about casper and lvm preseed partitionning ?
[02:42] <cjwatson> evand: thanks, but I think post-feisty at this point
[02:42] <saispo> hi cjwatson :)
[02:43] <saispo> :)
[02:44] <saispo> hi cr3
[02:53] <cjwatson> Fujitsu: fair - it's on the UDS schedule
[02:54] <cjwatson> the only reason it didn't happen for feisty was that cryptsetup and usplash didn't play nice together
[02:54] <cjwatson> saispo: what does casper have to do with LVM preseeding?
[02:54] <saispo> oups, not casper
[02:55] <saispo> about partman-lvm
[02:56] <Fujitsu> cjwatson: Great!
[02:59] <saispo> cjwatson: you have an idea why ? it's not hurry... but for final release it may be good
[03:00] <saispo> with latest updates i haven't got this
[03:00] <cjwatson> saispo: I have no idea what you're talking about
[03:00] <cjwatson> oh, the hfsplus thing?
[03:01] <cjwatson> that means that your archive has hfsplus at Priority: important or required, but it isn't in your minimal seed
[03:04] <saispo> no hfsplug, i correct this things
[03:04] <saispo> cjwatson: about lvm preseeding automatic, it's so slow...
[03:04] <saispo> before i haven't got this things
[03:09] <cjwatson> oh that's known, nothing to do with preseeding
[03:09] <cjwatson> too late to fix for feisty though
[03:10] <cjwatson> bug 105623
[03:10] <cjwatson> it'll just have to go in the release notes, and maybe be point-release material
[03:11] <saispo> ok :/
[03:11] <saispo> thanks :)
[03:12] <cjwatson> evand: let me know when you're around? I need to talk to you about weirdness in the m-a component
[03:13] <cjwatson> evand: specifically bug 105861
[03:59] <cjwatson> evand: panic over, decided it wasn't release-critical
[03:59] <evand> here now
[04:00] <evand> hrm, yikes
[04:03] <cjwatson> I don't think it's quite as weird as I initially thought, but it's weird
[04:03] <cjwatson> why does ma-ask need to be called from ok_handler?
[04:04] <evand> checking
[04:06] <cjwatson> dropping back into a UI loop from cleanup is freaky, but probably just about OK - it's the call of run_command from ok_handler that really makes me blink
[04:07] <evand> I'm still looking it over, but the general overview is that it calls ma-ask the first time to figure out what all the options are to populate the interface, it then calls ma-ask again to see if there were any errors in the options that the user selected.  If that pops up an error question, then it drops back to the interface and sets up an error, otherwise it proceeds to the next page
[04:08] <cjwatson> Hmm. In that case I think it would be best to refactor it so that you can avoid exiting ma-ask in between
[04:08] <cjwatson> (not feisty, but)
[04:09] <cjwatson> if that doesn't fit with the d-i control flow, something I've been known to do is create a wrapper script in scripts/ that does something extra that's ubiquity-specific
[04:10] <evand> Ah, that would make much more sense.  I tried to avoid making ma-ask ubquity specific, but I'm sure that can still be avoided
[04:10] <evand> ah, ok
[04:14] <cjwatson> which might for instance include a spurious db_input/db_go just to give you something to block on
[04:14] <cjwatson> -       mount $device $mountpoint
[04:14] <cjwatson> +       mount $device $mountpoint || error "failed to mount $device"
[04:14] <cjwatson> evand: doing that to ma-script-utils would be good - I have a partition on my test system here where mount fails because it's an LVM PV
[04:14] <cjwatson> which was the root cause of 105861
[04:14] <cjwatson> took me ages to find where the 32 was coming from :) turns out it's EX_FAIL from mount
[04:14] <evand> whoops, I thought I did check that mount succeeded, but apparently not
[04:14] <evand> ah, just in windows
[04:14] <evand> hilarious
[04:14] <cjwatson> and /home
[04:14] <evand> indeed
[04:14] <cjwatson> wah, escaping in summary is broken
[04:14] <evand> right!
[04:14] <evand> I forgot to mention that to you
[04:14] <cjwatson> fuck
[04:14] <evand> I noticed it as well
[04:14] <evand> heh
[04:14] <cjwatson> sadly not release-critical either, I think ...
[04:14] <cjwatson> trivial to fix
[04:15] <evand> I've got to bolt, I have work in less than an hour.  If you happen to cross any more crazyness in m-a, just let me know in here and I'll try to get back to you ASAP.
[04:15] <evand> I assigned that bug to myself.  I'll refactor for Gutsy.
[04:15] <cjwatson> thanks
[04:16] <cjwatson> yeah, it seems to be largely working ok, just some minor glitches
[04:16] <cjwatson> sorry for panicking :)
[04:17] <evand> no worries, I understand completely.  I can see from -devel that things are quite hectic right now for the release team.
[04:17] <cjwatson> just a bit
[04:21] <CIA-16> ubiquity: cjwatson * r2031 ubiquity/ (configure configure.ac): bump to 1.4.11
[04:23] <CIA-16> ubiquity: cjwatson * r2032 ubiquity/ (debian/changelog scripts/summary): * Fix escaping of migration-assistant summary.
[04:43] <CIA-16> ubiquity: cjwatson * r2033 ubiquity/ (debian/changelog scripts/summary):
[04:43] <CIA-16> ubiquity: * Only display partitions with scheduled migrations in the summary, and
[04:43] <CIA-16> ubiquity:  improve the formatting slightly.
[04:47] <CIA-16> ubiquity: cjwatson * r2034 ubiquity/ (debian/changelog scripts/summary): * Remove another stray newline from the migration-assistant summary.
[04:52] <CIA-16> ubiquity: cjwatson * r2035 ubiquity/ (configure configure.ac debian/changelog): bump to 1.5.0; this is all post-feisty
[04:54] <cjwatson> (whee)
[05:00] <CIA-16> ubiquity: cjwatson * r2036 ubiquity/ (debian/changelog ubiquity/frontend/ubiquity.glade):
[05:00] <CIA-16> ubiquity: * GTK frontend: Pressing Enter in the migration-assistant user/password
[05:00] <CIA-16> ubiquity:  entry boxes activates the Forward button.
[05:53] <cjwatson>  15 files changed, 285 insertions(+), 2167 deletions(-)
[05:53] <cjwatson> monkey like
[07:19] <CIA-16> ubiquity: cjwatson * r2032 feisty/debian/ (changelog po/fr.po po/tr.po):
[07:19] <CIA-16> ubiquity: * Fix broken French and Turkish translations of "Go Back" and "Continue"
[07:19] <CIA-16> ubiquity:  (LP: #105903).
[07:51] <CIA-16> ubiquity: cjwatson * r2033 feisty/debian/changelog: releasing version 1.4.11
[08:44] <penta> hi
[08:45] <penta> i have some problems with installing feisty on an existing lvm partition